DARDEN BUSINESS PUBLISHING

UNrvERsnYcfVIRGINIA

UVA-F-1152 Version 21

TELETECH CORPORATION 1996

Raider Dials Teletech

Wake-Up Call Needed Says Investor

New York (AP)--The reclusive billionaire Victor Y ossarian has acquired a 10 percent stake in Teletech Corporation and has demanded two seats on the firms board of directors The purchase was revealed yesterday in a filing with the Securities and Exchange Commission and separately in a letter to Teletechs CEO Maxwell Harper The firm is misusing its resources and not earning an adequate return the letter said The company should abandon its misguided entry into computers and sell the Product and Systems Segment Management must focus on creating value for shareholders Teletech issued a brief statement emphasizing the virtues ofa link between computer technology and telecommunications

Wall Street Daily News January 9 19

Margaret Weston Teletechs chief financial officer learned ofYossarians letter late one evening in early January 1996 Quickly she organized a team of lawyers and finance staff to assess the threat Maxwell Harper the firms CEO scheduled a teleconference meeting of the firms board of directors the next afternoon Harper and Weston agreed that before the meeting they needed to fashion a response to Yossarians assertions about the firms returns

Ironically returns had been the subject of debate within the firms circle ofsenior managers in recent months A number ofissues had been raised about the hurdle rate used by the company in evaluating performance and in setting the annual capital budget Since the company was expected to invest nearly $2 billion in capital projects in 1996 gaining closure and consensus on these issues had become an important priority for Margaret Weston Now Yossarians letter lent urg~ncy to the discussion In the short run she needed to respond to Y ossarian In the long run she needed to assess the competing viewpoints and recommend new policies as necessary What should be the hurdle rates for

Teletechs two business segments Was the Products and Systems segment really paying its way

This case was written by Robert F Bruner and is dedicated to the memory of Professor Robert F Vande II a scholar in I

corporate finance and investment analysis and the author of an antecedent case upon which the present case draws Teletech Corporation is a fictional company reflecting the issues facing actual firms and is used as a basis for class discussion rather than to illustrate effective or ineffective handling ofan administrative situation The financial support ofthe Batten Institute is gratefully acknowledged Copyright (0 1997 by the University of Virginia Darden School Foundation Charlottesville VA All rights reserved To order copies send an e-mailtosalesdardenpublishingcom No part ofthis publication may be reproduced stored in a retrieval system used in a spreadsheet or transmitted in any form or by any means-electronic mechanical photocopying recording or otherwise-without the permiSSion ofthe Darden School Foundation Rev1201 0

Jgt~gt6raquo0 Distributed by The European Case Clearing House England and USA if J North America phone +17812395884 tax +17812395885 e-mail ECCHBabsonaocom 6 t Rest oftbe Wodd phone +44 (0)1234750903 tax +44 (0)1234 751125 e-mail ECCHcrnnfieldacuk ~Jc g~-- All rights reserved Printed in UK and USA Web Site httpwwweccncraruieldacuk

-2shy UVA-F-1152 r The Company

Teletech Corporation headquartered in Dallas Texas defined itself as a provider of integrated information movement and management The firm had two main business segments Telecommunications Services and the manufacture of computing and telecommunications equipment a segment named Products and Systems In 1995 Telecommunications Services had earned a return on capital (ROC)i of98 percent Products and Systems had earned 120 percent The firms current book value of net assets was $16 billion consisting of $114 billion allocated to Telecommunications Services and $46 billion allocated to Products and Systems An internal analysis suggested that Telecommunications Services accounted for 75 percent of the market value ofTeletech while Products and Systems accounted for 25 percent The current capital expenditures proposed by Telecommunications Services offered prospective internal rates of return averaging of 98 percent the IRR for prospective Products and Systems projects averaged Teletech Share Prices vs Market and Industry 120 percent Overall it Indexes

appeared that the firms 170 -x-SampP100prospective return on capital

would be 1035 percent 150 ~ -0- Telephone Top management applied a 130 _ ~ i-a- TelecomEq

_x-~x I i

hurdle rate of 1041 percent 110 ~ - _ +-+-+-+--+ I-o-Corrputers +-+a+-+-+-+=-+- I i

to all capital projects and in 90 I f I I I I I I I I I I 1-+- Teletech iL-____

evaluating the performance OJOJlt tq )~ tcent) olJ-middot t

of business units laquo~

MonthOver the past 12 months the firms shares had not kept pace with the overall stock market indices or with industry indexes for telephone equipment or computer stocks Securities analysts had remarked on ~firm s lackluster earnings growth pointing especially to inshy

creasing competition in tele-communications as well as disappointing performance in the Products and Systems segment A prominent commentator on television opined that there was no precedent for a hostile takeover of a telephone eompany but in the case of Teletech there is every reason to try

Teletechs Telecommunications Services segment

The Telecommunications Services segment provided long-distance local and cellular telephone service to more than 7 million customer lines throughout the Southwest and Miltiwest Revenues in this segment grew at an average rate of 3 percent over the 1989-95 period Ifi~IIf~S segment revenues net operating profit after tax (NOPAT) and net assets were $11 billion $118

I Return on capital is calculated as the ratio of net operating profits after tax (NOPAT) to capital

o

UVA-F-1152

billion and $114 billion respectively Since the court-ordered breakup of the Bell System telephone monopoly in 1983 Teletech had coped with gradual deregulation of its industry through aggressive expansion into new services and geographic regions Most recently the firm had been a leading bidder for cellular telephone operations and for licenses to offer personal communications services (PCS) In addition the firm had purchased a number of telephone operating companies in privatization auctions in Latin America Finally the firm had invested aggressively in new technology-primarily digital switches and optical-fiber cables-in an effort to enhance its service quality All of these strategic moves had been costly the capital budget in this segment had varied between $15 and $2 billion in each of the previous 1 degyears

Unfortunately profit margins in the telecommunications segment had been under pressure for several years Government regulators had been slow to provide rate relief to Teletech for its capital investments Other leading telecommunications providers had expanded into Teletechs geographic markets and invested in new technology and quality enhancing assets Teletechs management noted that large cable TV companies might enter the teleconununications market and continue the pressure on margins

On the other hand Teletech was the dominant service provider in its geographic markets and product segments Customer surveys revealed that the company was the leader in product quality and customer satisfaction Teletechs management was confident that the company could conunand premium prices however the industry might evolve

Teletechs Products and Systems segment

Before 1990 teleconununications had been the companys core business supplemented by an equipment-manufacturing division that produced teleconununications components In 1990 the company acquired a leading computer workstation manufacturer with the goal ofapplying state-ofshythe-art computing technology to the design of telecommunications equipment The explosive growth in the microcomputer market and the increased use oftelephone lines to connect home- and office-based computers with mainframes oonvinced Teletech management ofthe potential value of marrying telecommunications equipment and computing technology Using Teletechs capital base borrowing ability and distribution network to catapult growth the Products and Systems segment increased its sales by nearly 40 percent in 1995 This segments 1995 NaPAT and net assets were $480 million and $46 billion respectively

Products and Systems was acknowledged to be a technology leader in the industry While this accounted for its rapid growth and pricing power maintenance of that leadership position required sizable investments in RampD and fixed assets The rate of technological change was increasing as witnessed by sudden major write-offs by Teletech on products that until recently management had thought were still competitive Major computer manufacturers were entering into the teleconununications-equipment industry Foreign manufacturers were proving to be stiff competitors in bidding on major supply contracts

-4- UVA-F-1l52 t

Focus on Value at Teletech

Teletechs mission statement said in part

We will create value by pursuing business activities that earn premium rates oreturn

Translating that statement into practice had been a challenge for Margaret Weston First it had been necessary to help managers of the segments and business units understand what create value meant for them Because the segments and smaller business units did not issue securities into the capital market the only objective measures of value were the securities prices of the whole corporation-but the activities ofany particular manager might not be significant enough to drive Teletechs securities prices Therefore the company had adopted a measure of value creation for use at the segment and business unit level that would provide a proxy for the way investors would view each units performance This measure called economic profit multiplied the excess rate of return of the business unit times the capital it used

Economic Profit = (ROC - Hurdle Rate) x Capital Employed

Where

NOPATROC = Return on Capital = --shy

Capital

NOPAT = Net Operating Profit After Taxes

Each year the segment and business unit executi ves were measured on the basis ofeconomic profit This measure was an important consideration in strategic decisions about capital allocation

manager promotion and the awarding of incentive compensation

A second way in which the value creation perspective influenced managers was in the assessment of capital-investment proposals For each investment projected cash flows were discounted to the present using the firms hurdle rate to give a measure of the net present value (or NPV) of each project A positive (negative) NPV indicated the amount by which the value of the firm would increase (decrease) ifthe project were undertaken The following equation shows how the hurdle rate was used in the familiar NPV equation

n Net Present Value E [ Free Cash FIOWI I] - Initial Investment

(1 + Hurdle Rate) t 1

Hurdle Rates

The hurdle rate used in the assessments of economic profit and NPV had been the focus of considerable debate in recent months This rate was based on an estimate of Teletechs weighted average cost of capital (WACC) Management was completely satisfied with the intellectual ~

c

-5- UVA-F-1152

relevance of a hurdle rate as an expression of the opportunity cost of money The notion that the W ACC represented this opportunity cost had been debated Its measurement was never considered wholly scientific but it had been accepted For instance Teletech was split-rated between AAshyand A+ An investment banker recently suggested that at these ratings new debt funds might cost Teletech 703 percent (about 422 percent after a 40 percent tax rate) With a beta of 1041 the cost of equity might be about 1177 percent At market-value weights of 18 percent for debt and 82 percent for equity the resulting W ACC would be 10Al percent Exhibit 1 summarizes the calculation The hurdle rate of lOA1 percent was applied to all investment and performanceshymeasurement analyses in the firm

Arguments for risk-adjusted hurdle rates

How the rate should be used within the company in evaluating projects was another point of debate Given the different natures ofthe two businesses and the risks each one faced differences of opinion arose at the segment level over the appropriateness of measuring all projects against the corporate hurdle rate of 10Al percent The chief advocate of multiple rates was Rick Phillips executive vice president ofTelecommunications Services who presented his views as follows

Each phase ofour business is different must compete differently and must draw on capital differently Until recently telecommunications was a regulated industry and the return on our total capital highly certain given the stable nature ofthe industry Because of the recognized safety of the investment many telecommunications companies can raise large quantities ofcapital from the debt markets In operations comparable to Telecommunications Services 75 percent of the necessary capital is raised in the debt markets at interest rates reflecting solid AA quality on averageshythis is better than the corporate bond rating ofAA- A+ Moreover I have to believe that the cost of equity of Telecommunications Services is lower than for Products and Systems I contrast this with the Products and Systems segment where although sales growth and profitability are strong risks are high Independent equipment manufacturers are financed by hi~her yield BBB-rated debt and more equity with higher expected total returns

In my book the hurdle rate for Products and Systems should reflect these higher costs offunds Without the risk-adjusted system of~dle rates Telecommunications Services will gradually starve for capital while Products and Systems will be forceshyfed-thats because our returns are less than the corporate hurdle rate and theirs are greater Telecommunications Services lowers the risk ofthe whole corporation and should not be penalized

Heres a rough graph of what I think is going on Telecommunications Services which can earn 98 percent on capital is actually profitable on a risk-adjusted basis even though it is not profitable compared to the corporate hurdle rate The triangle shape on the drawing shows about where Telecommunications Services is located

-6- UVA-F-1152 _ (

My hunch is that the reverse is true for Products and Systems which promises to earn 120 percent on capital P+S is located on the graph near the little circle

Constant vs Risk-Adjusted Hurdle Rates

17

- Teletech Corp Hurdle

-- Telecomm Services

-e- A-oducts and Systems

Risk Level

In deciding how much to loan us lenders will consider the composition of risks If money flows into safer investments over time the cost of their loans to us will decrease

Our stockholders are just as much concerned with risk Ifthey perceive our business as being more risky than other pound9mpanies they will not pay as high a price for our earnings Perhaps this is why our priceearnings ratio is below the industry average most ofthe time It is not a question ofwhether we adjust for risk-we already do infonnally The only question in my mind is whether we make these adjustments systematically or not

While multiple hurdle rates may not reflect capital-structure changes on a day -to-day basis over time they will reflect prospects more realistically At the moment as I understand it our real problem is an inadequate and very costly supply of equity funds If we are really rationing equity capital then we should be striving for the best returns on equity for the risk Multiple hurdle rates achieve this objective

Implicit in Phillipss argument as Weston understood it was the notion that ifeach segment in the company had a different hurdle rate the costs ofthe various fonns ofcapital would remain the same However the mix of capital used would change in the calculation Low-risk operations ~

c

-7- UVA-F-1152

would use leverage more extensively while the high-risk divisions would have little or no debt funds This lower-risk segment would have a lower hurdle rate

Opposition to risk-adjusted hurdle rates

Phillipss views were supported by several others within Teletech opposition was just as strong however particularly within the Products and Systems segment Helen Buono executive vice president for the segment expressed her opinion as follows

All money is green Investors cant know as much about our operations as we do To them the firm is an opaque box they hire us to take care ofwhat is inside the box andjudge us by the dividends coming out of the box We cant say that one part of the box has a different hurdle rate than another part ofthe box ifour investors dont think that way Like I say all money is green all investments at Teletech should be judged against one hurdle rate

Multiple hurdle rates are illogical Suppose that the hurdle rate for Teleshycommunications Services was much lower than the corporatewide hurdle rate Ifwe undertook investments that met the segment hurdle rate we would be destroying shareholder value because we werent meeting the (orporate hurdle rate shy

Ourjob as managers should be to put our money where the returns are best A single hurdle rate may deprive an underprofitable division of investments in order to channel more funds into a more profitable division but isnt that the aim of the process Our challenge today is simple we must earn the highest absolute rates of return we can get

In reality we dont finance each division separately The corporation raises capital based on its overall prospects and record The diversification of the company probably helps keep our capital C0sts down and enables us to borrow more in total than the sum of the capabilities of the divisions separately As a result developing separate hurdle rates is both unrealistic and misleading All our stockholders want is for us to invest our funds wisely in order to increase the value of their stock This happens when we pick the most promising projects~ irrespective of their source

Margaret Westons Concerns

As she listened to these arguments presented over the course of several months Weston became increasingly concerned with several related considerations First the corporate strategy directed the company toward integrating the two divisions One effect ofusing multiple hurdle rates would be to make justifying high-technology research and application proposals more difficult as the required rate ofreturn would be increased Perhaps she thought multiple hurdle rates were the

-8- UVA-F-1152 r f

right idea but the notion that they should be based on capital costs rather than strategic considerations might be wrong On the other hand perhaps multiple rates based on capital costs should be used but in allocating funds some qualitative adjustment should be made for unquantifiable strategic considerations In Westons mind theory was certainly not clear on how to achieve strategic objectives when allocating capital

Second using a single measure ofthe cost ofmoney (the hurdle rate or discount factor) made the net present value results consistent at least in economic terms If Teletech adopted multiple rates for discounting cash flows Weston was afraid the NPV and economic profit calculations would lose their meaning and comparability across business segments To her a performance criterion had to be consistent and understandable or it would not be useful

In addition Weston was concerned with the problem of attributing capital structures to divisions In Telecommunications Services a major new switching station might be financed by mortgage bonds But in Products and Systems it was not possible for the division to borrow directly indeed any financing was feasible only because the corporation guaranteed the debt Such projects were considered highly risky-perhaps at best warranting only a minimal debt structure Also Weston considered the debt-capacity decision difficult enough to make for the corporation as a whole let alone for each division Judgments could only be very crude

In further discussions with those in the organization about the use of multiple hurdle rates Weston ran across two predominant trains of thought One argument held that the investment C decision should never be mixed with the financing decisionr A firm should decide what its investments should be and then how to finance them most efficiently Adding leverage to a presentshyvalue calculation would distort the results Use ofmultiple hurdle rates was simply a way ofmixing financing with investment analysis This argument also held that a single rate left the risk decision clear-cut management could simply adjust its standard (NPV or economic profit) as risks increased

The contrasting line of reasoning noted that the weighted-average cost of capital tended to represent an average market reaction tol mixture ofrisks Lower-than-average-risk projects should probably be accepted even though they did not meet a weighted-average criterion Higher-than-normal-risk projects should provide a return premium While the multiple-hurdle-rate system was a crude way ofachieving this end it at least was a step in the right direction Moreover some argued that Teletechs objective should be to maximize return on equity funds and because

equity funds were and would remain a comparatively scarce resource a multiple-rate system would tend to maximize returns to stockholders better than a single-rate system

To help resolve these questions Weston asked her assistant Bernard Ingles to summarize academic thinking about multiple hurdle rates His memorandum is given in Exhibit 2 She also requested that he draw samples ofcomparable firms for Telecommunications Services and Products and Systems that might be used in deriving segment WACCs The summary of data is given in Exhibit 3 Information on capital-market conditions in January 1996 is given in Exhibit 4

-9- UVA-F-1l52

Conclusion

Weston could not realistically hope that all the issues before her would be resolved in time to influence Victor Yossarians attack on management But the attack did dictate the need for an objective assessment of the performance of Teletechs two segments-the choice of hurdle rates would be very important in this analysis However she did want to institute a pragmatic system of appropriate hurdle rates (or one rate) that would facilitate judgments in the changing circumstances Teletech faced What were the appropriate hurdle rates for the two segments Was Products and Systems underperforming as Yossarian suggested How should Teletech respond to the raider

-10- UVA-F-1152 shy

Exhibit 1

TELETECH CORPORATION 1996

Summary of W ACC Calculation for Teletech Corporation and

Segment Worksheet

Telecommunications Products and

Co]rate Services SYstems

MY Asset Weights 100 75 25

Bond Rating AA-A+ AA BBB-

Pre-tax Cost of Debt 703 700 778

Tax Rate 40 40 40

After-tax Cost of Debt 422 420 467

Equity Beta 104

Rf 604

RM-Rf 550

Cost of Equity 1177

Weight of Debt 180

Weight of Equity 820

WACC 1041 I

c

UVA-F-1152

Exhibit 2

TELETECH CORPORATION 1996

Theoretical Overview of Multiple Hurdle Rates

To Margaret Weston From Bernard Ingles Subject Theory of Segment Cost of Capital Date January 1996

You requested an overview of theories about multiple hurdle rates Without getting into minutiae the theories boil down to the following points

1 The central idea is that required returns should be driven by risk This is the dominant view in the field of investment management and is based on a mountain of theory and empirical research stretching over several decades The extension of this idea from investment management to corporate decision making is straightforward p~ least in theory

2 An underlying assumption is that the firm is transparent (Le that investors can see through the corporate veil and evaluate the activities going on inside) No one believes firms are completely transparent or that investors are perfectly informed But financial accounting standards have evolved toward making the firm more transparent And the investment community has grown tougher and sharper in its analysis Teletech now has 36 analysts publishing reports and forecasts on the firm The reality is that for big publicly held firms transparency is not a bad assumption

3 Another underlying assumption is that the value of the whole enterprise is simply the sum of its parts-this is the concept of Value Additivity We can define parts as either the business segments (on the left-hand side of the balance sheet) or the layers of the capital structure (on the right-hand side of the balance sheet) Market values (MVs) have to balance

MVTeteedl = ( MVTetlaquoomllllltlicaliolls Services + MV tmdmls+ iwellls) = ( MVdehl + MVltquiV)

Ifthese equalities did not hold then a raider could come along and exploit the inequality by buying or selling the whole and the parts This is arbitrage By buying and selling the actions of the raider would drive the MVs back into balance

-12- UVA-F-1152 r ~

Exhibit 2 (continued)

4 Investment theory tells us that the only risk that matters is nondiversifiable risk which is measured by beta Beta indicates the risk that an asset will add to a portfolio Because the investor is assumed to be diversified she is assumed to seek a return for only that risk that she cannot shed the nondiversifiabIe risk Now the important point here is that the beta ofa portfolio is equal to a weighted average of the betas of the portfolio components Extending this to the corporate environment the asset beta for the firm will equal a weighted average of the components of the firm-again the components of the firm can be defined in terms of either the right-hand side or the left-hand side of the balance sheet

Where

W = percentage weights based on market values

fJ Tel Serv fJ ps Asset betas for business segments

fJdebl fJ for the firms debt securities

fJeqUlIY= fJ of firms common stock (given by Bloomberg etc)

This is a very handy way to model the risk ofthe firm for it means that we can use the Capital Asset Pricing Model to estimate the cost of capital for a segment (ie using segment asset betas)

5 Given all the previous points it follows that the weighted average of the various costs ofcapital (K) for the firm (W ACC) which is the theoretically correct hurdle rate is simply a weighted average of segment W ACCs

Where

WTeCf WIS = market value weights

WACCTeLrv = (Wdebl TeLerv( K debl TelServ) + ( Weqllily TelServt K eltIUlly TelServ)

c

-13- UVA-F-1152

Exhibit 2 (continued)

6 The notion in point 5 may not hold exactly in practice First most of the components in the WACC formula are estimated with some error Second because oftaxes information asymmetries or other market imperfections assets may not be priced strictly in line with the model-for a company like Teletech it is reasonable to assume that any mispricings are just temporary Third the simple two-segment characterization ignores a hidden third segment the corporate treasury department that hedges and aims to finance the whole corporation optimally-this acts as a shock absorber for the financial policies of the segments Modeling the W ACC ofthe corporate treasury department is quite difficult Most companies assume that the impact of corporate treasury isnt very large and simply assume it away As a first cut we could do this too though it is an issue we should revisit

Conclusions

bull In theory the corporate WACC for Teletech is appropriate only for evaluating an asset having the same risk as the whole company It is not appropriate for assets having different risks than the whole company

bull Segment WACCs are computed similarly to corporate W ACCs

bull In concept the corporate W ACe is aweighted average of the segment W ACCs In practice the weighted-average concept may not hold due to imperfections in the market andor estimation errors

bull If we start computing segment W ACCs we must use the cost of debt cost of equity and weights appropriate to that segment We need a lot of information to do this correctly or else we really need to stretch to make assumptions

-14- UVA-F-1152_

Exhibit 3 TELETECH CORPORATION 1996

Samples of Comparable Firms

1995 Equity Asset Bond Book Val Pricel Mkt Val Mkt Val Pricel Revenues Seta Seta Iiaiiog OebtJCap Book DebtCap DebtlEq Eacniogs

Teletech Corporation $16000 1041 092 AA-tA+ 40 301 18 22 129

Telecommunications Services Industty ATampT $80000 090 085 AA 39 660 88 97 308 Alltel Corp 3160 075 063 A 49 299 243 322 160 Amerltech 13325 075 067 AA 47 472 158 188 169 BeD Atlantic 13500 080 068 AA 57 453 226 293 175 Bell South 17780 075 072 AAA 44 1190 62 66 190

Centul) Tel Enterprs 625 100 084 BBB+ 46 263 245 324 158

Cincinnati Bell 1350 080 069 AA 56 472 212 270 188 Citizens Utilities Co 1070 070 056 AA 42 168 301 431 158

Comsat 850 095 068 A 40 103 392 646 168

Frontier Corp 1750 080 074 A 42 550 116 132 283

GTE Corp 20250 080 066 BBB 69 652 254 341 169

MCI Communications 15100 125 114 A 24 187 144 169 179 NYNEX Corp 13425 075 059 Ashy 62 375 303 436 156

Pacific Telesis 9070 085 068 AAshy 74 694 291 410 136

SBC Communications 12560 090 080 A 54 559 173 210 180

Southem New England 1840 075 058 AA 5710 263 335 504 154

Sprint Corp

US West

Average

13550

9450

105

065

084

087

052

072

BBB

AAshy

52

66

51

313

467

452

257

294

228

347

416

311

150

143 C179

Telecommunications Equipment Industty ADC Telecomm Inc $586 135 135 NR 0 405 00 00 280

Aane-Cleveland 120 150 149 NR 1 151 07 07 128

ADen Group 325 160 155 NR 13 275 52 54 188

Andrew Corp 626 125 123 NR 13 440 33 34 197

DSC Communications 1450 130 126 NR 18 372 56 590 179

Newbridge NetworIs 675 155 155 NR 1 548 01 01 239

Qualcomm Inc 386 155 152 NR 9 341 28 29 583

Tellabs Inc 645 150 150 NR 1 796 01 01 266

Average 145 143 7 416 22 23 258

Computer and Network Equipment Industty Amdahl Corp $1500 13 120 NR 12 095 125 143 1~1

Bay Networks Inc 1342 175 174 NR 10 903 12 12 260

Cabletron Systems 1060 160 160 NR 0 657 00 00 218

Cisco Systems 1979 175 175 NR 0 1383 00 00 269

Digital Equipment 13813 110 104 NR 22 287 89 98 171

General Datacomm 221 185 179 NR 17 396 49 52 NMF

Hewlett-Packard 31519 125 124 NR 6 333 19 19 146

SCI Systems 2673 120 106 NR 32 220 176 214 127

Sequent Computer 535 195 192 NR 3 124 24 25 120

Standard Microsystems 345 160 152 NR 10 131 78 85 298

Stratus Computer 580 160 159 NR 2 136 15 15 133

Sun Microsystems 5902 155 154 NR 3 418 07 07 176

Tandem Computers 2285 155 150 NR 6 105 57 61 333

3ComCorp 1295 160 159 NR 12 1190 11 11 258

Average 155 151 10 456 47 53 204

-15- UVA-F-1152

Exhibit 4

TELETECH CORPORATION 1996

Debt Capital Market Conditions January 1996

Corporate Bond Yields by Rating US Treasurv Securities

Industrials AAA 650 Short-term bills 520 AA 700 Intermediate-term notes 543 A 764 Long-term bonds 604

BBB 778 BB 893 B 1049

Utilities AA 653 A 794 BBB 806

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

I

j

j

~~

c j

j

j

j

j

j

j

j

j

j

j

l j

j

j

j

UVA-F-1152

billion and $114 billion respectively Since the court-ordered breakup of the Bell System telephone monopoly in 1983 Teletech had coped with gradual deregulation of its industry through aggressive expansion into new services and geographic regions Most recently the firm had been a leading bidder for cellular telephone operations and for licenses to offer personal communications services (PCS) In addition the firm had purchased a number of telephone operating companies in privatization auctions in Latin America Finally the firm had invested aggressively in new technology-primarily digital switches and optical-fiber cables-in an effort to enhance its service quality All of these strategic moves had been costly the capital budget in this segment had varied between $15 and $2 billion in each of the previous 1 degyears

Unfortunately profit margins in the telecommunications segment had been under pressure for several years Government regulators had been slow to provide rate relief to Teletech for its capital investments Other leading telecommunications providers had expanded into Teletechs geographic markets and invested in new technology and quality enhancing assets Teletechs management noted that large cable TV companies might enter the teleconununications market and continue the pressure on margins

On the other hand Teletech was the dominant service provider in its geographic markets and product segments Customer surveys revealed that the company was the leader in product quality and customer satisfaction Teletechs management was confident that the company could conunand premium prices however the industry might evolve

Teletechs Products and Systems segment

Before 1990 teleconununications had been the companys core business supplemented by an equipment-manufacturing division that produced teleconununications components In 1990 the company acquired a leading computer workstation manufacturer with the goal ofapplying state-ofshythe-art computing technology to the design of telecommunications equipment The explosive growth in the microcomputer market and the increased use oftelephone lines to connect home- and office-based computers with mainframes oonvinced Teletech management ofthe potential value of marrying telecommunications equipment and computing technology Using Teletechs capital base borrowing ability and distribution network to catapult growth the Products and Systems segment increased its sales by nearly 40 percent in 1995 This segments 1995 NaPAT and net assets were $480 million and $46 billion respectively

Products and Systems was acknowledged to be a technology leader in the industry While this accounted for its rapid growth and pricing power maintenance of that leadership position required sizable investments in RampD and fixed assets The rate of technological change was increasing as witnessed by sudden major write-offs by Teletech on products that until recently management had thought were still competitive Major computer manufacturers were entering into the teleconununications-equipment industry Foreign manufacturers were proving to be stiff competitors in bidding on major supply contracts

-4- UVA-F-1l52 t

Focus on Value at Teletech

Teletechs mission statement said in part

We will create value by pursuing business activities that earn premium rates oreturn

Translating that statement into practice had been a challenge for Margaret Weston First it had been necessary to help managers of the segments and business units understand what create value meant for them Because the segments and smaller business units did not issue securities into the capital market the only objective measures of value were the securities prices of the whole corporation-but the activities ofany particular manager might not be significant enough to drive Teletechs securities prices Therefore the company had adopted a measure of value creation for use at the segment and business unit level that would provide a proxy for the way investors would view each units performance This measure called economic profit multiplied the excess rate of return of the business unit times the capital it used

Economic Profit = (ROC - Hurdle Rate) x Capital Employed

Where

NOPATROC = Return on Capital = --shy

Capital

NOPAT = Net Operating Profit After Taxes

Each year the segment and business unit executi ves were measured on the basis ofeconomic profit This measure was an important consideration in strategic decisions about capital allocation

manager promotion and the awarding of incentive compensation

A second way in which the value creation perspective influenced managers was in the assessment of capital-investment proposals For each investment projected cash flows were discounted to the present using the firms hurdle rate to give a measure of the net present value (or NPV) of each project A positive (negative) NPV indicated the amount by which the value of the firm would increase (decrease) ifthe project were undertaken The following equation shows how the hurdle rate was used in the familiar NPV equation

n Net Present Value E [ Free Cash FIOWI I] - Initial Investment

(1 + Hurdle Rate) t 1

Hurdle Rates

The hurdle rate used in the assessments of economic profit and NPV had been the focus of considerable debate in recent months This rate was based on an estimate of Teletechs weighted average cost of capital (WACC) Management was completely satisfied with the intellectual ~

c

-5- UVA-F-1152

relevance of a hurdle rate as an expression of the opportunity cost of money The notion that the W ACC represented this opportunity cost had been debated Its measurement was never considered wholly scientific but it had been accepted For instance Teletech was split-rated between AAshyand A+ An investment banker recently suggested that at these ratings new debt funds might cost Teletech 703 percent (about 422 percent after a 40 percent tax rate) With a beta of 1041 the cost of equity might be about 1177 percent At market-value weights of 18 percent for debt and 82 percent for equity the resulting W ACC would be 10Al percent Exhibit 1 summarizes the calculation The hurdle rate of lOA1 percent was applied to all investment and performanceshymeasurement analyses in the firm

Arguments for risk-adjusted hurdle rates

How the rate should be used within the company in evaluating projects was another point of debate Given the different natures ofthe two businesses and the risks each one faced differences of opinion arose at the segment level over the appropriateness of measuring all projects against the corporate hurdle rate of 10Al percent The chief advocate of multiple rates was Rick Phillips executive vice president ofTelecommunications Services who presented his views as follows

Each phase ofour business is different must compete differently and must draw on capital differently Until recently telecommunications was a regulated industry and the return on our total capital highly certain given the stable nature ofthe industry Because of the recognized safety of the investment many telecommunications companies can raise large quantities ofcapital from the debt markets In operations comparable to Telecommunications Services 75 percent of the necessary capital is raised in the debt markets at interest rates reflecting solid AA quality on averageshythis is better than the corporate bond rating ofAA- A+ Moreover I have to believe that the cost of equity of Telecommunications Services is lower than for Products and Systems I contrast this with the Products and Systems segment where although sales growth and profitability are strong risks are high Independent equipment manufacturers are financed by hi~her yield BBB-rated debt and more equity with higher expected total returns

In my book the hurdle rate for Products and Systems should reflect these higher costs offunds Without the risk-adjusted system of~dle rates Telecommunications Services will gradually starve for capital while Products and Systems will be forceshyfed-thats because our returns are less than the corporate hurdle rate and theirs are greater Telecommunications Services lowers the risk ofthe whole corporation and should not be penalized

Heres a rough graph of what I think is going on Telecommunications Services which can earn 98 percent on capital is actually profitable on a risk-adjusted basis even though it is not profitable compared to the corporate hurdle rate The triangle shape on the drawing shows about where Telecommunications Services is located

-6- UVA-F-1152 _ (

My hunch is that the reverse is true for Products and Systems which promises to earn 120 percent on capital P+S is located on the graph near the little circle

Constant vs Risk-Adjusted Hurdle Rates

17

- Teletech Corp Hurdle

-- Telecomm Services

-e- A-oducts and Systems

Risk Level

In deciding how much to loan us lenders will consider the composition of risks If money flows into safer investments over time the cost of their loans to us will decrease

Our stockholders are just as much concerned with risk Ifthey perceive our business as being more risky than other pound9mpanies they will not pay as high a price for our earnings Perhaps this is why our priceearnings ratio is below the industry average most ofthe time It is not a question ofwhether we adjust for risk-we already do infonnally The only question in my mind is whether we make these adjustments systematically or not

While multiple hurdle rates may not reflect capital-structure changes on a day -to-day basis over time they will reflect prospects more realistically At the moment as I understand it our real problem is an inadequate and very costly supply of equity funds If we are really rationing equity capital then we should be striving for the best returns on equity for the risk Multiple hurdle rates achieve this objective

Implicit in Phillipss argument as Weston understood it was the notion that ifeach segment in the company had a different hurdle rate the costs ofthe various fonns ofcapital would remain the same However the mix of capital used would change in the calculation Low-risk operations ~

c

-7- UVA-F-1152

would use leverage more extensively while the high-risk divisions would have little or no debt funds This lower-risk segment would have a lower hurdle rate

Opposition to risk-adjusted hurdle rates

Phillipss views were supported by several others within Teletech opposition was just as strong however particularly within the Products and Systems segment Helen Buono executive vice president for the segment expressed her opinion as follows

All money is green Investors cant know as much about our operations as we do To them the firm is an opaque box they hire us to take care ofwhat is inside the box andjudge us by the dividends coming out of the box We cant say that one part of the box has a different hurdle rate than another part ofthe box ifour investors dont think that way Like I say all money is green all investments at Teletech should be judged against one hurdle rate

Multiple hurdle rates are illogical Suppose that the hurdle rate for Teleshycommunications Services was much lower than the corporatewide hurdle rate Ifwe undertook investments that met the segment hurdle rate we would be destroying shareholder value because we werent meeting the (orporate hurdle rate shy

Ourjob as managers should be to put our money where the returns are best A single hurdle rate may deprive an underprofitable division of investments in order to channel more funds into a more profitable division but isnt that the aim of the process Our challenge today is simple we must earn the highest absolute rates of return we can get

In reality we dont finance each division separately The corporation raises capital based on its overall prospects and record The diversification of the company probably helps keep our capital C0sts down and enables us to borrow more in total than the sum of the capabilities of the divisions separately As a result developing separate hurdle rates is both unrealistic and misleading All our stockholders want is for us to invest our funds wisely in order to increase the value of their stock This happens when we pick the most promising projects~ irrespective of their source

Margaret Westons Concerns

As she listened to these arguments presented over the course of several months Weston became increasingly concerned with several related considerations First the corporate strategy directed the company toward integrating the two divisions One effect ofusing multiple hurdle rates would be to make justifying high-technology research and application proposals more difficult as the required rate ofreturn would be increased Perhaps she thought multiple hurdle rates were the

-8- UVA-F-1152 r f

right idea but the notion that they should be based on capital costs rather than strategic considerations might be wrong On the other hand perhaps multiple rates based on capital costs should be used but in allocating funds some qualitative adjustment should be made for unquantifiable strategic considerations In Westons mind theory was certainly not clear on how to achieve strategic objectives when allocating capital

Second using a single measure ofthe cost ofmoney (the hurdle rate or discount factor) made the net present value results consistent at least in economic terms If Teletech adopted multiple rates for discounting cash flows Weston was afraid the NPV and economic profit calculations would lose their meaning and comparability across business segments To her a performance criterion had to be consistent and understandable or it would not be useful

In addition Weston was concerned with the problem of attributing capital structures to divisions In Telecommunications Services a major new switching station might be financed by mortgage bonds But in Products and Systems it was not possible for the division to borrow directly indeed any financing was feasible only because the corporation guaranteed the debt Such projects were considered highly risky-perhaps at best warranting only a minimal debt structure Also Weston considered the debt-capacity decision difficult enough to make for the corporation as a whole let alone for each division Judgments could only be very crude

In further discussions with those in the organization about the use of multiple hurdle rates Weston ran across two predominant trains of thought One argument held that the investment C decision should never be mixed with the financing decisionr A firm should decide what its investments should be and then how to finance them most efficiently Adding leverage to a presentshyvalue calculation would distort the results Use ofmultiple hurdle rates was simply a way ofmixing financing with investment analysis This argument also held that a single rate left the risk decision clear-cut management could simply adjust its standard (NPV or economic profit) as risks increased

The contrasting line of reasoning noted that the weighted-average cost of capital tended to represent an average market reaction tol mixture ofrisks Lower-than-average-risk projects should probably be accepted even though they did not meet a weighted-average criterion Higher-than-normal-risk projects should provide a return premium While the multiple-hurdle-rate system was a crude way ofachieving this end it at least was a step in the right direction Moreover some argued that Teletechs objective should be to maximize return on equity funds and because

equity funds were and would remain a comparatively scarce resource a multiple-rate system would tend to maximize returns to stockholders better than a single-rate system

To help resolve these questions Weston asked her assistant Bernard Ingles to summarize academic thinking about multiple hurdle rates His memorandum is given in Exhibit 2 She also requested that he draw samples ofcomparable firms for Telecommunications Services and Products and Systems that might be used in deriving segment WACCs The summary of data is given in Exhibit 3 Information on capital-market conditions in January 1996 is given in Exhibit 4

-9- UVA-F-1l52

Conclusion

Weston could not realistically hope that all the issues before her would be resolved in time to influence Victor Yossarians attack on management But the attack did dictate the need for an objective assessment of the performance of Teletechs two segments-the choice of hurdle rates would be very important in this analysis However she did want to institute a pragmatic system of appropriate hurdle rates (or one rate) that would facilitate judgments in the changing circumstances Teletech faced What were the appropriate hurdle rates for the two segments Was Products and Systems underperforming as Yossarian suggested How should Teletech respond to the raider

-10- UVA-F-1152 shy

Exhibit 1

TELETECH CORPORATION 1996

Summary of W ACC Calculation for Teletech Corporation and

Segment Worksheet

Telecommunications Products and

Co]rate Services SYstems

MY Asset Weights 100 75 25

Bond Rating AA-A+ AA BBB-

Pre-tax Cost of Debt 703 700 778

Tax Rate 40 40 40

After-tax Cost of Debt 422 420 467

Equity Beta 104

Rf 604

RM-Rf 550

Cost of Equity 1177

Weight of Debt 180

Weight of Equity 820

WACC 1041 I

c

UVA-F-1152

Exhibit 2

TELETECH CORPORATION 1996

Theoretical Overview of Multiple Hurdle Rates

To Margaret Weston From Bernard Ingles Subject Theory of Segment Cost of Capital Date January 1996

You requested an overview of theories about multiple hurdle rates Without getting into minutiae the theories boil down to the following points

1 The central idea is that required returns should be driven by risk This is the dominant view in the field of investment management and is based on a mountain of theory and empirical research stretching over several decades The extension of this idea from investment management to corporate decision making is straightforward p~ least in theory

2 An underlying assumption is that the firm is transparent (Le that investors can see through the corporate veil and evaluate the activities going on inside) No one believes firms are completely transparent or that investors are perfectly informed But financial accounting standards have evolved toward making the firm more transparent And the investment community has grown tougher and sharper in its analysis Teletech now has 36 analysts publishing reports and forecasts on the firm The reality is that for big publicly held firms transparency is not a bad assumption

3 Another underlying assumption is that the value of the whole enterprise is simply the sum of its parts-this is the concept of Value Additivity We can define parts as either the business segments (on the left-hand side of the balance sheet) or the layers of the capital structure (on the right-hand side of the balance sheet) Market values (MVs) have to balance

MVTeteedl = ( MVTetlaquoomllllltlicaliolls Services + MV tmdmls+ iwellls) = ( MVdehl + MVltquiV)

Ifthese equalities did not hold then a raider could come along and exploit the inequality by buying or selling the whole and the parts This is arbitrage By buying and selling the actions of the raider would drive the MVs back into balance

-12- UVA-F-1152 r ~

Exhibit 2 (continued)

4 Investment theory tells us that the only risk that matters is nondiversifiable risk which is measured by beta Beta indicates the risk that an asset will add to a portfolio Because the investor is assumed to be diversified she is assumed to seek a return for only that risk that she cannot shed the nondiversifiabIe risk Now the important point here is that the beta ofa portfolio is equal to a weighted average of the betas of the portfolio components Extending this to the corporate environment the asset beta for the firm will equal a weighted average of the components of the firm-again the components of the firm can be defined in terms of either the right-hand side or the left-hand side of the balance sheet

Where

W = percentage weights based on market values

fJ Tel Serv fJ ps Asset betas for business segments

fJdebl fJ for the firms debt securities

fJeqUlIY= fJ of firms common stock (given by Bloomberg etc)

This is a very handy way to model the risk ofthe firm for it means that we can use the Capital Asset Pricing Model to estimate the cost of capital for a segment (ie using segment asset betas)

5 Given all the previous points it follows that the weighted average of the various costs ofcapital (K) for the firm (W ACC) which is the theoretically correct hurdle rate is simply a weighted average of segment W ACCs

Where

WTeCf WIS = market value weights

WACCTeLrv = (Wdebl TeLerv( K debl TelServ) + ( Weqllily TelServt K eltIUlly TelServ)

c

-13- UVA-F-1152

Exhibit 2 (continued)

6 The notion in point 5 may not hold exactly in practice First most of the components in the WACC formula are estimated with some error Second because oftaxes information asymmetries or other market imperfections assets may not be priced strictly in line with the model-for a company like Teletech it is reasonable to assume that any mispricings are just temporary Third the simple two-segment characterization ignores a hidden third segment the corporate treasury department that hedges and aims to finance the whole corporation optimally-this acts as a shock absorber for the financial policies of the segments Modeling the W ACC ofthe corporate treasury department is quite difficult Most companies assume that the impact of corporate treasury isnt very large and simply assume it away As a first cut we could do this too though it is an issue we should revisit

Conclusions

bull In theory the corporate WACC for Teletech is appropriate only for evaluating an asset having the same risk as the whole company It is not appropriate for assets having different risks than the whole company

bull Segment WACCs are computed similarly to corporate W ACCs

bull In concept the corporate W ACe is aweighted average of the segment W ACCs In practice the weighted-average concept may not hold due to imperfections in the market andor estimation errors

bull If we start computing segment W ACCs we must use the cost of debt cost of equity and weights appropriate to that segment We need a lot of information to do this correctly or else we really need to stretch to make assumptions

-14- UVA-F-1152_

Exhibit 3 TELETECH CORPORATION 1996

Samples of Comparable Firms

1995 Equity Asset Bond Book Val Pricel Mkt Val Mkt Val Pricel Revenues Seta Seta Iiaiiog OebtJCap Book DebtCap DebtlEq Eacniogs

Teletech Corporation $16000 1041 092 AA-tA+ 40 301 18 22 129

Telecommunications Services Industty ATampT $80000 090 085 AA 39 660 88 97 308 Alltel Corp 3160 075 063 A 49 299 243 322 160 Amerltech 13325 075 067 AA 47 472 158 188 169 BeD Atlantic 13500 080 068 AA 57 453 226 293 175 Bell South 17780 075 072 AAA 44 1190 62 66 190

Centul) Tel Enterprs 625 100 084 BBB+ 46 263 245 324 158

Cincinnati Bell 1350 080 069 AA 56 472 212 270 188 Citizens Utilities Co 1070 070 056 AA 42 168 301 431 158

Comsat 850 095 068 A 40 103 392 646 168

Frontier Corp 1750 080 074 A 42 550 116 132 283

GTE Corp 20250 080 066 BBB 69 652 254 341 169

MCI Communications 15100 125 114 A 24 187 144 169 179 NYNEX Corp 13425 075 059 Ashy 62 375 303 436 156

Pacific Telesis 9070 085 068 AAshy 74 694 291 410 136

SBC Communications 12560 090 080 A 54 559 173 210 180

Southem New England 1840 075 058 AA 5710 263 335 504 154

Sprint Corp

US West

Average

13550

9450

105

065

084

087

052

072

BBB

AAshy

52

66

51

313

467

452

257

294

228

347

416

311

150

143 C179

Telecommunications Equipment Industty ADC Telecomm Inc $586 135 135 NR 0 405 00 00 280

Aane-Cleveland 120 150 149 NR 1 151 07 07 128

ADen Group 325 160 155 NR 13 275 52 54 188

Andrew Corp 626 125 123 NR 13 440 33 34 197

DSC Communications 1450 130 126 NR 18 372 56 590 179

Newbridge NetworIs 675 155 155 NR 1 548 01 01 239

Qualcomm Inc 386 155 152 NR 9 341 28 29 583

Tellabs Inc 645 150 150 NR 1 796 01 01 266

Average 145 143 7 416 22 23 258

Computer and Network Equipment Industty Amdahl Corp $1500 13 120 NR 12 095 125 143 1~1

Bay Networks Inc 1342 175 174 NR 10 903 12 12 260

Cabletron Systems 1060 160 160 NR 0 657 00 00 218

Cisco Systems 1979 175 175 NR 0 1383 00 00 269

Digital Equipment 13813 110 104 NR 22 287 89 98 171

General Datacomm 221 185 179 NR 17 396 49 52 NMF

Hewlett-Packard 31519 125 124 NR 6 333 19 19 146

SCI Systems 2673 120 106 NR 32 220 176 214 127

Sequent Computer 535 195 192 NR 3 124 24 25 120

Standard Microsystems 345 160 152 NR 10 131 78 85 298

Stratus Computer 580 160 159 NR 2 136 15 15 133

Sun Microsystems 5902 155 154 NR 3 418 07 07 176

Tandem Computers 2285 155 150 NR 6 105 57 61 333

3ComCorp 1295 160 159 NR 12 1190 11 11 258

Average 155 151 10 456 47 53 204

-15- UVA-F-1152

Exhibit 4

TELETECH CORPORATION 1996

Debt Capital Market Conditions January 1996

Corporate Bond Yields by Rating US Treasurv Securities

Industrials AAA 650 Short-term bills 520 AA 700 Intermediate-term notes 543 A 764 Long-term bonds 604

BBB 778 BB 893 B 1049

Utilities AA 653 A 794 BBB 806

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

I

j

j

~~

c j

j

j

j

j

j

j

j

j

j

j

l j

j

j

j

-4- UVA-F-1l52 t

Focus on Value at Teletech

Teletechs mission statement said in part

We will create value by pursuing business activities that earn premium rates oreturn

Translating that statement into practice had been a challenge for Margaret Weston First it had been necessary to help managers of the segments and business units understand what create value meant for them Because the segments and smaller business units did not issue securities into the capital market the only objective measures of value were the securities prices of the whole corporation-but the activities ofany particular manager might not be significant enough to drive Teletechs securities prices Therefore the company had adopted a measure of value creation for use at the segment and business unit level that would provide a proxy for the way investors would view each units performance This measure called economic profit multiplied the excess rate of return of the business unit times the capital it used

Economic Profit = (ROC - Hurdle Rate) x Capital Employed

Where

NOPATROC = Return on Capital = --shy

Capital

NOPAT = Net Operating Profit After Taxes

Each year the segment and business unit executi ves were measured on the basis ofeconomic profit This measure was an important consideration in strategic decisions about capital allocation

manager promotion and the awarding of incentive compensation

A second way in which the value creation perspective influenced managers was in the assessment of capital-investment proposals For each investment projected cash flows were discounted to the present using the firms hurdle rate to give a measure of the net present value (or NPV) of each project A positive (negative) NPV indicated the amount by which the value of the firm would increase (decrease) ifthe project were undertaken The following equation shows how the hurdle rate was used in the familiar NPV equation

n Net Present Value E [ Free Cash FIOWI I] - Initial Investment

(1 + Hurdle Rate) t 1

Hurdle Rates

The hurdle rate used in the assessments of economic profit and NPV had been the focus of considerable debate in recent months This rate was based on an estimate of Teletechs weighted average cost of capital (WACC) Management was completely satisfied with the intellectual ~

c

-5- UVA-F-1152

relevance of a hurdle rate as an expression of the opportunity cost of money The notion that the W ACC represented this opportunity cost had been debated Its measurement was never considered wholly scientific but it had been accepted For instance Teletech was split-rated between AAshyand A+ An investment banker recently suggested that at these ratings new debt funds might cost Teletech 703 percent (about 422 percent after a 40 percent tax rate) With a beta of 1041 the cost of equity might be about 1177 percent At market-value weights of 18 percent for debt and 82 percent for equity the resulting W ACC would be 10Al percent Exhibit 1 summarizes the calculation The hurdle rate of lOA1 percent was applied to all investment and performanceshymeasurement analyses in the firm

Arguments for risk-adjusted hurdle rates

How the rate should be used within the company in evaluating projects was another point of debate Given the different natures ofthe two businesses and the risks each one faced differences of opinion arose at the segment level over the appropriateness of measuring all projects against the corporate hurdle rate of 10Al percent The chief advocate of multiple rates was Rick Phillips executive vice president ofTelecommunications Services who presented his views as follows

Each phase ofour business is different must compete differently and must draw on capital differently Until recently telecommunications was a regulated industry and the return on our total capital highly certain given the stable nature ofthe industry Because of the recognized safety of the investment many telecommunications companies can raise large quantities ofcapital from the debt markets In operations comparable to Telecommunications Services 75 percent of the necessary capital is raised in the debt markets at interest rates reflecting solid AA quality on averageshythis is better than the corporate bond rating ofAA- A+ Moreover I have to believe that the cost of equity of Telecommunications Services is lower than for Products and Systems I contrast this with the Products and Systems segment where although sales growth and profitability are strong risks are high Independent equipment manufacturers are financed by hi~her yield BBB-rated debt and more equity with higher expected total returns

In my book the hurdle rate for Products and Systems should reflect these higher costs offunds Without the risk-adjusted system of~dle rates Telecommunications Services will gradually starve for capital while Products and Systems will be forceshyfed-thats because our returns are less than the corporate hurdle rate and theirs are greater Telecommunications Services lowers the risk ofthe whole corporation and should not be penalized

Heres a rough graph of what I think is going on Telecommunications Services which can earn 98 percent on capital is actually profitable on a risk-adjusted basis even though it is not profitable compared to the corporate hurdle rate The triangle shape on the drawing shows about where Telecommunications Services is located

-6- UVA-F-1152 _ (

My hunch is that the reverse is true for Products and Systems which promises to earn 120 percent on capital P+S is located on the graph near the little circle

Constant vs Risk-Adjusted Hurdle Rates

17

- Teletech Corp Hurdle

-- Telecomm Services

-e- A-oducts and Systems

Risk Level

In deciding how much to loan us lenders will consider the composition of risks If money flows into safer investments over time the cost of their loans to us will decrease

Our stockholders are just as much concerned with risk Ifthey perceive our business as being more risky than other pound9mpanies they will not pay as high a price for our earnings Perhaps this is why our priceearnings ratio is below the industry average most ofthe time It is not a question ofwhether we adjust for risk-we already do infonnally The only question in my mind is whether we make these adjustments systematically or not

While multiple hurdle rates may not reflect capital-structure changes on a day -to-day basis over time they will reflect prospects more realistically At the moment as I understand it our real problem is an inadequate and very costly supply of equity funds If we are really rationing equity capital then we should be striving for the best returns on equity for the risk Multiple hurdle rates achieve this objective

Implicit in Phillipss argument as Weston understood it was the notion that ifeach segment in the company had a different hurdle rate the costs ofthe various fonns ofcapital would remain the same However the mix of capital used would change in the calculation Low-risk operations ~

c

-7- UVA-F-1152

would use leverage more extensively while the high-risk divisions would have little or no debt funds This lower-risk segment would have a lower hurdle rate

Opposition to risk-adjusted hurdle rates

Phillipss views were supported by several others within Teletech opposition was just as strong however particularly within the Products and Systems segment Helen Buono executive vice president for the segment expressed her opinion as follows

All money is green Investors cant know as much about our operations as we do To them the firm is an opaque box they hire us to take care ofwhat is inside the box andjudge us by the dividends coming out of the box We cant say that one part of the box has a different hurdle rate than another part ofthe box ifour investors dont think that way Like I say all money is green all investments at Teletech should be judged against one hurdle rate

Multiple hurdle rates are illogical Suppose that the hurdle rate for Teleshycommunications Services was much lower than the corporatewide hurdle rate Ifwe undertook investments that met the segment hurdle rate we would be destroying shareholder value because we werent meeting the (orporate hurdle rate shy

Ourjob as managers should be to put our money where the returns are best A single hurdle rate may deprive an underprofitable division of investments in order to channel more funds into a more profitable division but isnt that the aim of the process Our challenge today is simple we must earn the highest absolute rates of return we can get

In reality we dont finance each division separately The corporation raises capital based on its overall prospects and record The diversification of the company probably helps keep our capital C0sts down and enables us to borrow more in total than the sum of the capabilities of the divisions separately As a result developing separate hurdle rates is both unrealistic and misleading All our stockholders want is for us to invest our funds wisely in order to increase the value of their stock This happens when we pick the most promising projects~ irrespective of their source

Margaret Westons Concerns

As she listened to these arguments presented over the course of several months Weston became increasingly concerned with several related considerations First the corporate strategy directed the company toward integrating the two divisions One effect ofusing multiple hurdle rates would be to make justifying high-technology research and application proposals more difficult as the required rate ofreturn would be increased Perhaps she thought multiple hurdle rates were the

-8- UVA-F-1152 r f

right idea but the notion that they should be based on capital costs rather than strategic considerations might be wrong On the other hand perhaps multiple rates based on capital costs should be used but in allocating funds some qualitative adjustment should be made for unquantifiable strategic considerations In Westons mind theory was certainly not clear on how to achieve strategic objectives when allocating capital

Second using a single measure ofthe cost ofmoney (the hurdle rate or discount factor) made the net present value results consistent at least in economic terms If Teletech adopted multiple rates for discounting cash flows Weston was afraid the NPV and economic profit calculations would lose their meaning and comparability across business segments To her a performance criterion had to be consistent and understandable or it would not be useful

In addition Weston was concerned with the problem of attributing capital structures to divisions In Telecommunications Services a major new switching station might be financed by mortgage bonds But in Products and Systems it was not possible for the division to borrow directly indeed any financing was feasible only because the corporation guaranteed the debt Such projects were considered highly risky-perhaps at best warranting only a minimal debt structure Also Weston considered the debt-capacity decision difficult enough to make for the corporation as a whole let alone for each division Judgments could only be very crude

In further discussions with those in the organization about the use of multiple hurdle rates Weston ran across two predominant trains of thought One argument held that the investment C decision should never be mixed with the financing decisionr A firm should decide what its investments should be and then how to finance them most efficiently Adding leverage to a presentshyvalue calculation would distort the results Use ofmultiple hurdle rates was simply a way ofmixing financing with investment analysis This argument also held that a single rate left the risk decision clear-cut management could simply adjust its standard (NPV or economic profit) as risks increased

The contrasting line of reasoning noted that the weighted-average cost of capital tended to represent an average market reaction tol mixture ofrisks Lower-than-average-risk projects should probably be accepted even though they did not meet a weighted-average criterion Higher-than-normal-risk projects should provide a return premium While the multiple-hurdle-rate system was a crude way ofachieving this end it at least was a step in the right direction Moreover some argued that Teletechs objective should be to maximize return on equity funds and because

equity funds were and would remain a comparatively scarce resource a multiple-rate system would tend to maximize returns to stockholders better than a single-rate system

To help resolve these questions Weston asked her assistant Bernard Ingles to summarize academic thinking about multiple hurdle rates His memorandum is given in Exhibit 2 She also requested that he draw samples ofcomparable firms for Telecommunications Services and Products and Systems that might be used in deriving segment WACCs The summary of data is given in Exhibit 3 Information on capital-market conditions in January 1996 is given in Exhibit 4

-9- UVA-F-1l52

Conclusion

Weston could not realistically hope that all the issues before her would be resolved in time to influence Victor Yossarians attack on management But the attack did dictate the need for an objective assessment of the performance of Teletechs two segments-the choice of hurdle rates would be very important in this analysis However she did want to institute a pragmatic system of appropriate hurdle rates (or one rate) that would facilitate judgments in the changing circumstances Teletech faced What were the appropriate hurdle rates for the two segments Was Products and Systems underperforming as Yossarian suggested How should Teletech respond to the raider

-10- UVA-F-1152 shy

Exhibit 1

TELETECH CORPORATION 1996

Summary of W ACC Calculation for Teletech Corporation and

Segment Worksheet

Telecommunications Products and

Co]rate Services SYstems

MY Asset Weights 100 75 25

Bond Rating AA-A+ AA BBB-

Pre-tax Cost of Debt 703 700 778

Tax Rate 40 40 40

After-tax Cost of Debt 422 420 467

Equity Beta 104

Rf 604

RM-Rf 550

Cost of Equity 1177

Weight of Debt 180

Weight of Equity 820

WACC 1041 I

c

UVA-F-1152

Exhibit 2

TELETECH CORPORATION 1996

Theoretical Overview of Multiple Hurdle Rates

To Margaret Weston From Bernard Ingles Subject Theory of Segment Cost of Capital Date January 1996

You requested an overview of theories about multiple hurdle rates Without getting into minutiae the theories boil down to the following points

1 The central idea is that required returns should be driven by risk This is the dominant view in the field of investment management and is based on a mountain of theory and empirical research stretching over several decades The extension of this idea from investment management to corporate decision making is straightforward p~ least in theory

2 An underlying assumption is that the firm is transparent (Le that investors can see through the corporate veil and evaluate the activities going on inside) No one believes firms are completely transparent or that investors are perfectly informed But financial accounting standards have evolved toward making the firm more transparent And the investment community has grown tougher and sharper in its analysis Teletech now has 36 analysts publishing reports and forecasts on the firm The reality is that for big publicly held firms transparency is not a bad assumption

3 Another underlying assumption is that the value of the whole enterprise is simply the sum of its parts-this is the concept of Value Additivity We can define parts as either the business segments (on the left-hand side of the balance sheet) or the layers of the capital structure (on the right-hand side of the balance sheet) Market values (MVs) have to balance

MVTeteedl = ( MVTetlaquoomllllltlicaliolls Services + MV tmdmls+ iwellls) = ( MVdehl + MVltquiV)

Ifthese equalities did not hold then a raider could come along and exploit the inequality by buying or selling the whole and the parts This is arbitrage By buying and selling the actions of the raider would drive the MVs back into balance

-12- UVA-F-1152 r ~

Exhibit 2 (continued)

4 Investment theory tells us that the only risk that matters is nondiversifiable risk which is measured by beta Beta indicates the risk that an asset will add to a portfolio Because the investor is assumed to be diversified she is assumed to seek a return for only that risk that she cannot shed the nondiversifiabIe risk Now the important point here is that the beta ofa portfolio is equal to a weighted average of the betas of the portfolio components Extending this to the corporate environment the asset beta for the firm will equal a weighted average of the components of the firm-again the components of the firm can be defined in terms of either the right-hand side or the left-hand side of the balance sheet

Where

W = percentage weights based on market values

fJ Tel Serv fJ ps Asset betas for business segments

fJdebl fJ for the firms debt securities

fJeqUlIY= fJ of firms common stock (given by Bloomberg etc)

This is a very handy way to model the risk ofthe firm for it means that we can use the Capital Asset Pricing Model to estimate the cost of capital for a segment (ie using segment asset betas)

5 Given all the previous points it follows that the weighted average of the various costs ofcapital (K) for the firm (W ACC) which is the theoretically correct hurdle rate is simply a weighted average of segment W ACCs

Where

WTeCf WIS = market value weights

WACCTeLrv = (Wdebl TeLerv( K debl TelServ) + ( Weqllily TelServt K eltIUlly TelServ)

c

-13- UVA-F-1152

Exhibit 2 (continued)

6 The notion in point 5 may not hold exactly in practice First most of the components in the WACC formula are estimated with some error Second because oftaxes information asymmetries or other market imperfections assets may not be priced strictly in line with the model-for a company like Teletech it is reasonable to assume that any mispricings are just temporary Third the simple two-segment characterization ignores a hidden third segment the corporate treasury department that hedges and aims to finance the whole corporation optimally-this acts as a shock absorber for the financial policies of the segments Modeling the W ACC ofthe corporate treasury department is quite difficult Most companies assume that the impact of corporate treasury isnt very large and simply assume it away As a first cut we could do this too though it is an issue we should revisit

Conclusions

bull In theory the corporate WACC for Teletech is appropriate only for evaluating an asset having the same risk as the whole company It is not appropriate for assets having different risks than the whole company

bull Segment WACCs are computed similarly to corporate W ACCs

bull In concept the corporate W ACe is aweighted average of the segment W ACCs In practice the weighted-average concept may not hold due to imperfections in the market andor estimation errors

bull If we start computing segment W ACCs we must use the cost of debt cost of equity and weights appropriate to that segment We need a lot of information to do this correctly or else we really need to stretch to make assumptions

-14- UVA-F-1152_

Exhibit 3 TELETECH CORPORATION 1996

Samples of Comparable Firms

1995 Equity Asset Bond Book Val Pricel Mkt Val Mkt Val Pricel Revenues Seta Seta Iiaiiog OebtJCap Book DebtCap DebtlEq Eacniogs

Teletech Corporation $16000 1041 092 AA-tA+ 40 301 18 22 129

Telecommunications Services Industty ATampT $80000 090 085 AA 39 660 88 97 308 Alltel Corp 3160 075 063 A 49 299 243 322 160 Amerltech 13325 075 067 AA 47 472 158 188 169 BeD Atlantic 13500 080 068 AA 57 453 226 293 175 Bell South 17780 075 072 AAA 44 1190 62 66 190

Centul) Tel Enterprs 625 100 084 BBB+ 46 263 245 324 158

Cincinnati Bell 1350 080 069 AA 56 472 212 270 188 Citizens Utilities Co 1070 070 056 AA 42 168 301 431 158

Comsat 850 095 068 A 40 103 392 646 168

Frontier Corp 1750 080 074 A 42 550 116 132 283

GTE Corp 20250 080 066 BBB 69 652 254 341 169

MCI Communications 15100 125 114 A 24 187 144 169 179 NYNEX Corp 13425 075 059 Ashy 62 375 303 436 156

Pacific Telesis 9070 085 068 AAshy 74 694 291 410 136

SBC Communications 12560 090 080 A 54 559 173 210 180

Southem New England 1840 075 058 AA 5710 263 335 504 154

Sprint Corp

US West

Average

13550

9450

105

065

084

087

052

072

BBB

AAshy

52

66

51

313

467

452

257

294

228

347

416

311

150

143 C179

Telecommunications Equipment Industty ADC Telecomm Inc $586 135 135 NR 0 405 00 00 280

Aane-Cleveland 120 150 149 NR 1 151 07 07 128

ADen Group 325 160 155 NR 13 275 52 54 188

Andrew Corp 626 125 123 NR 13 440 33 34 197

DSC Communications 1450 130 126 NR 18 372 56 590 179

Newbridge NetworIs 675 155 155 NR 1 548 01 01 239

Qualcomm Inc 386 155 152 NR 9 341 28 29 583

Tellabs Inc 645 150 150 NR 1 796 01 01 266

Average 145 143 7 416 22 23 258

Computer and Network Equipment Industty Amdahl Corp $1500 13 120 NR 12 095 125 143 1~1

Bay Networks Inc 1342 175 174 NR 10 903 12 12 260

Cabletron Systems 1060 160 160 NR 0 657 00 00 218

Cisco Systems 1979 175 175 NR 0 1383 00 00 269

Digital Equipment 13813 110 104 NR 22 287 89 98 171

General Datacomm 221 185 179 NR 17 396 49 52 NMF

Hewlett-Packard 31519 125 124 NR 6 333 19 19 146

SCI Systems 2673 120 106 NR 32 220 176 214 127

Sequent Computer 535 195 192 NR 3 124 24 25 120

Standard Microsystems 345 160 152 NR 10 131 78 85 298

Stratus Computer 580 160 159 NR 2 136 15 15 133

Sun Microsystems 5902 155 154 NR 3 418 07 07 176

Tandem Computers 2285 155 150 NR 6 105 57 61 333

3ComCorp 1295 160 159 NR 12 1190 11 11 258

Average 155 151 10 456 47 53 204

-15- UVA-F-1152

Exhibit 4

TELETECH CORPORATION 1996

Debt Capital Market Conditions January 1996

Corporate Bond Yields by Rating US Treasurv Securities

Industrials AAA 650 Short-term bills 520 AA 700 Intermediate-term notes 543 A 764 Long-term bonds 604

BBB 778 BB 893 B 1049

Utilities AA 653 A 794 BBB 806

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

I

j

j

~~

c j

j

j

j

j

j

j

j

j

j

j

l j

j

j

j

-5- UVA-F-1152

relevance of a hurdle rate as an expression of the opportunity cost of money The notion that the W ACC represented this opportunity cost had been debated Its measurement was never considered wholly scientific but it had been accepted For instance Teletech was split-rated between AAshyand A+ An investment banker recently suggested that at these ratings new debt funds might cost Teletech 703 percent (about 422 percent after a 40 percent tax rate) With a beta of 1041 the cost of equity might be about 1177 percent At market-value weights of 18 percent for debt and 82 percent for equity the resulting W ACC would be 10Al percent Exhibit 1 summarizes the calculation The hurdle rate of lOA1 percent was applied to all investment and performanceshymeasurement analyses in the firm

Arguments for risk-adjusted hurdle rates

How the rate should be used within the company in evaluating projects was another point of debate Given the different natures ofthe two businesses and the risks each one faced differences of opinion arose at the segment level over the appropriateness of measuring all projects against the corporate hurdle rate of 10Al percent The chief advocate of multiple rates was Rick Phillips executive vice president ofTelecommunications Services who presented his views as follows

Each phase ofour business is different must compete differently and must draw on capital differently Until recently telecommunications was a regulated industry and the return on our total capital highly certain given the stable nature ofthe industry Because of the recognized safety of the investment many telecommunications companies can raise large quantities ofcapital from the debt markets In operations comparable to Telecommunications Services 75 percent of the necessary capital is raised in the debt markets at interest rates reflecting solid AA quality on averageshythis is better than the corporate bond rating ofAA- A+ Moreover I have to believe that the cost of equity of Telecommunications Services is lower than for Products and Systems I contrast this with the Products and Systems segment where although sales growth and profitability are strong risks are high Independent equipment manufacturers are financed by hi~her yield BBB-rated debt and more equity with higher expected total returns

In my book the hurdle rate for Products and Systems should reflect these higher costs offunds Without the risk-adjusted system of~dle rates Telecommunications Services will gradually starve for capital while Products and Systems will be forceshyfed-thats because our returns are less than the corporate hurdle rate and theirs are greater Telecommunications Services lowers the risk ofthe whole corporation and should not be penalized

Heres a rough graph of what I think is going on Telecommunications Services which can earn 98 percent on capital is actually profitable on a risk-adjusted basis even though it is not profitable compared to the corporate hurdle rate The triangle shape on the drawing shows about where Telecommunications Services is located

-6- UVA-F-1152 _ (

My hunch is that the reverse is true for Products and Systems which promises to earn 120 percent on capital P+S is located on the graph near the little circle

Constant vs Risk-Adjusted Hurdle Rates

17

- Teletech Corp Hurdle

-- Telecomm Services

-e- A-oducts and Systems

Risk Level

In deciding how much to loan us lenders will consider the composition of risks If money flows into safer investments over time the cost of their loans to us will decrease

Our stockholders are just as much concerned with risk Ifthey perceive our business as being more risky than other pound9mpanies they will not pay as high a price for our earnings Perhaps this is why our priceearnings ratio is below the industry average most ofthe time It is not a question ofwhether we adjust for risk-we already do infonnally The only question in my mind is whether we make these adjustments systematically or not

While multiple hurdle rates may not reflect capital-structure changes on a day -to-day basis over time they will reflect prospects more realistically At the moment as I understand it our real problem is an inadequate and very costly supply of equity funds If we are really rationing equity capital then we should be striving for the best returns on equity for the risk Multiple hurdle rates achieve this objective

Implicit in Phillipss argument as Weston understood it was the notion that ifeach segment in the company had a different hurdle rate the costs ofthe various fonns ofcapital would remain the same However the mix of capital used would change in the calculation Low-risk operations ~

c

-7- UVA-F-1152

would use leverage more extensively while the high-risk divisions would have little or no debt funds This lower-risk segment would have a lower hurdle rate

Opposition to risk-adjusted hurdle rates

Phillipss views were supported by several others within Teletech opposition was just as strong however particularly within the Products and Systems segment Helen Buono executive vice president for the segment expressed her opinion as follows

All money is green Investors cant know as much about our operations as we do To them the firm is an opaque box they hire us to take care ofwhat is inside the box andjudge us by the dividends coming out of the box We cant say that one part of the box has a different hurdle rate than another part ofthe box ifour investors dont think that way Like I say all money is green all investments at Teletech should be judged against one hurdle rate

Multiple hurdle rates are illogical Suppose that the hurdle rate for Teleshycommunications Services was much lower than the corporatewide hurdle rate Ifwe undertook investments that met the segment hurdle rate we would be destroying shareholder value because we werent meeting the (orporate hurdle rate shy

Ourjob as managers should be to put our money where the returns are best A single hurdle rate may deprive an underprofitable division of investments in order to channel more funds into a more profitable division but isnt that the aim of the process Our challenge today is simple we must earn the highest absolute rates of return we can get

In reality we dont finance each division separately The corporation raises capital based on its overall prospects and record The diversification of the company probably helps keep our capital C0sts down and enables us to borrow more in total than the sum of the capabilities of the divisions separately As a result developing separate hurdle rates is both unrealistic and misleading All our stockholders want is for us to invest our funds wisely in order to increase the value of their stock This happens when we pick the most promising projects~ irrespective of their source

Margaret Westons Concerns

As she listened to these arguments presented over the course of several months Weston became increasingly concerned with several related considerations First the corporate strategy directed the company toward integrating the two divisions One effect ofusing multiple hurdle rates would be to make justifying high-technology research and application proposals more difficult as the required rate ofreturn would be increased Perhaps she thought multiple hurdle rates were the

-8- UVA-F-1152 r f

right idea but the notion that they should be based on capital costs rather than strategic considerations might be wrong On the other hand perhaps multiple rates based on capital costs should be used but in allocating funds some qualitative adjustment should be made for unquantifiable strategic considerations In Westons mind theory was certainly not clear on how to achieve strategic objectives when allocating capital

Second using a single measure ofthe cost ofmoney (the hurdle rate or discount factor) made the net present value results consistent at least in economic terms If Teletech adopted multiple rates for discounting cash flows Weston was afraid the NPV and economic profit calculations would lose their meaning and comparability across business segments To her a performance criterion had to be consistent and understandable or it would not be useful

In addition Weston was concerned with the problem of attributing capital structures to divisions In Telecommunications Services a major new switching station might be financed by mortgage bonds But in Products and Systems it was not possible for the division to borrow directly indeed any financing was feasible only because the corporation guaranteed the debt Such projects were considered highly risky-perhaps at best warranting only a minimal debt structure Also Weston considered the debt-capacity decision difficult enough to make for the corporation as a whole let alone for each division Judgments could only be very crude

In further discussions with those in the organization about the use of multiple hurdle rates Weston ran across two predominant trains of thought One argument held that the investment C decision should never be mixed with the financing decisionr A firm should decide what its investments should be and then how to finance them most efficiently Adding leverage to a presentshyvalue calculation would distort the results Use ofmultiple hurdle rates was simply a way ofmixing financing with investment analysis This argument also held that a single rate left the risk decision clear-cut management could simply adjust its standard (NPV or economic profit) as risks increased

The contrasting line of reasoning noted that the weighted-average cost of capital tended to represent an average market reaction tol mixture ofrisks Lower-than-average-risk projects should probably be accepted even though they did not meet a weighted-average criterion Higher-than-normal-risk projects should provide a return premium While the multiple-hurdle-rate system was a crude way ofachieving this end it at least was a step in the right direction Moreover some argued that Teletechs objective should be to maximize return on equity funds and because

equity funds were and would remain a comparatively scarce resource a multiple-rate system would tend to maximize returns to stockholders better than a single-rate system

To help resolve these questions Weston asked her assistant Bernard Ingles to summarize academic thinking about multiple hurdle rates His memorandum is given in Exhibit 2 She also requested that he draw samples ofcomparable firms for Telecommunications Services and Products and Systems that might be used in deriving segment WACCs The summary of data is given in Exhibit 3 Information on capital-market conditions in January 1996 is given in Exhibit 4

-9- UVA-F-1l52

Conclusion

Weston could not realistically hope that all the issues before her would be resolved in time to influence Victor Yossarians attack on management But the attack did dictate the need for an objective assessment of the performance of Teletechs two segments-the choice of hurdle rates would be very important in this analysis However she did want to institute a pragmatic system of appropriate hurdle rates (or one rate) that would facilitate judgments in the changing circumstances Teletech faced What were the appropriate hurdle rates for the two segments Was Products and Systems underperforming as Yossarian suggested How should Teletech respond to the raider

-10- UVA-F-1152 shy

Exhibit 1

TELETECH CORPORATION 1996

Summary of W ACC Calculation for Teletech Corporation and

Segment Worksheet

Telecommunications Products and

Co]rate Services SYstems

MY Asset Weights 100 75 25

Bond Rating AA-A+ AA BBB-

Pre-tax Cost of Debt 703 700 778

Tax Rate 40 40 40

After-tax Cost of Debt 422 420 467

Equity Beta 104

Rf 604

RM-Rf 550

Cost of Equity 1177

Weight of Debt 180

Weight of Equity 820

WACC 1041 I

c

UVA-F-1152

Exhibit 2

TELETECH CORPORATION 1996

Theoretical Overview of Multiple Hurdle Rates

To Margaret Weston From Bernard Ingles Subject Theory of Segment Cost of Capital Date January 1996

You requested an overview of theories about multiple hurdle rates Without getting into minutiae the theories boil down to the following points

1 The central idea is that required returns should be driven by risk This is the dominant view in the field of investment management and is based on a mountain of theory and empirical research stretching over several decades The extension of this idea from investment management to corporate decision making is straightforward p~ least in theory

2 An underlying assumption is that the firm is transparent (Le that investors can see through the corporate veil and evaluate the activities going on inside) No one believes firms are completely transparent or that investors are perfectly informed But financial accounting standards have evolved toward making the firm more transparent And the investment community has grown tougher and sharper in its analysis Teletech now has 36 analysts publishing reports and forecasts on the firm The reality is that for big publicly held firms transparency is not a bad assumption

3 Another underlying assumption is that the value of the whole enterprise is simply the sum of its parts-this is the concept of Value Additivity We can define parts as either the business segments (on the left-hand side of the balance sheet) or the layers of the capital structure (on the right-hand side of the balance sheet) Market values (MVs) have to balance

MVTeteedl = ( MVTetlaquoomllllltlicaliolls Services + MV tmdmls+ iwellls) = ( MVdehl + MVltquiV)

Ifthese equalities did not hold then a raider could come along and exploit the inequality by buying or selling the whole and the parts This is arbitrage By buying and selling the actions of the raider would drive the MVs back into balance

-12- UVA-F-1152 r ~

Exhibit 2 (continued)

4 Investment theory tells us that the only risk that matters is nondiversifiable risk which is measured by beta Beta indicates the risk that an asset will add to a portfolio Because the investor is assumed to be diversified she is assumed to seek a return for only that risk that she cannot shed the nondiversifiabIe risk Now the important point here is that the beta ofa portfolio is equal to a weighted average of the betas of the portfolio components Extending this to the corporate environment the asset beta for the firm will equal a weighted average of the components of the firm-again the components of the firm can be defined in terms of either the right-hand side or the left-hand side of the balance sheet

Where

W = percentage weights based on market values

fJ Tel Serv fJ ps Asset betas for business segments

fJdebl fJ for the firms debt securities

fJeqUlIY= fJ of firms common stock (given by Bloomberg etc)

This is a very handy way to model the risk ofthe firm for it means that we can use the Capital Asset Pricing Model to estimate the cost of capital for a segment (ie using segment asset betas)

5 Given all the previous points it follows that the weighted average of the various costs ofcapital (K) for the firm (W ACC) which is the theoretically correct hurdle rate is simply a weighted average of segment W ACCs

Where

WTeCf WIS = market value weights

WACCTeLrv = (Wdebl TeLerv( K debl TelServ) + ( Weqllily TelServt K eltIUlly TelServ)

c

-13- UVA-F-1152

Exhibit 2 (continued)

6 The notion in point 5 may not hold exactly in practice First most of the components in the WACC formula are estimated with some error Second because oftaxes information asymmetries or other market imperfections assets may not be priced strictly in line with the model-for a company like Teletech it is reasonable to assume that any mispricings are just temporary Third the simple two-segment characterization ignores a hidden third segment the corporate treasury department that hedges and aims to finance the whole corporation optimally-this acts as a shock absorber for the financial policies of the segments Modeling the W ACC ofthe corporate treasury department is quite difficult Most companies assume that the impact of corporate treasury isnt very large and simply assume it away As a first cut we could do this too though it is an issue we should revisit

Conclusions

bull In theory the corporate WACC for Teletech is appropriate only for evaluating an asset having the same risk as the whole company It is not appropriate for assets having different risks than the whole company

bull Segment WACCs are computed similarly to corporate W ACCs

bull In concept the corporate W ACe is aweighted average of the segment W ACCs In practice the weighted-average concept may not hold due to imperfections in the market andor estimation errors

bull If we start computing segment W ACCs we must use the cost of debt cost of equity and weights appropriate to that segment We need a lot of information to do this correctly or else we really need to stretch to make assumptions

-14- UVA-F-1152_

Exhibit 3 TELETECH CORPORATION 1996

Samples of Comparable Firms

1995 Equity Asset Bond Book Val Pricel Mkt Val Mkt Val Pricel Revenues Seta Seta Iiaiiog OebtJCap Book DebtCap DebtlEq Eacniogs

Teletech Corporation $16000 1041 092 AA-tA+ 40 301 18 22 129

Telecommunications Services Industty ATampT $80000 090 085 AA 39 660 88 97 308 Alltel Corp 3160 075 063 A 49 299 243 322 160 Amerltech 13325 075 067 AA 47 472 158 188 169 BeD Atlantic 13500 080 068 AA 57 453 226 293 175 Bell South 17780 075 072 AAA 44 1190 62 66 190

Centul) Tel Enterprs 625 100 084 BBB+ 46 263 245 324 158

Cincinnati Bell 1350 080 069 AA 56 472 212 270 188 Citizens Utilities Co 1070 070 056 AA 42 168 301 431 158

Comsat 850 095 068 A 40 103 392 646 168

Frontier Corp 1750 080 074 A 42 550 116 132 283

GTE Corp 20250 080 066 BBB 69 652 254 341 169

MCI Communications 15100 125 114 A 24 187 144 169 179 NYNEX Corp 13425 075 059 Ashy 62 375 303 436 156

Pacific Telesis 9070 085 068 AAshy 74 694 291 410 136

SBC Communications 12560 090 080 A 54 559 173 210 180

Southem New England 1840 075 058 AA 5710 263 335 504 154

Sprint Corp

US West

Average

13550

9450

105

065

084

087

052

072

BBB

AAshy

52

66

51

313

467

452

257

294

228

347

416

311

150

143 C179

Telecommunications Equipment Industty ADC Telecomm Inc $586 135 135 NR 0 405 00 00 280

Aane-Cleveland 120 150 149 NR 1 151 07 07 128

ADen Group 325 160 155 NR 13 275 52 54 188

Andrew Corp 626 125 123 NR 13 440 33 34 197

DSC Communications 1450 130 126 NR 18 372 56 590 179

Newbridge NetworIs 675 155 155 NR 1 548 01 01 239

Qualcomm Inc 386 155 152 NR 9 341 28 29 583

Tellabs Inc 645 150 150 NR 1 796 01 01 266

Average 145 143 7 416 22 23 258

Computer and Network Equipment Industty Amdahl Corp $1500 13 120 NR 12 095 125 143 1~1

Bay Networks Inc 1342 175 174 NR 10 903 12 12 260

Cabletron Systems 1060 160 160 NR 0 657 00 00 218

Cisco Systems 1979 175 175 NR 0 1383 00 00 269

Digital Equipment 13813 110 104 NR 22 287 89 98 171

General Datacomm 221 185 179 NR 17 396 49 52 NMF

Hewlett-Packard 31519 125 124 NR 6 333 19 19 146

SCI Systems 2673 120 106 NR 32 220 176 214 127

Sequent Computer 535 195 192 NR 3 124 24 25 120

Standard Microsystems 345 160 152 NR 10 131 78 85 298

Stratus Computer 580 160 159 NR 2 136 15 15 133

Sun Microsystems 5902 155 154 NR 3 418 07 07 176

Tandem Computers 2285 155 150 NR 6 105 57 61 333

3ComCorp 1295 160 159 NR 12 1190 11 11 258

Average 155 151 10 456 47 53 204

-15- UVA-F-1152

Exhibit 4

TELETECH CORPORATION 1996

Debt Capital Market Conditions January 1996

Corporate Bond Yields by Rating US Treasurv Securities

Industrials AAA 650 Short-term bills 520 AA 700 Intermediate-term notes 543 A 764 Long-term bonds 604

BBB 778 BB 893 B 1049

Utilities AA 653 A 794 BBB 806

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

I

j

j

~~

c j

j

j

j

j

j

j

j

j

j

j

l j

j

j

j

-6- UVA-F-1152 _ (

My hunch is that the reverse is true for Products and Systems which promises to earn 120 percent on capital P+S is located on the graph near the little circle

Constant vs Risk-Adjusted Hurdle Rates

17

- Teletech Corp Hurdle

-- Telecomm Services

-e- A-oducts and Systems

Risk Level

In deciding how much to loan us lenders will consider the composition of risks If money flows into safer investments over time the cost of their loans to us will decrease

Our stockholders are just as much concerned with risk Ifthey perceive our business as being more risky than other pound9mpanies they will not pay as high a price for our earnings Perhaps this is why our priceearnings ratio is below the industry average most ofthe time It is not a question ofwhether we adjust for risk-we already do infonnally The only question in my mind is whether we make these adjustments systematically or not

While multiple hurdle rates may not reflect capital-structure changes on a day -to-day basis over time they will reflect prospects more realistically At the moment as I understand it our real problem is an inadequate and very costly supply of equity funds If we are really rationing equity capital then we should be striving for the best returns on equity for the risk Multiple hurdle rates achieve this objective

Implicit in Phillipss argument as Weston understood it was the notion that ifeach segment in the company had a different hurdle rate the costs ofthe various fonns ofcapital would remain the same However the mix of capital used would change in the calculation Low-risk operations ~

c

-7- UVA-F-1152

would use leverage more extensively while the high-risk divisions would have little or no debt funds This lower-risk segment would have a lower hurdle rate

Opposition to risk-adjusted hurdle rates

Phillipss views were supported by several others within Teletech opposition was just as strong however particularly within the Products and Systems segment Helen Buono executive vice president for the segment expressed her opinion as follows

All money is green Investors cant know as much about our operations as we do To them the firm is an opaque box they hire us to take care ofwhat is inside the box andjudge us by the dividends coming out of the box We cant say that one part of the box has a different hurdle rate than another part ofthe box ifour investors dont think that way Like I say all money is green all investments at Teletech should be judged against one hurdle rate

Multiple hurdle rates are illogical Suppose that the hurdle rate for Teleshycommunications Services was much lower than the corporatewide hurdle rate Ifwe undertook investments that met the segment hurdle rate we would be destroying shareholder value because we werent meeting the (orporate hurdle rate shy

Ourjob as managers should be to put our money where the returns are best A single hurdle rate may deprive an underprofitable division of investments in order to channel more funds into a more profitable division but isnt that the aim of the process Our challenge today is simple we must earn the highest absolute rates of return we can get

In reality we dont finance each division separately The corporation raises capital based on its overall prospects and record The diversification of the company probably helps keep our capital C0sts down and enables us to borrow more in total than the sum of the capabilities of the divisions separately As a result developing separate hurdle rates is both unrealistic and misleading All our stockholders want is for us to invest our funds wisely in order to increase the value of their stock This happens when we pick the most promising projects~ irrespective of their source

Margaret Westons Concerns

As she listened to these arguments presented over the course of several months Weston became increasingly concerned with several related considerations First the corporate strategy directed the company toward integrating the two divisions One effect ofusing multiple hurdle rates would be to make justifying high-technology research and application proposals more difficult as the required rate ofreturn would be increased Perhaps she thought multiple hurdle rates were the

-8- UVA-F-1152 r f

right idea but the notion that they should be based on capital costs rather than strategic considerations might be wrong On the other hand perhaps multiple rates based on capital costs should be used but in allocating funds some qualitative adjustment should be made for unquantifiable strategic considerations In Westons mind theory was certainly not clear on how to achieve strategic objectives when allocating capital

Second using a single measure ofthe cost ofmoney (the hurdle rate or discount factor) made the net present value results consistent at least in economic terms If Teletech adopted multiple rates for discounting cash flows Weston was afraid the NPV and economic profit calculations would lose their meaning and comparability across business segments To her a performance criterion had to be consistent and understandable or it would not be useful

In addition Weston was concerned with the problem of attributing capital structures to divisions In Telecommunications Services a major new switching station might be financed by mortgage bonds But in Products and Systems it was not possible for the division to borrow directly indeed any financing was feasible only because the corporation guaranteed the debt Such projects were considered highly risky-perhaps at best warranting only a minimal debt structure Also Weston considered the debt-capacity decision difficult enough to make for the corporation as a whole let alone for each division Judgments could only be very crude

In further discussions with those in the organization about the use of multiple hurdle rates Weston ran across two predominant trains of thought One argument held that the investment C decision should never be mixed with the financing decisionr A firm should decide what its investments should be and then how to finance them most efficiently Adding leverage to a presentshyvalue calculation would distort the results Use ofmultiple hurdle rates was simply a way ofmixing financing with investment analysis This argument also held that a single rate left the risk decision clear-cut management could simply adjust its standard (NPV or economic profit) as risks increased

The contrasting line of reasoning noted that the weighted-average cost of capital tended to represent an average market reaction tol mixture ofrisks Lower-than-average-risk projects should probably be accepted even though they did not meet a weighted-average criterion Higher-than-normal-risk projects should provide a return premium While the multiple-hurdle-rate system was a crude way ofachieving this end it at least was a step in the right direction Moreover some argued that Teletechs objective should be to maximize return on equity funds and because

equity funds were and would remain a comparatively scarce resource a multiple-rate system would tend to maximize returns to stockholders better than a single-rate system

To help resolve these questions Weston asked her assistant Bernard Ingles to summarize academic thinking about multiple hurdle rates His memorandum is given in Exhibit 2 She also requested that he draw samples ofcomparable firms for Telecommunications Services and Products and Systems that might be used in deriving segment WACCs The summary of data is given in Exhibit 3 Information on capital-market conditions in January 1996 is given in Exhibit 4

-9- UVA-F-1l52

Conclusion

Weston could not realistically hope that all the issues before her would be resolved in time to influence Victor Yossarians attack on management But the attack did dictate the need for an objective assessment of the performance of Teletechs two segments-the choice of hurdle rates would be very important in this analysis However she did want to institute a pragmatic system of appropriate hurdle rates (or one rate) that would facilitate judgments in the changing circumstances Teletech faced What were the appropriate hurdle rates for the two segments Was Products and Systems underperforming as Yossarian suggested How should Teletech respond to the raider

-10- UVA-F-1152 shy

Exhibit 1

TELETECH CORPORATION 1996

Summary of W ACC Calculation for Teletech Corporation and

Segment Worksheet

Telecommunications Products and

Co]rate Services SYstems

MY Asset Weights 100 75 25

Bond Rating AA-A+ AA BBB-

Pre-tax Cost of Debt 703 700 778

Tax Rate 40 40 40

After-tax Cost of Debt 422 420 467

Equity Beta 104

Rf 604

RM-Rf 550

Cost of Equity 1177

Weight of Debt 180

Weight of Equity 820

WACC 1041 I

c

UVA-F-1152

Exhibit 2

TELETECH CORPORATION 1996

Theoretical Overview of Multiple Hurdle Rates

To Margaret Weston From Bernard Ingles Subject Theory of Segment Cost of Capital Date January 1996

You requested an overview of theories about multiple hurdle rates Without getting into minutiae the theories boil down to the following points

1 The central idea is that required returns should be driven by risk This is the dominant view in the field of investment management and is based on a mountain of theory and empirical research stretching over several decades The extension of this idea from investment management to corporate decision making is straightforward p~ least in theory

2 An underlying assumption is that the firm is transparent (Le that investors can see through the corporate veil and evaluate the activities going on inside) No one believes firms are completely transparent or that investors are perfectly informed But financial accounting standards have evolved toward making the firm more transparent And the investment community has grown tougher and sharper in its analysis Teletech now has 36 analysts publishing reports and forecasts on the firm The reality is that for big publicly held firms transparency is not a bad assumption

3 Another underlying assumption is that the value of the whole enterprise is simply the sum of its parts-this is the concept of Value Additivity We can define parts as either the business segments (on the left-hand side of the balance sheet) or the layers of the capital structure (on the right-hand side of the balance sheet) Market values (MVs) have to balance

MVTeteedl = ( MVTetlaquoomllllltlicaliolls Services + MV tmdmls+ iwellls) = ( MVdehl + MVltquiV)

Ifthese equalities did not hold then a raider could come along and exploit the inequality by buying or selling the whole and the parts This is arbitrage By buying and selling the actions of the raider would drive the MVs back into balance

-12- UVA-F-1152 r ~

Exhibit 2 (continued)

4 Investment theory tells us that the only risk that matters is nondiversifiable risk which is measured by beta Beta indicates the risk that an asset will add to a portfolio Because the investor is assumed to be diversified she is assumed to seek a return for only that risk that she cannot shed the nondiversifiabIe risk Now the important point here is that the beta ofa portfolio is equal to a weighted average of the betas of the portfolio components Extending this to the corporate environment the asset beta for the firm will equal a weighted average of the components of the firm-again the components of the firm can be defined in terms of either the right-hand side or the left-hand side of the balance sheet

Where

W = percentage weights based on market values

fJ Tel Serv fJ ps Asset betas for business segments

fJdebl fJ for the firms debt securities

fJeqUlIY= fJ of firms common stock (given by Bloomberg etc)

This is a very handy way to model the risk ofthe firm for it means that we can use the Capital Asset Pricing Model to estimate the cost of capital for a segment (ie using segment asset betas)

5 Given all the previous points it follows that the weighted average of the various costs ofcapital (K) for the firm (W ACC) which is the theoretically correct hurdle rate is simply a weighted average of segment W ACCs

Where

WTeCf WIS = market value weights

WACCTeLrv = (Wdebl TeLerv( K debl TelServ) + ( Weqllily TelServt K eltIUlly TelServ)

c

-13- UVA-F-1152

Exhibit 2 (continued)

6 The notion in point 5 may not hold exactly in practice First most of the components in the WACC formula are estimated with some error Second because oftaxes information asymmetries or other market imperfections assets may not be priced strictly in line with the model-for a company like Teletech it is reasonable to assume that any mispricings are just temporary Third the simple two-segment characterization ignores a hidden third segment the corporate treasury department that hedges and aims to finance the whole corporation optimally-this acts as a shock absorber for the financial policies of the segments Modeling the W ACC ofthe corporate treasury department is quite difficult Most companies assume that the impact of corporate treasury isnt very large and simply assume it away As a first cut we could do this too though it is an issue we should revisit

Conclusions

bull In theory the corporate WACC for Teletech is appropriate only for evaluating an asset having the same risk as the whole company It is not appropriate for assets having different risks than the whole company

bull Segment WACCs are computed similarly to corporate W ACCs

bull In concept the corporate W ACe is aweighted average of the segment W ACCs In practice the weighted-average concept may not hold due to imperfections in the market andor estimation errors

bull If we start computing segment W ACCs we must use the cost of debt cost of equity and weights appropriate to that segment We need a lot of information to do this correctly or else we really need to stretch to make assumptions

-14- UVA-F-1152_

Exhibit 3 TELETECH CORPORATION 1996

Samples of Comparable Firms

1995 Equity Asset Bond Book Val Pricel Mkt Val Mkt Val Pricel Revenues Seta Seta Iiaiiog OebtJCap Book DebtCap DebtlEq Eacniogs

Teletech Corporation $16000 1041 092 AA-tA+ 40 301 18 22 129

Telecommunications Services Industty ATampT $80000 090 085 AA 39 660 88 97 308 Alltel Corp 3160 075 063 A 49 299 243 322 160 Amerltech 13325 075 067 AA 47 472 158 188 169 BeD Atlantic 13500 080 068 AA 57 453 226 293 175 Bell South 17780 075 072 AAA 44 1190 62 66 190

Centul) Tel Enterprs 625 100 084 BBB+ 46 263 245 324 158

Cincinnati Bell 1350 080 069 AA 56 472 212 270 188 Citizens Utilities Co 1070 070 056 AA 42 168 301 431 158

Comsat 850 095 068 A 40 103 392 646 168

Frontier Corp 1750 080 074 A 42 550 116 132 283

GTE Corp 20250 080 066 BBB 69 652 254 341 169

MCI Communications 15100 125 114 A 24 187 144 169 179 NYNEX Corp 13425 075 059 Ashy 62 375 303 436 156

Pacific Telesis 9070 085 068 AAshy 74 694 291 410 136

SBC Communications 12560 090 080 A 54 559 173 210 180

Southem New England 1840 075 058 AA 5710 263 335 504 154

Sprint Corp

US West

Average

13550

9450

105

065

084

087

052

072

BBB

AAshy

52

66

51

313

467

452

257

294

228

347

416

311

150

143 C179

Telecommunications Equipment Industty ADC Telecomm Inc $586 135 135 NR 0 405 00 00 280

Aane-Cleveland 120 150 149 NR 1 151 07 07 128

ADen Group 325 160 155 NR 13 275 52 54 188

Andrew Corp 626 125 123 NR 13 440 33 34 197

DSC Communications 1450 130 126 NR 18 372 56 590 179

Newbridge NetworIs 675 155 155 NR 1 548 01 01 239

Qualcomm Inc 386 155 152 NR 9 341 28 29 583

Tellabs Inc 645 150 150 NR 1 796 01 01 266

Average 145 143 7 416 22 23 258

Computer and Network Equipment Industty Amdahl Corp $1500 13 120 NR 12 095 125 143 1~1

Bay Networks Inc 1342 175 174 NR 10 903 12 12 260

Cabletron Systems 1060 160 160 NR 0 657 00 00 218

Cisco Systems 1979 175 175 NR 0 1383 00 00 269

Digital Equipment 13813 110 104 NR 22 287 89 98 171

General Datacomm 221 185 179 NR 17 396 49 52 NMF

Hewlett-Packard 31519 125 124 NR 6 333 19 19 146

SCI Systems 2673 120 106 NR 32 220 176 214 127

Sequent Computer 535 195 192 NR 3 124 24 25 120

Standard Microsystems 345 160 152 NR 10 131 78 85 298

Stratus Computer 580 160 159 NR 2 136 15 15 133

Sun Microsystems 5902 155 154 NR 3 418 07 07 176

Tandem Computers 2285 155 150 NR 6 105 57 61 333

3ComCorp 1295 160 159 NR 12 1190 11 11 258

Average 155 151 10 456 47 53 204

-15- UVA-F-1152

Exhibit 4

TELETECH CORPORATION 1996

Debt Capital Market Conditions January 1996

Corporate Bond Yields by Rating US Treasurv Securities

Industrials AAA 650 Short-term bills 520 AA 700 Intermediate-term notes 543 A 764 Long-term bonds 604

BBB 778 BB 893 B 1049

Utilities AA 653 A 794 BBB 806

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

I

j

j

~~

c j

j

j

j

j

j

j

j

j

j

j

l j

j

j

j

-7- UVA-F-1152

would use leverage more extensively while the high-risk divisions would have little or no debt funds This lower-risk segment would have a lower hurdle rate

Opposition to risk-adjusted hurdle rates

Phillipss views were supported by several others within Teletech opposition was just as strong however particularly within the Products and Systems segment Helen Buono executive vice president for the segment expressed her opinion as follows

All money is green Investors cant know as much about our operations as we do To them the firm is an opaque box they hire us to take care ofwhat is inside the box andjudge us by the dividends coming out of the box We cant say that one part of the box has a different hurdle rate than another part ofthe box ifour investors dont think that way Like I say all money is green all investments at Teletech should be judged against one hurdle rate

Multiple hurdle rates are illogical Suppose that the hurdle rate for Teleshycommunications Services was much lower than the corporatewide hurdle rate Ifwe undertook investments that met the segment hurdle rate we would be destroying shareholder value because we werent meeting the (orporate hurdle rate shy

Ourjob as managers should be to put our money where the returns are best A single hurdle rate may deprive an underprofitable division of investments in order to channel more funds into a more profitable division but isnt that the aim of the process Our challenge today is simple we must earn the highest absolute rates of return we can get

In reality we dont finance each division separately The corporation raises capital based on its overall prospects and record The diversification of the company probably helps keep our capital C0sts down and enables us to borrow more in total than the sum of the capabilities of the divisions separately As a result developing separate hurdle rates is both unrealistic and misleading All our stockholders want is for us to invest our funds wisely in order to increase the value of their stock This happens when we pick the most promising projects~ irrespective of their source

Margaret Westons Concerns

As she listened to these arguments presented over the course of several months Weston became increasingly concerned with several related considerations First the corporate strategy directed the company toward integrating the two divisions One effect ofusing multiple hurdle rates would be to make justifying high-technology research and application proposals more difficult as the required rate ofreturn would be increased Perhaps she thought multiple hurdle rates were the

-8- UVA-F-1152 r f

right idea but the notion that they should be based on capital costs rather than strategic considerations might be wrong On the other hand perhaps multiple rates based on capital costs should be used but in allocating funds some qualitative adjustment should be made for unquantifiable strategic considerations In Westons mind theory was certainly not clear on how to achieve strategic objectives when allocating capital

Second using a single measure ofthe cost ofmoney (the hurdle rate or discount factor) made the net present value results consistent at least in economic terms If Teletech adopted multiple rates for discounting cash flows Weston was afraid the NPV and economic profit calculations would lose their meaning and comparability across business segments To her a performance criterion had to be consistent and understandable or it would not be useful

In addition Weston was concerned with the problem of attributing capital structures to divisions In Telecommunications Services a major new switching station might be financed by mortgage bonds But in Products and Systems it was not possible for the division to borrow directly indeed any financing was feasible only because the corporation guaranteed the debt Such projects were considered highly risky-perhaps at best warranting only a minimal debt structure Also Weston considered the debt-capacity decision difficult enough to make for the corporation as a whole let alone for each division Judgments could only be very crude

In further discussions with those in the organization about the use of multiple hurdle rates Weston ran across two predominant trains of thought One argument held that the investment C decision should never be mixed with the financing decisionr A firm should decide what its investments should be and then how to finance them most efficiently Adding leverage to a presentshyvalue calculation would distort the results Use ofmultiple hurdle rates was simply a way ofmixing financing with investment analysis This argument also held that a single rate left the risk decision clear-cut management could simply adjust its standard (NPV or economic profit) as risks increased

The contrasting line of reasoning noted that the weighted-average cost of capital tended to represent an average market reaction tol mixture ofrisks Lower-than-average-risk projects should probably be accepted even though they did not meet a weighted-average criterion Higher-than-normal-risk projects should provide a return premium While the multiple-hurdle-rate system was a crude way ofachieving this end it at least was a step in the right direction Moreover some argued that Teletechs objective should be to maximize return on equity funds and because

equity funds were and would remain a comparatively scarce resource a multiple-rate system would tend to maximize returns to stockholders better than a single-rate system

To help resolve these questions Weston asked her assistant Bernard Ingles to summarize academic thinking about multiple hurdle rates His memorandum is given in Exhibit 2 She also requested that he draw samples ofcomparable firms for Telecommunications Services and Products and Systems that might be used in deriving segment WACCs The summary of data is given in Exhibit 3 Information on capital-market conditions in January 1996 is given in Exhibit 4

-9- UVA-F-1l52

Conclusion

Weston could not realistically hope that all the issues before her would be resolved in time to influence Victor Yossarians attack on management But the attack did dictate the need for an objective assessment of the performance of Teletechs two segments-the choice of hurdle rates would be very important in this analysis However she did want to institute a pragmatic system of appropriate hurdle rates (or one rate) that would facilitate judgments in the changing circumstances Teletech faced What were the appropriate hurdle rates for the two segments Was Products and Systems underperforming as Yossarian suggested How should Teletech respond to the raider

-10- UVA-F-1152 shy

Exhibit 1

TELETECH CORPORATION 1996

Summary of W ACC Calculation for Teletech Corporation and

Segment Worksheet

Telecommunications Products and

Co]rate Services SYstems

MY Asset Weights 100 75 25

Bond Rating AA-A+ AA BBB-

Pre-tax Cost of Debt 703 700 778

Tax Rate 40 40 40

After-tax Cost of Debt 422 420 467

Equity Beta 104

Rf 604

RM-Rf 550

Cost of Equity 1177

Weight of Debt 180

Weight of Equity 820

WACC 1041 I

c

UVA-F-1152

Exhibit 2

TELETECH CORPORATION 1996

Theoretical Overview of Multiple Hurdle Rates

To Margaret Weston From Bernard Ingles Subject Theory of Segment Cost of Capital Date January 1996

You requested an overview of theories about multiple hurdle rates Without getting into minutiae the theories boil down to the following points

1 The central idea is that required returns should be driven by risk This is the dominant view in the field of investment management and is based on a mountain of theory and empirical research stretching over several decades The extension of this idea from investment management to corporate decision making is straightforward p~ least in theory

2 An underlying assumption is that the firm is transparent (Le that investors can see through the corporate veil and evaluate the activities going on inside) No one believes firms are completely transparent or that investors are perfectly informed But financial accounting standards have evolved toward making the firm more transparent And the investment community has grown tougher and sharper in its analysis Teletech now has 36 analysts publishing reports and forecasts on the firm The reality is that for big publicly held firms transparency is not a bad assumption

3 Another underlying assumption is that the value of the whole enterprise is simply the sum of its parts-this is the concept of Value Additivity We can define parts as either the business segments (on the left-hand side of the balance sheet) or the layers of the capital structure (on the right-hand side of the balance sheet) Market values (MVs) have to balance

MVTeteedl = ( MVTetlaquoomllllltlicaliolls Services + MV tmdmls+ iwellls) = ( MVdehl + MVltquiV)

Ifthese equalities did not hold then a raider could come along and exploit the inequality by buying or selling the whole and the parts This is arbitrage By buying and selling the actions of the raider would drive the MVs back into balance

-12- UVA-F-1152 r ~

Exhibit 2 (continued)

4 Investment theory tells us that the only risk that matters is nondiversifiable risk which is measured by beta Beta indicates the risk that an asset will add to a portfolio Because the investor is assumed to be diversified she is assumed to seek a return for only that risk that she cannot shed the nondiversifiabIe risk Now the important point here is that the beta ofa portfolio is equal to a weighted average of the betas of the portfolio components Extending this to the corporate environment the asset beta for the firm will equal a weighted average of the components of the firm-again the components of the firm can be defined in terms of either the right-hand side or the left-hand side of the balance sheet

Where

W = percentage weights based on market values

fJ Tel Serv fJ ps Asset betas for business segments

fJdebl fJ for the firms debt securities

fJeqUlIY= fJ of firms common stock (given by Bloomberg etc)

This is a very handy way to model the risk ofthe firm for it means that we can use the Capital Asset Pricing Model to estimate the cost of capital for a segment (ie using segment asset betas)

5 Given all the previous points it follows that the weighted average of the various costs ofcapital (K) for the firm (W ACC) which is the theoretically correct hurdle rate is simply a weighted average of segment W ACCs

Where

WTeCf WIS = market value weights

WACCTeLrv = (Wdebl TeLerv( K debl TelServ) + ( Weqllily TelServt K eltIUlly TelServ)

c

-13- UVA-F-1152

Exhibit 2 (continued)

6 The notion in point 5 may not hold exactly in practice First most of the components in the WACC formula are estimated with some error Second because oftaxes information asymmetries or other market imperfections assets may not be priced strictly in line with the model-for a company like Teletech it is reasonable to assume that any mispricings are just temporary Third the simple two-segment characterization ignores a hidden third segment the corporate treasury department that hedges and aims to finance the whole corporation optimally-this acts as a shock absorber for the financial policies of the segments Modeling the W ACC ofthe corporate treasury department is quite difficult Most companies assume that the impact of corporate treasury isnt very large and simply assume it away As a first cut we could do this too though it is an issue we should revisit

Conclusions

bull In theory the corporate WACC for Teletech is appropriate only for evaluating an asset having the same risk as the whole company It is not appropriate for assets having different risks than the whole company

bull Segment WACCs are computed similarly to corporate W ACCs

bull In concept the corporate W ACe is aweighted average of the segment W ACCs In practice the weighted-average concept may not hold due to imperfections in the market andor estimation errors

bull If we start computing segment W ACCs we must use the cost of debt cost of equity and weights appropriate to that segment We need a lot of information to do this correctly or else we really need to stretch to make assumptions

-14- UVA-F-1152_

Exhibit 3 TELETECH CORPORATION 1996

Samples of Comparable Firms

1995 Equity Asset Bond Book Val Pricel Mkt Val Mkt Val Pricel Revenues Seta Seta Iiaiiog OebtJCap Book DebtCap DebtlEq Eacniogs

Teletech Corporation $16000 1041 092 AA-tA+ 40 301 18 22 129

Telecommunications Services Industty ATampT $80000 090 085 AA 39 660 88 97 308 Alltel Corp 3160 075 063 A 49 299 243 322 160 Amerltech 13325 075 067 AA 47 472 158 188 169 BeD Atlantic 13500 080 068 AA 57 453 226 293 175 Bell South 17780 075 072 AAA 44 1190 62 66 190

Centul) Tel Enterprs 625 100 084 BBB+ 46 263 245 324 158

Cincinnati Bell 1350 080 069 AA 56 472 212 270 188 Citizens Utilities Co 1070 070 056 AA 42 168 301 431 158

Comsat 850 095 068 A 40 103 392 646 168

Frontier Corp 1750 080 074 A 42 550 116 132 283

GTE Corp 20250 080 066 BBB 69 652 254 341 169

MCI Communications 15100 125 114 A 24 187 144 169 179 NYNEX Corp 13425 075 059 Ashy 62 375 303 436 156

Pacific Telesis 9070 085 068 AAshy 74 694 291 410 136

SBC Communications 12560 090 080 A 54 559 173 210 180

Southem New England 1840 075 058 AA 5710 263 335 504 154

Sprint Corp

US West

Average

13550

9450

105

065

084

087

052

072

BBB

AAshy

52

66

51

313

467

452

257

294

228

347

416

311

150

143 C179

Telecommunications Equipment Industty ADC Telecomm Inc $586 135 135 NR 0 405 00 00 280

Aane-Cleveland 120 150 149 NR 1 151 07 07 128

ADen Group 325 160 155 NR 13 275 52 54 188

Andrew Corp 626 125 123 NR 13 440 33 34 197

DSC Communications 1450 130 126 NR 18 372 56 590 179

Newbridge NetworIs 675 155 155 NR 1 548 01 01 239

Qualcomm Inc 386 155 152 NR 9 341 28 29 583

Tellabs Inc 645 150 150 NR 1 796 01 01 266

Average 145 143 7 416 22 23 258

Computer and Network Equipment Industty Amdahl Corp $1500 13 120 NR 12 095 125 143 1~1

Bay Networks Inc 1342 175 174 NR 10 903 12 12 260

Cabletron Systems 1060 160 160 NR 0 657 00 00 218

Cisco Systems 1979 175 175 NR 0 1383 00 00 269

Digital Equipment 13813 110 104 NR 22 287 89 98 171

General Datacomm 221 185 179 NR 17 396 49 52 NMF

Hewlett-Packard 31519 125 124 NR 6 333 19 19 146

SCI Systems 2673 120 106 NR 32 220 176 214 127

Sequent Computer 535 195 192 NR 3 124 24 25 120

Standard Microsystems 345 160 152 NR 10 131 78 85 298

Stratus Computer 580 160 159 NR 2 136 15 15 133

Sun Microsystems 5902 155 154 NR 3 418 07 07 176

Tandem Computers 2285 155 150 NR 6 105 57 61 333

3ComCorp 1295 160 159 NR 12 1190 11 11 258

Average 155 151 10 456 47 53 204

-15- UVA-F-1152

Exhibit 4

TELETECH CORPORATION 1996

Debt Capital Market Conditions January 1996

Corporate Bond Yields by Rating US Treasurv Securities

Industrials AAA 650 Short-term bills 520 AA 700 Intermediate-term notes 543 A 764 Long-term bonds 604

BBB 778 BB 893 B 1049

Utilities AA 653 A 794 BBB 806

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

I

j

j

~~

c j

j

j

j

j

j

j

j

j

j

j

l j

j

j

j

-8- UVA-F-1152 r f

right idea but the notion that they should be based on capital costs rather than strategic considerations might be wrong On the other hand perhaps multiple rates based on capital costs should be used but in allocating funds some qualitative adjustment should be made for unquantifiable strategic considerations In Westons mind theory was certainly not clear on how to achieve strategic objectives when allocating capital

Second using a single measure ofthe cost ofmoney (the hurdle rate or discount factor) made the net present value results consistent at least in economic terms If Teletech adopted multiple rates for discounting cash flows Weston was afraid the NPV and economic profit calculations would lose their meaning and comparability across business segments To her a performance criterion had to be consistent and understandable or it would not be useful

In addition Weston was concerned with the problem of attributing capital structures to divisions In Telecommunications Services a major new switching station might be financed by mortgage bonds But in Products and Systems it was not possible for the division to borrow directly indeed any financing was feasible only because the corporation guaranteed the debt Such projects were considered highly risky-perhaps at best warranting only a minimal debt structure Also Weston considered the debt-capacity decision difficult enough to make for the corporation as a whole let alone for each division Judgments could only be very crude

In further discussions with those in the organization about the use of multiple hurdle rates Weston ran across two predominant trains of thought One argument held that the investment C decision should never be mixed with the financing decisionr A firm should decide what its investments should be and then how to finance them most efficiently Adding leverage to a presentshyvalue calculation would distort the results Use ofmultiple hurdle rates was simply a way ofmixing financing with investment analysis This argument also held that a single rate left the risk decision clear-cut management could simply adjust its standard (NPV or economic profit) as risks increased

The contrasting line of reasoning noted that the weighted-average cost of capital tended to represent an average market reaction tol mixture ofrisks Lower-than-average-risk projects should probably be accepted even though they did not meet a weighted-average criterion Higher-than-normal-risk projects should provide a return premium While the multiple-hurdle-rate system was a crude way ofachieving this end it at least was a step in the right direction Moreover some argued that Teletechs objective should be to maximize return on equity funds and because

equity funds were and would remain a comparatively scarce resource a multiple-rate system would tend to maximize returns to stockholders better than a single-rate system

To help resolve these questions Weston asked her assistant Bernard Ingles to summarize academic thinking about multiple hurdle rates His memorandum is given in Exhibit 2 She also requested that he draw samples ofcomparable firms for Telecommunications Services and Products and Systems that might be used in deriving segment WACCs The summary of data is given in Exhibit 3 Information on capital-market conditions in January 1996 is given in Exhibit 4

-9- UVA-F-1l52

Conclusion

Weston could not realistically hope that all the issues before her would be resolved in time to influence Victor Yossarians attack on management But the attack did dictate the need for an objective assessment of the performance of Teletechs two segments-the choice of hurdle rates would be very important in this analysis However she did want to institute a pragmatic system of appropriate hurdle rates (or one rate) that would facilitate judgments in the changing circumstances Teletech faced What were the appropriate hurdle rates for the two segments Was Products and Systems underperforming as Yossarian suggested How should Teletech respond to the raider

-10- UVA-F-1152 shy

Exhibit 1

TELETECH CORPORATION 1996

Summary of W ACC Calculation for Teletech Corporation and

Segment Worksheet

Telecommunications Products and

Co]rate Services SYstems

MY Asset Weights 100 75 25

Bond Rating AA-A+ AA BBB-

Pre-tax Cost of Debt 703 700 778

Tax Rate 40 40 40

After-tax Cost of Debt 422 420 467

Equity Beta 104

Rf 604

RM-Rf 550

Cost of Equity 1177

Weight of Debt 180

Weight of Equity 820

WACC 1041 I

c

UVA-F-1152

Exhibit 2

TELETECH CORPORATION 1996

Theoretical Overview of Multiple Hurdle Rates

To Margaret Weston From Bernard Ingles Subject Theory of Segment Cost of Capital Date January 1996

You requested an overview of theories about multiple hurdle rates Without getting into minutiae the theories boil down to the following points

1 The central idea is that required returns should be driven by risk This is the dominant view in the field of investment management and is based on a mountain of theory and empirical research stretching over several decades The extension of this idea from investment management to corporate decision making is straightforward p~ least in theory

2 An underlying assumption is that the firm is transparent (Le that investors can see through the corporate veil and evaluate the activities going on inside) No one believes firms are completely transparent or that investors are perfectly informed But financial accounting standards have evolved toward making the firm more transparent And the investment community has grown tougher and sharper in its analysis Teletech now has 36 analysts publishing reports and forecasts on the firm The reality is that for big publicly held firms transparency is not a bad assumption

3 Another underlying assumption is that the value of the whole enterprise is simply the sum of its parts-this is the concept of Value Additivity We can define parts as either the business segments (on the left-hand side of the balance sheet) or the layers of the capital structure (on the right-hand side of the balance sheet) Market values (MVs) have to balance

MVTeteedl = ( MVTetlaquoomllllltlicaliolls Services + MV tmdmls+ iwellls) = ( MVdehl + MVltquiV)

Ifthese equalities did not hold then a raider could come along and exploit the inequality by buying or selling the whole and the parts This is arbitrage By buying and selling the actions of the raider would drive the MVs back into balance

-12- UVA-F-1152 r ~

Exhibit 2 (continued)

4 Investment theory tells us that the only risk that matters is nondiversifiable risk which is measured by beta Beta indicates the risk that an asset will add to a portfolio Because the investor is assumed to be diversified she is assumed to seek a return for only that risk that she cannot shed the nondiversifiabIe risk Now the important point here is that the beta ofa portfolio is equal to a weighted average of the betas of the portfolio components Extending this to the corporate environment the asset beta for the firm will equal a weighted average of the components of the firm-again the components of the firm can be defined in terms of either the right-hand side or the left-hand side of the balance sheet

Where

W = percentage weights based on market values

fJ Tel Serv fJ ps Asset betas for business segments

fJdebl fJ for the firms debt securities

fJeqUlIY= fJ of firms common stock (given by Bloomberg etc)

This is a very handy way to model the risk ofthe firm for it means that we can use the Capital Asset Pricing Model to estimate the cost of capital for a segment (ie using segment asset betas)

5 Given all the previous points it follows that the weighted average of the various costs ofcapital (K) for the firm (W ACC) which is the theoretically correct hurdle rate is simply a weighted average of segment W ACCs

Where

WTeCf WIS = market value weights

WACCTeLrv = (Wdebl TeLerv( K debl TelServ) + ( Weqllily TelServt K eltIUlly TelServ)

c

-13- UVA-F-1152

Exhibit 2 (continued)

6 The notion in point 5 may not hold exactly in practice First most of the components in the WACC formula are estimated with some error Second because oftaxes information asymmetries or other market imperfections assets may not be priced strictly in line with the model-for a company like Teletech it is reasonable to assume that any mispricings are just temporary Third the simple two-segment characterization ignores a hidden third segment the corporate treasury department that hedges and aims to finance the whole corporation optimally-this acts as a shock absorber for the financial policies of the segments Modeling the W ACC ofthe corporate treasury department is quite difficult Most companies assume that the impact of corporate treasury isnt very large and simply assume it away As a first cut we could do this too though it is an issue we should revisit

Conclusions

bull In theory the corporate WACC for Teletech is appropriate only for evaluating an asset having the same risk as the whole company It is not appropriate for assets having different risks than the whole company

bull Segment WACCs are computed similarly to corporate W ACCs

bull In concept the corporate W ACe is aweighted average of the segment W ACCs In practice the weighted-average concept may not hold due to imperfections in the market andor estimation errors

bull If we start computing segment W ACCs we must use the cost of debt cost of equity and weights appropriate to that segment We need a lot of information to do this correctly or else we really need to stretch to make assumptions

-14- UVA-F-1152_

Exhibit 3 TELETECH CORPORATION 1996

Samples of Comparable Firms

1995 Equity Asset Bond Book Val Pricel Mkt Val Mkt Val Pricel Revenues Seta Seta Iiaiiog OebtJCap Book DebtCap DebtlEq Eacniogs

Teletech Corporation $16000 1041 092 AA-tA+ 40 301 18 22 129

Telecommunications Services Industty ATampT $80000 090 085 AA 39 660 88 97 308 Alltel Corp 3160 075 063 A 49 299 243 322 160 Amerltech 13325 075 067 AA 47 472 158 188 169 BeD Atlantic 13500 080 068 AA 57 453 226 293 175 Bell South 17780 075 072 AAA 44 1190 62 66 190

Centul) Tel Enterprs 625 100 084 BBB+ 46 263 245 324 158

Cincinnati Bell 1350 080 069 AA 56 472 212 270 188 Citizens Utilities Co 1070 070 056 AA 42 168 301 431 158

Comsat 850 095 068 A 40 103 392 646 168

Frontier Corp 1750 080 074 A 42 550 116 132 283

GTE Corp 20250 080 066 BBB 69 652 254 341 169

MCI Communications 15100 125 114 A 24 187 144 169 179 NYNEX Corp 13425 075 059 Ashy 62 375 303 436 156

Pacific Telesis 9070 085 068 AAshy 74 694 291 410 136

SBC Communications 12560 090 080 A 54 559 173 210 180

Southem New England 1840 075 058 AA 5710 263 335 504 154

Sprint Corp

US West

Average

13550

9450

105

065

084

087

052

072

BBB

AAshy

52

66

51

313

467

452

257

294

228

347

416

311

150

143 C179

Telecommunications Equipment Industty ADC Telecomm Inc $586 135 135 NR 0 405 00 00 280

Aane-Cleveland 120 150 149 NR 1 151 07 07 128

ADen Group 325 160 155 NR 13 275 52 54 188

Andrew Corp 626 125 123 NR 13 440 33 34 197

DSC Communications 1450 130 126 NR 18 372 56 590 179

Newbridge NetworIs 675 155 155 NR 1 548 01 01 239

Qualcomm Inc 386 155 152 NR 9 341 28 29 583

Tellabs Inc 645 150 150 NR 1 796 01 01 266

Average 145 143 7 416 22 23 258

Computer and Network Equipment Industty Amdahl Corp $1500 13 120 NR 12 095 125 143 1~1

Bay Networks Inc 1342 175 174 NR 10 903 12 12 260

Cabletron Systems 1060 160 160 NR 0 657 00 00 218

Cisco Systems 1979 175 175 NR 0 1383 00 00 269

Digital Equipment 13813 110 104 NR 22 287 89 98 171

General Datacomm 221 185 179 NR 17 396 49 52 NMF

Hewlett-Packard 31519 125 124 NR 6 333 19 19 146

SCI Systems 2673 120 106 NR 32 220 176 214 127

Sequent Computer 535 195 192 NR 3 124 24 25 120

Standard Microsystems 345 160 152 NR 10 131 78 85 298

Stratus Computer 580 160 159 NR 2 136 15 15 133

Sun Microsystems 5902 155 154 NR 3 418 07 07 176

Tandem Computers 2285 155 150 NR 6 105 57 61 333

3ComCorp 1295 160 159 NR 12 1190 11 11 258

Average 155 151 10 456 47 53 204

-15- UVA-F-1152

Exhibit 4

TELETECH CORPORATION 1996

Debt Capital Market Conditions January 1996

Corporate Bond Yields by Rating US Treasurv Securities

Industrials AAA 650 Short-term bills 520 AA 700 Intermediate-term notes 543 A 764 Long-term bonds 604

BBB 778 BB 893 B 1049

Utilities AA 653 A 794 BBB 806

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

I

j

j

~~

c j

j

j

j

j

j

j

j

j

j

j

l j

j

j

j

-9- UVA-F-1l52

Conclusion

Weston could not realistically hope that all the issues before her would be resolved in time to influence Victor Yossarians attack on management But the attack did dictate the need for an objective assessment of the performance of Teletechs two segments-the choice of hurdle rates would be very important in this analysis However she did want to institute a pragmatic system of appropriate hurdle rates (or one rate) that would facilitate judgments in the changing circumstances Teletech faced What were the appropriate hurdle rates for the two segments Was Products and Systems underperforming as Yossarian suggested How should Teletech respond to the raider

-10- UVA-F-1152 shy

Exhibit 1

TELETECH CORPORATION 1996

Summary of W ACC Calculation for Teletech Corporation and

Segment Worksheet

Telecommunications Products and

Co]rate Services SYstems

MY Asset Weights 100 75 25

Bond Rating AA-A+ AA BBB-

Pre-tax Cost of Debt 703 700 778

Tax Rate 40 40 40

After-tax Cost of Debt 422 420 467

Equity Beta 104

Rf 604

RM-Rf 550

Cost of Equity 1177

Weight of Debt 180

Weight of Equity 820

WACC 1041 I

c

UVA-F-1152

Exhibit 2

TELETECH CORPORATION 1996

Theoretical Overview of Multiple Hurdle Rates

To Margaret Weston From Bernard Ingles Subject Theory of Segment Cost of Capital Date January 1996

You requested an overview of theories about multiple hurdle rates Without getting into minutiae the theories boil down to the following points

1 The central idea is that required returns should be driven by risk This is the dominant view in the field of investment management and is based on a mountain of theory and empirical research stretching over several decades The extension of this idea from investment management to corporate decision making is straightforward p~ least in theory

2 An underlying assumption is that the firm is transparent (Le that investors can see through the corporate veil and evaluate the activities going on inside) No one believes firms are completely transparent or that investors are perfectly informed But financial accounting standards have evolved toward making the firm more transparent And the investment community has grown tougher and sharper in its analysis Teletech now has 36 analysts publishing reports and forecasts on the firm The reality is that for big publicly held firms transparency is not a bad assumption

3 Another underlying assumption is that the value of the whole enterprise is simply the sum of its parts-this is the concept of Value Additivity We can define parts as either the business segments (on the left-hand side of the balance sheet) or the layers of the capital structure (on the right-hand side of the balance sheet) Market values (MVs) have to balance

MVTeteedl = ( MVTetlaquoomllllltlicaliolls Services + MV tmdmls+ iwellls) = ( MVdehl + MVltquiV)

Ifthese equalities did not hold then a raider could come along and exploit the inequality by buying or selling the whole and the parts This is arbitrage By buying and selling the actions of the raider would drive the MVs back into balance

-12- UVA-F-1152 r ~

Exhibit 2 (continued)

4 Investment theory tells us that the only risk that matters is nondiversifiable risk which is measured by beta Beta indicates the risk that an asset will add to a portfolio Because the investor is assumed to be diversified she is assumed to seek a return for only that risk that she cannot shed the nondiversifiabIe risk Now the important point here is that the beta ofa portfolio is equal to a weighted average of the betas of the portfolio components Extending this to the corporate environment the asset beta for the firm will equal a weighted average of the components of the firm-again the components of the firm can be defined in terms of either the right-hand side or the left-hand side of the balance sheet

Where

W = percentage weights based on market values

fJ Tel Serv fJ ps Asset betas for business segments

fJdebl fJ for the firms debt securities

fJeqUlIY= fJ of firms common stock (given by Bloomberg etc)

This is a very handy way to model the risk ofthe firm for it means that we can use the Capital Asset Pricing Model to estimate the cost of capital for a segment (ie using segment asset betas)

5 Given all the previous points it follows that the weighted average of the various costs ofcapital (K) for the firm (W ACC) which is the theoretically correct hurdle rate is simply a weighted average of segment W ACCs

Where

WTeCf WIS = market value weights

WACCTeLrv = (Wdebl TeLerv( K debl TelServ) + ( Weqllily TelServt K eltIUlly TelServ)

c

-13- UVA-F-1152

Exhibit 2 (continued)

6 The notion in point 5 may not hold exactly in practice First most of the components in the WACC formula are estimated with some error Second because oftaxes information asymmetries or other market imperfections assets may not be priced strictly in line with the model-for a company like Teletech it is reasonable to assume that any mispricings are just temporary Third the simple two-segment characterization ignores a hidden third segment the corporate treasury department that hedges and aims to finance the whole corporation optimally-this acts as a shock absorber for the financial policies of the segments Modeling the W ACC ofthe corporate treasury department is quite difficult Most companies assume that the impact of corporate treasury isnt very large and simply assume it away As a first cut we could do this too though it is an issue we should revisit

Conclusions

bull In theory the corporate WACC for Teletech is appropriate only for evaluating an asset having the same risk as the whole company It is not appropriate for assets having different risks than the whole company

bull Segment WACCs are computed similarly to corporate W ACCs

bull In concept the corporate W ACe is aweighted average of the segment W ACCs In practice the weighted-average concept may not hold due to imperfections in the market andor estimation errors

bull If we start computing segment W ACCs we must use the cost of debt cost of equity and weights appropriate to that segment We need a lot of information to do this correctly or else we really need to stretch to make assumptions

-14- UVA-F-1152_

Exhibit 3 TELETECH CORPORATION 1996

Samples of Comparable Firms

1995 Equity Asset Bond Book Val Pricel Mkt Val Mkt Val Pricel Revenues Seta Seta Iiaiiog OebtJCap Book DebtCap DebtlEq Eacniogs

Teletech Corporation $16000 1041 092 AA-tA+ 40 301 18 22 129

Telecommunications Services Industty ATampT $80000 090 085 AA 39 660 88 97 308 Alltel Corp 3160 075 063 A 49 299 243 322 160 Amerltech 13325 075 067 AA 47 472 158 188 169 BeD Atlantic 13500 080 068 AA 57 453 226 293 175 Bell South 17780 075 072 AAA 44 1190 62 66 190

Centul) Tel Enterprs 625 100 084 BBB+ 46 263 245 324 158

Cincinnati Bell 1350 080 069 AA 56 472 212 270 188 Citizens Utilities Co 1070 070 056 AA 42 168 301 431 158

Comsat 850 095 068 A 40 103 392 646 168

Frontier Corp 1750 080 074 A 42 550 116 132 283

GTE Corp 20250 080 066 BBB 69 652 254 341 169

MCI Communications 15100 125 114 A 24 187 144 169 179 NYNEX Corp 13425 075 059 Ashy 62 375 303 436 156

Pacific Telesis 9070 085 068 AAshy 74 694 291 410 136

SBC Communications 12560 090 080 A 54 559 173 210 180

Southem New England 1840 075 058 AA 5710 263 335 504 154

Sprint Corp

US West

Average

13550

9450

105

065

084

087

052

072

BBB

AAshy

52

66

51

313

467

452

257

294

228

347

416

311

150

143 C179

Telecommunications Equipment Industty ADC Telecomm Inc $586 135 135 NR 0 405 00 00 280

Aane-Cleveland 120 150 149 NR 1 151 07 07 128

ADen Group 325 160 155 NR 13 275 52 54 188

Andrew Corp 626 125 123 NR 13 440 33 34 197

DSC Communications 1450 130 126 NR 18 372 56 590 179

Newbridge NetworIs 675 155 155 NR 1 548 01 01 239

Qualcomm Inc 386 155 152 NR 9 341 28 29 583

Tellabs Inc 645 150 150 NR 1 796 01 01 266

Average 145 143 7 416 22 23 258

Computer and Network Equipment Industty Amdahl Corp $1500 13 120 NR 12 095 125 143 1~1

Bay Networks Inc 1342 175 174 NR 10 903 12 12 260

Cabletron Systems 1060 160 160 NR 0 657 00 00 218

Cisco Systems 1979 175 175 NR 0 1383 00 00 269

Digital Equipment 13813 110 104 NR 22 287 89 98 171

General Datacomm 221 185 179 NR 17 396 49 52 NMF

Hewlett-Packard 31519 125 124 NR 6 333 19 19 146

SCI Systems 2673 120 106 NR 32 220 176 214 127

Sequent Computer 535 195 192 NR 3 124 24 25 120

Standard Microsystems 345 160 152 NR 10 131 78 85 298

Stratus Computer 580 160 159 NR 2 136 15 15 133

Sun Microsystems 5902 155 154 NR 3 418 07 07 176

Tandem Computers 2285 155 150 NR 6 105 57 61 333

3ComCorp 1295 160 159 NR 12 1190 11 11 258

Average 155 151 10 456 47 53 204

-15- UVA-F-1152

Exhibit 4

TELETECH CORPORATION 1996

Debt Capital Market Conditions January 1996

Corporate Bond Yields by Rating US Treasurv Securities

Industrials AAA 650 Short-term bills 520 AA 700 Intermediate-term notes 543 A 764 Long-term bonds 604

BBB 778 BB 893 B 1049

Utilities AA 653 A 794 BBB 806

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

I

j

j

~~

c j

j

j

j

j

j

j

j

j

j

j

l j

j

j

j

-10- UVA-F-1152 shy

Exhibit 1

TELETECH CORPORATION 1996

Summary of W ACC Calculation for Teletech Corporation and

Segment Worksheet

Telecommunications Products and

Co]rate Services SYstems

MY Asset Weights 100 75 25

Bond Rating AA-A+ AA BBB-

Pre-tax Cost of Debt 703 700 778

Tax Rate 40 40 40

After-tax Cost of Debt 422 420 467

Equity Beta 104

Rf 604

RM-Rf 550

Cost of Equity 1177

Weight of Debt 180

Weight of Equity 820

WACC 1041 I

c

UVA-F-1152

Exhibit 2

TELETECH CORPORATION 1996

Theoretical Overview of Multiple Hurdle Rates

To Margaret Weston From Bernard Ingles Subject Theory of Segment Cost of Capital Date January 1996

You requested an overview of theories about multiple hurdle rates Without getting into minutiae the theories boil down to the following points

1 The central idea is that required returns should be driven by risk This is the dominant view in the field of investment management and is based on a mountain of theory and empirical research stretching over several decades The extension of this idea from investment management to corporate decision making is straightforward p~ least in theory

2 An underlying assumption is that the firm is transparent (Le that investors can see through the corporate veil and evaluate the activities going on inside) No one believes firms are completely transparent or that investors are perfectly informed But financial accounting standards have evolved toward making the firm more transparent And the investment community has grown tougher and sharper in its analysis Teletech now has 36 analysts publishing reports and forecasts on the firm The reality is that for big publicly held firms transparency is not a bad assumption

3 Another underlying assumption is that the value of the whole enterprise is simply the sum of its parts-this is the concept of Value Additivity We can define parts as either the business segments (on the left-hand side of the balance sheet) or the layers of the capital structure (on the right-hand side of the balance sheet) Market values (MVs) have to balance

MVTeteedl = ( MVTetlaquoomllllltlicaliolls Services + MV tmdmls+ iwellls) = ( MVdehl + MVltquiV)

Ifthese equalities did not hold then a raider could come along and exploit the inequality by buying or selling the whole and the parts This is arbitrage By buying and selling the actions of the raider would drive the MVs back into balance

-12- UVA-F-1152 r ~

Exhibit 2 (continued)

4 Investment theory tells us that the only risk that matters is nondiversifiable risk which is measured by beta Beta indicates the risk that an asset will add to a portfolio Because the investor is assumed to be diversified she is assumed to seek a return for only that risk that she cannot shed the nondiversifiabIe risk Now the important point here is that the beta ofa portfolio is equal to a weighted average of the betas of the portfolio components Extending this to the corporate environment the asset beta for the firm will equal a weighted average of the components of the firm-again the components of the firm can be defined in terms of either the right-hand side or the left-hand side of the balance sheet

Where

W = percentage weights based on market values

fJ Tel Serv fJ ps Asset betas for business segments

fJdebl fJ for the firms debt securities

fJeqUlIY= fJ of firms common stock (given by Bloomberg etc)

This is a very handy way to model the risk ofthe firm for it means that we can use the Capital Asset Pricing Model to estimate the cost of capital for a segment (ie using segment asset betas)

5 Given all the previous points it follows that the weighted average of the various costs ofcapital (K) for the firm (W ACC) which is the theoretically correct hurdle rate is simply a weighted average of segment W ACCs

Where

WTeCf WIS = market value weights

WACCTeLrv = (Wdebl TeLerv( K debl TelServ) + ( Weqllily TelServt K eltIUlly TelServ)

c

-13- UVA-F-1152

Exhibit 2 (continued)

6 The notion in point 5 may not hold exactly in practice First most of the components in the WACC formula are estimated with some error Second because oftaxes information asymmetries or other market imperfections assets may not be priced strictly in line with the model-for a company like Teletech it is reasonable to assume that any mispricings are just temporary Third the simple two-segment characterization ignores a hidden third segment the corporate treasury department that hedges and aims to finance the whole corporation optimally-this acts as a shock absorber for the financial policies of the segments Modeling the W ACC ofthe corporate treasury department is quite difficult Most companies assume that the impact of corporate treasury isnt very large and simply assume it away As a first cut we could do this too though it is an issue we should revisit

Conclusions

bull In theory the corporate WACC for Teletech is appropriate only for evaluating an asset having the same risk as the whole company It is not appropriate for assets having different risks than the whole company

bull Segment WACCs are computed similarly to corporate W ACCs

bull In concept the corporate W ACe is aweighted average of the segment W ACCs In practice the weighted-average concept may not hold due to imperfections in the market andor estimation errors

bull If we start computing segment W ACCs we must use the cost of debt cost of equity and weights appropriate to that segment We need a lot of information to do this correctly or else we really need to stretch to make assumptions

-14- UVA-F-1152_

Exhibit 3 TELETECH CORPORATION 1996

Samples of Comparable Firms

1995 Equity Asset Bond Book Val Pricel Mkt Val Mkt Val Pricel Revenues Seta Seta Iiaiiog OebtJCap Book DebtCap DebtlEq Eacniogs

Teletech Corporation $16000 1041 092 AA-tA+ 40 301 18 22 129

Telecommunications Services Industty ATampT $80000 090 085 AA 39 660 88 97 308 Alltel Corp 3160 075 063 A 49 299 243 322 160 Amerltech 13325 075 067 AA 47 472 158 188 169 BeD Atlantic 13500 080 068 AA 57 453 226 293 175 Bell South 17780 075 072 AAA 44 1190 62 66 190

Centul) Tel Enterprs 625 100 084 BBB+ 46 263 245 324 158

Cincinnati Bell 1350 080 069 AA 56 472 212 270 188 Citizens Utilities Co 1070 070 056 AA 42 168 301 431 158

Comsat 850 095 068 A 40 103 392 646 168

Frontier Corp 1750 080 074 A 42 550 116 132 283

GTE Corp 20250 080 066 BBB 69 652 254 341 169

MCI Communications 15100 125 114 A 24 187 144 169 179 NYNEX Corp 13425 075 059 Ashy 62 375 303 436 156

Pacific Telesis 9070 085 068 AAshy 74 694 291 410 136

SBC Communications 12560 090 080 A 54 559 173 210 180

Southem New England 1840 075 058 AA 5710 263 335 504 154

Sprint Corp

US West

Average

13550

9450

105

065

084

087

052

072

BBB

AAshy

52

66

51

313

467

452

257

294

228

347

416

311

150

143 C179

Telecommunications Equipment Industty ADC Telecomm Inc $586 135 135 NR 0 405 00 00 280

Aane-Cleveland 120 150 149 NR 1 151 07 07 128

ADen Group 325 160 155 NR 13 275 52 54 188

Andrew Corp 626 125 123 NR 13 440 33 34 197

DSC Communications 1450 130 126 NR 18 372 56 590 179

Newbridge NetworIs 675 155 155 NR 1 548 01 01 239

Qualcomm Inc 386 155 152 NR 9 341 28 29 583

Tellabs Inc 645 150 150 NR 1 796 01 01 266

Average 145 143 7 416 22 23 258

Computer and Network Equipment Industty Amdahl Corp $1500 13 120 NR 12 095 125 143 1~1

Bay Networks Inc 1342 175 174 NR 10 903 12 12 260

Cabletron Systems 1060 160 160 NR 0 657 00 00 218

Cisco Systems 1979 175 175 NR 0 1383 00 00 269

Digital Equipment 13813 110 104 NR 22 287 89 98 171

General Datacomm 221 185 179 NR 17 396 49 52 NMF

Hewlett-Packard 31519 125 124 NR 6 333 19 19 146

SCI Systems 2673 120 106 NR 32 220 176 214 127

Sequent Computer 535 195 192 NR 3 124 24 25 120

Standard Microsystems 345 160 152 NR 10 131 78 85 298

Stratus Computer 580 160 159 NR 2 136 15 15 133

Sun Microsystems 5902 155 154 NR 3 418 07 07 176

Tandem Computers 2285 155 150 NR 6 105 57 61 333

3ComCorp 1295 160 159 NR 12 1190 11 11 258

Average 155 151 10 456 47 53 204

-15- UVA-F-1152

Exhibit 4

TELETECH CORPORATION 1996

Debt Capital Market Conditions January 1996

Corporate Bond Yields by Rating US Treasurv Securities

Industrials AAA 650 Short-term bills 520 AA 700 Intermediate-term notes 543 A 764 Long-term bonds 604

BBB 778 BB 893 B 1049

Utilities AA 653 A 794 BBB 806

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

I

j

j

~~

c j

j

j

j

j

j

j

j

j

j

j

l j

j

j

j

UVA-F-1152

Exhibit 2

TELETECH CORPORATION 1996

Theoretical Overview of Multiple Hurdle Rates

To Margaret Weston From Bernard Ingles Subject Theory of Segment Cost of Capital Date January 1996

You requested an overview of theories about multiple hurdle rates Without getting into minutiae the theories boil down to the following points

1 The central idea is that required returns should be driven by risk This is the dominant view in the field of investment management and is based on a mountain of theory and empirical research stretching over several decades The extension of this idea from investment management to corporate decision making is straightforward p~ least in theory

2 An underlying assumption is that the firm is transparent (Le that investors can see through the corporate veil and evaluate the activities going on inside) No one believes firms are completely transparent or that investors are perfectly informed But financial accounting standards have evolved toward making the firm more transparent And the investment community has grown tougher and sharper in its analysis Teletech now has 36 analysts publishing reports and forecasts on the firm The reality is that for big publicly held firms transparency is not a bad assumption

3 Another underlying assumption is that the value of the whole enterprise is simply the sum of its parts-this is the concept of Value Additivity We can define parts as either the business segments (on the left-hand side of the balance sheet) or the layers of the capital structure (on the right-hand side of the balance sheet) Market values (MVs) have to balance

MVTeteedl = ( MVTetlaquoomllllltlicaliolls Services + MV tmdmls+ iwellls) = ( MVdehl + MVltquiV)

Ifthese equalities did not hold then a raider could come along and exploit the inequality by buying or selling the whole and the parts This is arbitrage By buying and selling the actions of the raider would drive the MVs back into balance

-12- UVA-F-1152 r ~

Exhibit 2 (continued)

4 Investment theory tells us that the only risk that matters is nondiversifiable risk which is measured by beta Beta indicates the risk that an asset will add to a portfolio Because the investor is assumed to be diversified she is assumed to seek a return for only that risk that she cannot shed the nondiversifiabIe risk Now the important point here is that the beta ofa portfolio is equal to a weighted average of the betas of the portfolio components Extending this to the corporate environment the asset beta for the firm will equal a weighted average of the components of the firm-again the components of the firm can be defined in terms of either the right-hand side or the left-hand side of the balance sheet

Where

W = percentage weights based on market values

fJ Tel Serv fJ ps Asset betas for business segments

fJdebl fJ for the firms debt securities

fJeqUlIY= fJ of firms common stock (given by Bloomberg etc)

This is a very handy way to model the risk ofthe firm for it means that we can use the Capital Asset Pricing Model to estimate the cost of capital for a segment (ie using segment asset betas)

5 Given all the previous points it follows that the weighted average of the various costs ofcapital (K) for the firm (W ACC) which is the theoretically correct hurdle rate is simply a weighted average of segment W ACCs

Where

WTeCf WIS = market value weights

WACCTeLrv = (Wdebl TeLerv( K debl TelServ) + ( Weqllily TelServt K eltIUlly TelServ)

c

-13- UVA-F-1152

Exhibit 2 (continued)

6 The notion in point 5 may not hold exactly in practice First most of the components in the WACC formula are estimated with some error Second because oftaxes information asymmetries or other market imperfections assets may not be priced strictly in line with the model-for a company like Teletech it is reasonable to assume that any mispricings are just temporary Third the simple two-segment characterization ignores a hidden third segment the corporate treasury department that hedges and aims to finance the whole corporation optimally-this acts as a shock absorber for the financial policies of the segments Modeling the W ACC ofthe corporate treasury department is quite difficult Most companies assume that the impact of corporate treasury isnt very large and simply assume it away As a first cut we could do this too though it is an issue we should revisit

Conclusions

bull In theory the corporate WACC for Teletech is appropriate only for evaluating an asset having the same risk as the whole company It is not appropriate for assets having different risks than the whole company

bull Segment WACCs are computed similarly to corporate W ACCs

bull In concept the corporate W ACe is aweighted average of the segment W ACCs In practice the weighted-average concept may not hold due to imperfections in the market andor estimation errors

bull If we start computing segment W ACCs we must use the cost of debt cost of equity and weights appropriate to that segment We need a lot of information to do this correctly or else we really need to stretch to make assumptions

-14- UVA-F-1152_

Exhibit 3 TELETECH CORPORATION 1996

Samples of Comparable Firms

1995 Equity Asset Bond Book Val Pricel Mkt Val Mkt Val Pricel Revenues Seta Seta Iiaiiog OebtJCap Book DebtCap DebtlEq Eacniogs

Teletech Corporation $16000 1041 092 AA-tA+ 40 301 18 22 129

Telecommunications Services Industty ATampT $80000 090 085 AA 39 660 88 97 308 Alltel Corp 3160 075 063 A 49 299 243 322 160 Amerltech 13325 075 067 AA 47 472 158 188 169 BeD Atlantic 13500 080 068 AA 57 453 226 293 175 Bell South 17780 075 072 AAA 44 1190 62 66 190

Centul) Tel Enterprs 625 100 084 BBB+ 46 263 245 324 158

Cincinnati Bell 1350 080 069 AA 56 472 212 270 188 Citizens Utilities Co 1070 070 056 AA 42 168 301 431 158

Comsat 850 095 068 A 40 103 392 646 168

Frontier Corp 1750 080 074 A 42 550 116 132 283

GTE Corp 20250 080 066 BBB 69 652 254 341 169

MCI Communications 15100 125 114 A 24 187 144 169 179 NYNEX Corp 13425 075 059 Ashy 62 375 303 436 156

Pacific Telesis 9070 085 068 AAshy 74 694 291 410 136

SBC Communications 12560 090 080 A 54 559 173 210 180

Southem New England 1840 075 058 AA 5710 263 335 504 154

Sprint Corp

US West

Average

13550

9450

105

065

084

087

052

072

BBB

AAshy

52

66

51

313

467

452

257

294

228

347

416

311

150

143 C179

Telecommunications Equipment Industty ADC Telecomm Inc $586 135 135 NR 0 405 00 00 280

Aane-Cleveland 120 150 149 NR 1 151 07 07 128

ADen Group 325 160 155 NR 13 275 52 54 188

Andrew Corp 626 125 123 NR 13 440 33 34 197

DSC Communications 1450 130 126 NR 18 372 56 590 179

Newbridge NetworIs 675 155 155 NR 1 548 01 01 239

Qualcomm Inc 386 155 152 NR 9 341 28 29 583

Tellabs Inc 645 150 150 NR 1 796 01 01 266

Average 145 143 7 416 22 23 258

Computer and Network Equipment Industty Amdahl Corp $1500 13 120 NR 12 095 125 143 1~1

Bay Networks Inc 1342 175 174 NR 10 903 12 12 260

Cabletron Systems 1060 160 160 NR 0 657 00 00 218

Cisco Systems 1979 175 175 NR 0 1383 00 00 269

Digital Equipment 13813 110 104 NR 22 287 89 98 171

General Datacomm 221 185 179 NR 17 396 49 52 NMF

Hewlett-Packard 31519 125 124 NR 6 333 19 19 146

SCI Systems 2673 120 106 NR 32 220 176 214 127

Sequent Computer 535 195 192 NR 3 124 24 25 120

Standard Microsystems 345 160 152 NR 10 131 78 85 298

Stratus Computer 580 160 159 NR 2 136 15 15 133

Sun Microsystems 5902 155 154 NR 3 418 07 07 176

Tandem Computers 2285 155 150 NR 6 105 57 61 333

3ComCorp 1295 160 159 NR 12 1190 11 11 258

Average 155 151 10 456 47 53 204

-15- UVA-F-1152

Exhibit 4

TELETECH CORPORATION 1996

Debt Capital Market Conditions January 1996

Corporate Bond Yields by Rating US Treasurv Securities

Industrials AAA 650 Short-term bills 520 AA 700 Intermediate-term notes 543 A 764 Long-term bonds 604

BBB 778 BB 893 B 1049

Utilities AA 653 A 794 BBB 806

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

I

j

j

~~

c j

j

j

j

j

j

j

j

j

j

j

l j

j

j

j

-12- UVA-F-1152 r ~

Exhibit 2 (continued)

4 Investment theory tells us that the only risk that matters is nondiversifiable risk which is measured by beta Beta indicates the risk that an asset will add to a portfolio Because the investor is assumed to be diversified she is assumed to seek a return for only that risk that she cannot shed the nondiversifiabIe risk Now the important point here is that the beta ofa portfolio is equal to a weighted average of the betas of the portfolio components Extending this to the corporate environment the asset beta for the firm will equal a weighted average of the components of the firm-again the components of the firm can be defined in terms of either the right-hand side or the left-hand side of the balance sheet

Where

W = percentage weights based on market values

fJ Tel Serv fJ ps Asset betas for business segments

fJdebl fJ for the firms debt securities

fJeqUlIY= fJ of firms common stock (given by Bloomberg etc)

This is a very handy way to model the risk ofthe firm for it means that we can use the Capital Asset Pricing Model to estimate the cost of capital for a segment (ie using segment asset betas)

5 Given all the previous points it follows that the weighted average of the various costs ofcapital (K) for the firm (W ACC) which is the theoretically correct hurdle rate is simply a weighted average of segment W ACCs

Where

WTeCf WIS = market value weights

WACCTeLrv = (Wdebl TeLerv( K debl TelServ) + ( Weqllily TelServt K eltIUlly TelServ)

c

-13- UVA-F-1152

Exhibit 2 (continued)

6 The notion in point 5 may not hold exactly in practice First most of the components in the WACC formula are estimated with some error Second because oftaxes information asymmetries or other market imperfections assets may not be priced strictly in line with the model-for a company like Teletech it is reasonable to assume that any mispricings are just temporary Third the simple two-segment characterization ignores a hidden third segment the corporate treasury department that hedges and aims to finance the whole corporation optimally-this acts as a shock absorber for the financial policies of the segments Modeling the W ACC ofthe corporate treasury department is quite difficult Most companies assume that the impact of corporate treasury isnt very large and simply assume it away As a first cut we could do this too though it is an issue we should revisit

Conclusions

bull In theory the corporate WACC for Teletech is appropriate only for evaluating an asset having the same risk as the whole company It is not appropriate for assets having different risks than the whole company

bull Segment WACCs are computed similarly to corporate W ACCs

bull In concept the corporate W ACe is aweighted average of the segment W ACCs In practice the weighted-average concept may not hold due to imperfections in the market andor estimation errors

bull If we start computing segment W ACCs we must use the cost of debt cost of equity and weights appropriate to that segment We need a lot of information to do this correctly or else we really need to stretch to make assumptions

-14- UVA-F-1152_

Exhibit 3 TELETECH CORPORATION 1996

Samples of Comparable Firms

1995 Equity Asset Bond Book Val Pricel Mkt Val Mkt Val Pricel Revenues Seta Seta Iiaiiog OebtJCap Book DebtCap DebtlEq Eacniogs

Teletech Corporation $16000 1041 092 AA-tA+ 40 301 18 22 129

Telecommunications Services Industty ATampT $80000 090 085 AA 39 660 88 97 308 Alltel Corp 3160 075 063 A 49 299 243 322 160 Amerltech 13325 075 067 AA 47 472 158 188 169 BeD Atlantic 13500 080 068 AA 57 453 226 293 175 Bell South 17780 075 072 AAA 44 1190 62 66 190

Centul) Tel Enterprs 625 100 084 BBB+ 46 263 245 324 158

Cincinnati Bell 1350 080 069 AA 56 472 212 270 188 Citizens Utilities Co 1070 070 056 AA 42 168 301 431 158

Comsat 850 095 068 A 40 103 392 646 168

Frontier Corp 1750 080 074 A 42 550 116 132 283

GTE Corp 20250 080 066 BBB 69 652 254 341 169

MCI Communications 15100 125 114 A 24 187 144 169 179 NYNEX Corp 13425 075 059 Ashy 62 375 303 436 156

Pacific Telesis 9070 085 068 AAshy 74 694 291 410 136

SBC Communications 12560 090 080 A 54 559 173 210 180

Southem New England 1840 075 058 AA 5710 263 335 504 154

Sprint Corp

US West

Average

13550

9450

105

065

084

087

052

072

BBB

AAshy

52

66

51

313

467

452

257

294

228

347

416

311

150

143 C179

Telecommunications Equipment Industty ADC Telecomm Inc $586 135 135 NR 0 405 00 00 280

Aane-Cleveland 120 150 149 NR 1 151 07 07 128

ADen Group 325 160 155 NR 13 275 52 54 188

Andrew Corp 626 125 123 NR 13 440 33 34 197

DSC Communications 1450 130 126 NR 18 372 56 590 179

Newbridge NetworIs 675 155 155 NR 1 548 01 01 239

Qualcomm Inc 386 155 152 NR 9 341 28 29 583

Tellabs Inc 645 150 150 NR 1 796 01 01 266

Average 145 143 7 416 22 23 258

Computer and Network Equipment Industty Amdahl Corp $1500 13 120 NR 12 095 125 143 1~1

Bay Networks Inc 1342 175 174 NR 10 903 12 12 260

Cabletron Systems 1060 160 160 NR 0 657 00 00 218

Cisco Systems 1979 175 175 NR 0 1383 00 00 269

Digital Equipment 13813 110 104 NR 22 287 89 98 171

General Datacomm 221 185 179 NR 17 396 49 52 NMF

Hewlett-Packard 31519 125 124 NR 6 333 19 19 146

SCI Systems 2673 120 106 NR 32 220 176 214 127

Sequent Computer 535 195 192 NR 3 124 24 25 120

Standard Microsystems 345 160 152 NR 10 131 78 85 298

Stratus Computer 580 160 159 NR 2 136 15 15 133

Sun Microsystems 5902 155 154 NR 3 418 07 07 176

Tandem Computers 2285 155 150 NR 6 105 57 61 333

3ComCorp 1295 160 159 NR 12 1190 11 11 258

Average 155 151 10 456 47 53 204

-15- UVA-F-1152

Exhibit 4

TELETECH CORPORATION 1996

Debt Capital Market Conditions January 1996

Corporate Bond Yields by Rating US Treasurv Securities

Industrials AAA 650 Short-term bills 520 AA 700 Intermediate-term notes 543 A 764 Long-term bonds 604

BBB 778 BB 893 B 1049

Utilities AA 653 A 794 BBB 806

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

I

j

j

~~

c j

j

j

j

j

j

j

j

j

j

j

l j

j

j

j

-13- UVA-F-1152

Exhibit 2 (continued)

6 The notion in point 5 may not hold exactly in practice First most of the components in the WACC formula are estimated with some error Second because oftaxes information asymmetries or other market imperfections assets may not be priced strictly in line with the model-for a company like Teletech it is reasonable to assume that any mispricings are just temporary Third the simple two-segment characterization ignores a hidden third segment the corporate treasury department that hedges and aims to finance the whole corporation optimally-this acts as a shock absorber for the financial policies of the segments Modeling the W ACC ofthe corporate treasury department is quite difficult Most companies assume that the impact of corporate treasury isnt very large and simply assume it away As a first cut we could do this too though it is an issue we should revisit

Conclusions

bull In theory the corporate WACC for Teletech is appropriate only for evaluating an asset having the same risk as the whole company It is not appropriate for assets having different risks than the whole company

bull Segment WACCs are computed similarly to corporate W ACCs

bull In concept the corporate W ACe is aweighted average of the segment W ACCs In practice the weighted-average concept may not hold due to imperfections in the market andor estimation errors

bull If we start computing segment W ACCs we must use the cost of debt cost of equity and weights appropriate to that segment We need a lot of information to do this correctly or else we really need to stretch to make assumptions

-14- UVA-F-1152_

Exhibit 3 TELETECH CORPORATION 1996

Samples of Comparable Firms

1995 Equity Asset Bond Book Val Pricel Mkt Val Mkt Val Pricel Revenues Seta Seta Iiaiiog OebtJCap Book DebtCap DebtlEq Eacniogs

Teletech Corporation $16000 1041 092 AA-tA+ 40 301 18 22 129

Telecommunications Services Industty ATampT $80000 090 085 AA 39 660 88 97 308 Alltel Corp 3160 075 063 A 49 299 243 322 160 Amerltech 13325 075 067 AA 47 472 158 188 169 BeD Atlantic 13500 080 068 AA 57 453 226 293 175 Bell South 17780 075 072 AAA 44 1190 62 66 190

Centul) Tel Enterprs 625 100 084 BBB+ 46 263 245 324 158

Cincinnati Bell 1350 080 069 AA 56 472 212 270 188 Citizens Utilities Co 1070 070 056 AA 42 168 301 431 158

Comsat 850 095 068 A 40 103 392 646 168

Frontier Corp 1750 080 074 A 42 550 116 132 283

GTE Corp 20250 080 066 BBB 69 652 254 341 169

MCI Communications 15100 125 114 A 24 187 144 169 179 NYNEX Corp 13425 075 059 Ashy 62 375 303 436 156

Pacific Telesis 9070 085 068 AAshy 74 694 291 410 136

SBC Communications 12560 090 080 A 54 559 173 210 180

Southem New England 1840 075 058 AA 5710 263 335 504 154

Sprint Corp

US West

Average

13550

9450

105

065

084

087

052

072

BBB

AAshy

52

66

51

313

467

452

257

294

228

347

416

311

150

143 C179

Telecommunications Equipment Industty ADC Telecomm Inc $586 135 135 NR 0 405 00 00 280

Aane-Cleveland 120 150 149 NR 1 151 07 07 128

ADen Group 325 160 155 NR 13 275 52 54 188

Andrew Corp 626 125 123 NR 13 440 33 34 197

DSC Communications 1450 130 126 NR 18 372 56 590 179

Newbridge NetworIs 675 155 155 NR 1 548 01 01 239

Qualcomm Inc 386 155 152 NR 9 341 28 29 583

Tellabs Inc 645 150 150 NR 1 796 01 01 266

Average 145 143 7 416 22 23 258

Computer and Network Equipment Industty Amdahl Corp $1500 13 120 NR 12 095 125 143 1~1

Bay Networks Inc 1342 175 174 NR 10 903 12 12 260

Cabletron Systems 1060 160 160 NR 0 657 00 00 218

Cisco Systems 1979 175 175 NR 0 1383 00 00 269

Digital Equipment 13813 110 104 NR 22 287 89 98 171

General Datacomm 221 185 179 NR 17 396 49 52 NMF

Hewlett-Packard 31519 125 124 NR 6 333 19 19 146

SCI Systems 2673 120 106 NR 32 220 176 214 127

Sequent Computer 535 195 192 NR 3 124 24 25 120

Standard Microsystems 345 160 152 NR 10 131 78 85 298

Stratus Computer 580 160 159 NR 2 136 15 15 133

Sun Microsystems 5902 155 154 NR 3 418 07 07 176

Tandem Computers 2285 155 150 NR 6 105 57 61 333

3ComCorp 1295 160 159 NR 12 1190 11 11 258

Average 155 151 10 456 47 53 204

-15- UVA-F-1152

Exhibit 4

TELETECH CORPORATION 1996

Debt Capital Market Conditions January 1996

Corporate Bond Yields by Rating US Treasurv Securities

Industrials AAA 650 Short-term bills 520 AA 700 Intermediate-term notes 543 A 764 Long-term bonds 604

BBB 778 BB 893 B 1049

Utilities AA 653 A 794 BBB 806

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

I

j

j

~~

c j

j

j

j

j

j

j

j

j

j

j

l j

j

j

j

-14- UVA-F-1152_

Exhibit 3 TELETECH CORPORATION 1996

Samples of Comparable Firms

1995 Equity Asset Bond Book Val Pricel Mkt Val Mkt Val Pricel Revenues Seta Seta Iiaiiog OebtJCap Book DebtCap DebtlEq Eacniogs

Teletech Corporation $16000 1041 092 AA-tA+ 40 301 18 22 129

Telecommunications Services Industty ATampT $80000 090 085 AA 39 660 88 97 308 Alltel Corp 3160 075 063 A 49 299 243 322 160 Amerltech 13325 075 067 AA 47 472 158 188 169 BeD Atlantic 13500 080 068 AA 57 453 226 293 175 Bell South 17780 075 072 AAA 44 1190 62 66 190

Centul) Tel Enterprs 625 100 084 BBB+ 46 263 245 324 158

Cincinnati Bell 1350 080 069 AA 56 472 212 270 188 Citizens Utilities Co 1070 070 056 AA 42 168 301 431 158

Comsat 850 095 068 A 40 103 392 646 168

Frontier Corp 1750 080 074 A 42 550 116 132 283

GTE Corp 20250 080 066 BBB 69 652 254 341 169

MCI Communications 15100 125 114 A 24 187 144 169 179 NYNEX Corp 13425 075 059 Ashy 62 375 303 436 156

Pacific Telesis 9070 085 068 AAshy 74 694 291 410 136

SBC Communications 12560 090 080 A 54 559 173 210 180

Southem New England 1840 075 058 AA 5710 263 335 504 154

Sprint Corp

US West

Average

13550

9450

105

065

084

087

052

072

BBB

AAshy

52

66

51

313

467

452

257

294

228

347

416

311

150

143 C179

Telecommunications Equipment Industty ADC Telecomm Inc $586 135 135 NR 0 405 00 00 280

Aane-Cleveland 120 150 149 NR 1 151 07 07 128

ADen Group 325 160 155 NR 13 275 52 54 188

Andrew Corp 626 125 123 NR 13 440 33 34 197

DSC Communications 1450 130 126 NR 18 372 56 590 179

Newbridge NetworIs 675 155 155 NR 1 548 01 01 239

Qualcomm Inc 386 155 152 NR 9 341 28 29 583

Tellabs Inc 645 150 150 NR 1 796 01 01 266

Average 145 143 7 416 22 23 258

Computer and Network Equipment Industty Amdahl Corp $1500 13 120 NR 12 095 125 143 1~1

Bay Networks Inc 1342 175 174 NR 10 903 12 12 260

Cabletron Systems 1060 160 160 NR 0 657 00 00 218

Cisco Systems 1979 175 175 NR 0 1383 00 00 269

Digital Equipment 13813 110 104 NR 22 287 89 98 171

General Datacomm 221 185 179 NR 17 396 49 52 NMF

Hewlett-Packard 31519 125 124 NR 6 333 19 19 146

SCI Systems 2673 120 106 NR 32 220 176 214 127

Sequent Computer 535 195 192 NR 3 124 24 25 120

Standard Microsystems 345 160 152 NR 10 131 78 85 298

Stratus Computer 580 160 159 NR 2 136 15 15 133

Sun Microsystems 5902 155 154 NR 3 418 07 07 176

Tandem Computers 2285 155 150 NR 6 105 57 61 333

3ComCorp 1295 160 159 NR 12 1190 11 11 258

Average 155 151 10 456 47 53 204

-15- UVA-F-1152

Exhibit 4

TELETECH CORPORATION 1996

Debt Capital Market Conditions January 1996

Corporate Bond Yields by Rating US Treasurv Securities

Industrials AAA 650 Short-term bills 520 AA 700 Intermediate-term notes 543 A 764 Long-term bonds 604

BBB 778 BB 893 B 1049

Utilities AA 653 A 794 BBB 806

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

I

j

j

~~

c j

j

j

j

j

j

j

j

j

j

j

l j

j

j

j

-15- UVA-F-1152

Exhibit 4

TELETECH CORPORATION 1996

Debt Capital Market Conditions January 1996

Corporate Bond Yields by Rating US Treasurv Securities

Industrials AAA 650 Short-term bills 520 AA 700 Intermediate-term notes 543 A 764 Long-term bonds 604

BBB 778 BB 893 B 1049

Utilities AA 653 A 794 BBB 806

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

I

j

j

~~

c j

j

j

j

j

j

j

j

j

j

j

l j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

j

I

j

j

~~

c j

j

j

j

j

j

j

j

j

j

j

l j

j

j

j