andrew lo

Can financial engineering cure Can financial engineering cure cancer?cancer?A new approach to funding large scale biomedical innovation

Andrew W. LoAndrew W. Lo(joint work with Jose‐Maria Fernandez and Roger M. Stein)

Sloan Reunion 2012Sloan Reunion 2012June 8, 2012

The contents of this presentation are based on the research carried out at the MIT Laboratory for Financial Engineering by Jose‐Maria Fernandez, Roger M. Stein and Andrew W. Lo. For total or partial reproduction please contact jose‐, g p p p [email protected]

2012 by Jose‐Maria Fernandez, Roger M. Stein, Andrew W. LoAll Rights Reserved

Summary Biomedical innovation has three important characteristics:

– Expensive, risky, and lengthyExpensive, risky, and lengthy

Traditional funding vehicles are not ideal (private/public equity)

Ri k i f th i i ( l it FDA t t liff) Risk is one of the main issues (complexity, FDA, patent cliff)

Diversification is the solution: large pooled vehicles lower risk

With sufficient diversification, long‐term debt is possible– More patient capital, but still imposes discipline (interest payments)– Much larger pool of capital, allowing greater participation– The financial technology and institutions already exist

The nat ral scale req ired is $ to $30 billion The natural scale required is $5 to $30 billion

Using financial engineering, this is achievable today

© 2012 by Fernandez, Stein, and Lo All Rights ReservedMIT LFE Slide 2

Lessons from the financial crisis?

Pension

Commercial BanksMutualF d

PensionFunds

Ra

Age

$$$$$$

Investment Banks

Mortgage Companies

Fannie Mae

Funds

SovereignWealth Funds

atingencies

InC $$$$$$Fannie Mae

Freddie Mac

PoliticiansFoundations,Endowments

nsurance &D

S Market $$$$$$

HedgeFunds

© 2012 by Fernandez, Stein, and Lo All Rights Reserved Slide 3MIT LFE


Real Home Price Index (1890-2006)200

160

180

Rat

e

120

140

or In

tere

st R

80

100

120

Inde

x o

60

80

1880 1900 1920 1940 1960 1980 2000 2020

Source: Robert J. Shiller, Irrational Exuberance, 2nd. Edition.

Year To get back to long-run average, we need 45% real drop( 35% i l d )


(= 35% nominal drop)

Consider simple securitization example

I.O.U.$1,000 $1,000

90%Price = 90% × $1,000 + 10% × $0

= $900

$0 (Default)10%

I.O.U.$1,000 $1,000

90%Price = 90% × $1,000 + 10% × $0

= $900

$0 (Default)10%

© 2010 by Andrew W. Lo All Rights Reserved9/21/2010 Slide 5


I.O.U.$1,000

MBS$1,000

Senior TranchePortfolio

I.O.U.$1,000

MBS$1,000

Junior Tranche



Assuming Independent

I.O.U.$1,000

Portfolio

Assuming IndependentDefaults

Value Prob.

$2,000 81%Portfolio $1,000 18%

$0 1%

I.O.U.$1,000

$0 1%



Assuming Independent Defaults

MBS$1,000 Portfolio

Value Prob. Senior Tranche

Junior Tranche

Senior Tranche

$2,000 81% $1,000 $1,000

$1,000 18% $1,000 $0$ , $ , $

$0 1% $0 $0

MBS$1,000 Bad State

For SeniorTranche (1%)

Bad StateFor Junior

Junior Tranche

Tranche (1%) For JuniorTranche (19%)




MBS$1,000 Portfolio


Junior Tranche

Senior Tranche

$2,000 81% $1,000 $1,000

$1,000 18% $1,000 $0

$0 1% $0 $0

MBS$1,000 Price for Senior Tranche = 99% × $1,000 + 1% × $0

= $990

Price for Junior Tranche = 81% × $1 000 + 19% × $0

Junior Tranche

Price for Junior Tranche = 81% × $1,000 + 19% × $0= $810



Non‐Investment Grade

“Similar” to Junior

Tranche?

“Similar” to S i

Tranche?

Senior Tranche?

© 2012 by Fernandez, Stein, and Lo All Rights Reserved Slide 10MIT LFE Source: Moody’s.


U.S. Bond Market Debt Issuance ($Billions)

Corporate Federal AgencyMunicipal Treasury1 Mortgage‐Related2 Debt3 Securities Asset‐Backed Total

1996 185.2 612.4 479.7 343.7 277.9 168.4 2,067.21997 220.7 540.0 577.6 466.0 323.1 223.1 2,350.51998 286.8 438.4 1,118.1 610.7 596.4 286.6 3,336.91999 227.5 364.6 985.4 629.2 548.0 287.1 3,041.82000 200.8 312.4 660.0 587.5 446.6 281.5 2,488.82001 287.7 380.7 1,663.9 776.1 941.0 326.2 4,375.62002 357.5 571.6 2,283.0 636.7 1,041.5 373.9 5,264.22002 357.5 571.6 2,283.0 636.7 1,041.5 373.9 5,264.22003 382.7 745.2 3,084.3 775.8 1,267.5 461.5 6,717.02004 359.8 853.3 1,879.0 780.7 881.8(4) 651.5 4,524.32005 408.2 746.2 2,182.4 752.8 669.0 753.5 5,512.12006 386.5 788.5 2,088.8 1,058.9 747.3 753.9 5,823.92007 429 3 752 3 2 186 2 1 127 5 941 8 509 7 5 946 82007 429.3 752.3 2,186.2 1,127.5 941.8 509.7 5,946.82008 389.5 1,037.3 1,362.2 707.2 984.5 139.5 4,620.22009 409.8 2,185.5 2,041.4 901.8 1,117.0 150.9 6,806.42010 433.1 2,303.9 1,742.7 1,062.7 1,032.6 109.4 6,684.5

1 Interest bearing marketable coupon public debt.2 Includes GNMA, FNMA, and FHLMC mortgage‐backed securities and CMOs and private‐label MBS/CMOs.3 Includes all non‐convertible debt, MTNs and Yankee bonds, but excludes CDs and federal agency debt. 4 Beginning with 2004 Sallie Mae has been excluded due to privatization Source: SIFMA


4 Beginning with 2004, Sallie Mae has been excluded due to privatization. Source: SIFMA



MBS$1,000 Portfolio


Junior Tranche

Senior Tranche

$2,000 81% $1,000 $1,000

$1,000 18% $1,000 $0

$0 1% $0 $0

MBS$1,000

But What If Defaults BecomeHi hl C l t d?

Junior Tranche

Highly Correlated?



20250

Effective Fed Funds Rate June 2006: 226.29

S&P Case Shiller Composite 10 Index (nominal)

16200

S&P/CaseShiller Composite 10

Updated June 1, 201233.5%

12150

s R

ate

(%)

evel

8100

Fed

Fund

s

Inde

x Le

April 2009: 150.44July 1996: 78.10

450

Effe

ctiv

e

00

Jan-

87Ja

n-88

Jan-

89Ja

n-90

Jan-

91Ja

n-92

Jan-

93Ja

n-94

Jan-

95Ja

n-96

Jan-

97Ja

n-98

Jan-

99Ja

n-00

Jan-

01Ja

n-02

Jan-

03Ja

n-04

Jan-

05Ja

n-06

Jan-

07Ja

n-08

Jan-

09Ja

n-10

Jan-

11Ja

n-12

Jan-

13


J J J J J J J J J J J J J J J J J J J J J J J J J J J


Assuming Perfectly Correlated Defaults

MBS$1,000 Portfolio


Junior Tranche

Senior Tranche

$2,000 90% $1,000 $1,000

$0 10% $0 $0

Bad StateFor Senior

MBS$1,000

For SeniorTranche (10%) Bad State

For JuniorTranche (10%)

Junior Tranche

Tranche (10%)



Assuming Perfectly Correlated Defaults

MBS$1,000 Portfolio


Junior Tranche

Senior Tranche

$2,000 90% $1,000 $1,000

$0 10% $0 $0

Price for Senior Tranche = 90% × $1,000 + 10% × $0$900 ( $990)

MBS$1,000 = $900 (was $990)

Price for Junior Tranche = 90% × $1,000 + 10% × $0= $900 (was $810)

Junior Tranche

$ ( $ )


Where did all the money go??

Mortgages Outstanding and Equity Extractions

© 2009 by Khandani, Lo, and Merton

All Rights Reserved9/10/2009 Slide 16

How could this have happened to us?

Hall of Shame?: HomeownersHomeowners Commercial banks Investment banks and other issuers of MBSs, CDOs, and CDSs Mortgage lenders, brokers, servicers, trustees Credit rating agencies (S&P, Moody’s, Fitch)I i ( ltili li ) Insurance companies (multiline, monoline) Investors (hedge funds, pension funds, mutual funds, others) Regulators (SEC OCC CFTC Fed etc )Regulators (SEC, OCC, CFTC, Fed, etc.) Government sponsored enterprises Politicians

Is the financial sector becoming too big?


Is the financial sector becoming bigger? Yes!


Why?

World Population, 10,000 BC to 2011 AD

Source: U S Census


Source: U.S. Census.

Innovation requires collaboration

Collaboration requires financial infrastructure…P i t i t t Private investment

Well‐functioning capital marketsP d i f i i Proper design of securities

Accounting, legal, regulatory structuresb l Systemic stability

and aligned incentives (human behavior) …and aligned incentives (human behavior) Fear works faster Greed lasts longer Greed lasts longer


An example…The DARPA network challenge Find 10 red balloons in fixed locations in U S Find 10 red balloons in fixed locations in U.S. $40K prize to first team to find all 10 locations MIT team won MIT team won 8:52:41! How??

Financial engineering! Financial engineering! $4,000 per balloon if they won= $2,000 to locator +

h k d l $1,000 to person who asked locator +$ 500 to person who asked person + …+ remainder to charity


An example…Real Home Price Index (1890-2006)

200

E = mcE = mc22

160

180

Rat

e

120

140

or In

tere

st R

80

100

120

Inde

x o

60

80

1880 1900 1920 1940 1960 1980 2000 2020

Source: Robert J. Shiller, Irrational Exuberance, 2nd. Edition.

Year


The power of global capital markets

What if we could focus this power for ”good”? Financial engineering with a “conscience” Financial engineering with a conscience Apply expertise to tackling society’s biggest challenges Facilitate collaboration collective intelligence Facilitate collaboration, collective intelligence

With the proper financial engineering, we believe we can solve the following problems within the next 20 years:1. Cancer2. Energy crisis3. Global warming

How?


Curing cancer Cancer is a complex and growing conjoint of over 200 diseases Many possible mechanisms for fighting cancer:

over 900 drugs and vaccines being developed Inventory of 20+ years of late‐stage drugs to be financed (Stewart and Naeymi‐Rad, 2010)

Pharma cannot pursue more than a few; VC resources are limitedp R&D productivity has fallen over the past few years

FDA New Drug Approvals

53

3940

50

60

New molecular entities

Biologics license application

2521

29 3035

2724

1721

31

18 18 1621 19

15

24

20

30

40

total app

rovals

15

3 2 03

6 73 2

5 7 6 52 4 2 3

6 6 6

0

10

1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011


1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

Curing cancer: would you invest?

Would you invest in a single drug discovery project? Consider the following stylized example:Consider the following stylized example:

Investment per new compound $200MM 5% success rate, 10‐year development cycle5% success rate, 10 year development cycle Net income of approved blockbuster drug = $2b/year for 10 years PV10 (y 11‐20@ 10% cost of capital) = $12.3b



Would you invest in a single drug discovery project? Consider the following stylized example:Consider the following stylized example:

Investment per new compound $200MM 5% success rate, 10‐year development cycle5% success rate, 10 year development cycle Net income of approved blockbuster drug = $2b/year for 10 years PV10 (y 11‐20@ 10% cost of capital) = $12.3b

Return: SUCCESS: 51% = (12.3/0.2)1/10 1

E[R] = 11.9%

Ri k S d d d i i f li d ROI %%

FAILURE: 100%

Risk: Standard deviation of annualized ROI: 423.5%423.5%


Curing cancer: would you invest?What if you had the option of investing in 150 new drug programs at the same time?

ROI and Probabilities Risk

Standard deviation of annualized ROI = Probability of at least 5 successes is 87.4 % Standard deviation of annualized ROI = 423% √150) = 34.6%

Probability of at least 5 successes is 87.4 %ROI( exactly 5 success) = (5*12.3/30)^0.1‐1 = 7.4%



Phase IIPhase III

Approved

WD/SoldPhase I

Phase II

WD/SoldWD/Sold

WD/Sold

/ ldPhase III

Approved1

WD/Sold

Phase IPhase II

WD/SoldWD/Sold

WD/Sold

2 WD/SoldWD/Sold

Ph IIIApproved

Phase IPhase II

Phase III

WD/SoldWD/Sold

150 WD/SoldWD/Sold

Year 1 Year 2 Year 3 Year 4 Year 10

150


Year 1 Year 2 Year 3 Year 4 … Year 10


R id lE i • ResidualEquity

• Interest PaymentsA Bonds

I t t P tA B d • Interest PaymentsAa Bonds

• Interest PaymentsAaa Bonds

Year 1 Year 2 Year 3 Year 4 Year 10


Year 1 Year 2 Year 3 Year 4 … Year 10


Given the reduction in risk, debt‐financing is now possible!

Default probability is determined by the probability of hits Default probability is determined by the probability of hits:

$16.8B of Aaa debt can be issued With securitization, debt capacity may be even larger


Research Based Obligations (RBOs)

Business structure of a biomedical megafund SPVFunds are raised from retail or institutional investors (1) through the capital markets issuance (2) of various typesof debt and equity. These funds are invested in molecules being developed to cure cancer (3). Some funds arereserved to pay for later clinical development costs and, if required, to cover the first few periods of couponpayments. The portfolio of drugs is developed over time (4). At any time a compound can be discontinued, moveto the next phase or even two phases ahead based on the results of the trials done. It is also possible thatcompounds are sold prior to their FDA approval for marketing if it is necessary to monetize them to cover someof the fund interest or principal payments. Any com‐ pound that is approved for marketing as a new drug is soldto a biopharma company. At the end of the life of the fund, all remaining compounds in the portfolio are sold (5).After bondholders are paid back (6), the residual cash is used to pay back the equity holders (7).



Is there capacity/appetite from investors? In 2010:U S b d k t $ T U.S. bond market: $35T

Mutual funds: $12T Money‐market funds: $1T Norwegian sovereign wealth fund: $537BNorwegian sovereign wealth fund: $537B CalPERS: $226B T t t f 6 bli f d ( ) 8% Target return of 126 public funds (2010): 8%

I th i f th ti VC i d t 6BIn 2010, the size of the entire VC industry was $176B



How does this model differ from existing solutions? Debt financing thro gh risk mitigation (compared to 0 companies!) Debt‐financing through risk mitigation (compared to 150 companies!)

Industry risk instead of single‐company risk

Investing in intellectual property not existing companies Investing in intellectual property, not existing companies

Investing in early stage projects (“valley of death”), not just mature products

Capital structure of the fund better matches investor preferences Capital structure of the fund better matches investor preferences

Access to vastly larger pools of potential investors

Long term debt facilitates potentially longer investment horizons Long‐term debt facilitates potentially longer investment horizons

All stakeholders benefit (investors, managers, researchers, insurers, regulators patients)regulators, patients)

Funding mechanism creates a new, more powerful, and sustainable bridge between capital markets and sciencep



How would various stakeholders be affected? VCs reduces future finance risks to take on projects currently unfunded VCs: reduces future‐finance risks to take on projects currently unfunded

Biopharma: larger supply of drugs to distribute, lower risk, greater liquidity

Drug development companies more business resources and liquidity Drug development companies: more business, resources, and liquidity

Investors: new supply of investable assets with less correlated returns

Health insurance companies new ways to hedge their balance sheets Health insurance companies: new ways to hedge their balance sheets

Investment bankers: new business using existing infrastructure, PR

Government regulators new opportunities to increase private sector Government, regulators: new opportunities to increase private‐sector involvement in the funding of innovation; jobs; politically popular issue

Foundations: new channel through which to support their goalsFoundations: new channel through which to support their goals

Society:New, faster, cheaper cures!



What are some of the potential challenges? Size: managing large portfolios of complex R&D projects may require new Size: managing large portfolios of complex R&D projects may require new management and governance structures (e.g., Manhattan Project)

Centralization: must preserve the benefits of diversity as scale increases p y

Capacity: is the talent pool large enough to match the scale of this venture?

Complexity: can investors understand the risks and rewards of RBOs?p y

Excesses: if successful, the potential for abuse will also increase

Ethics: how to balance profit motive vs. social objectives for cures?Ethics: how to balance profit motive vs. social objectives for cures?



With Some Imagination, This Can Be Achieved! Imagine creating a $30B “Cure For Cancer” megafund Imagine creating a $30B Cure For Cancer megafund Imagine creating an advisory board of experts:

Francis Collins Eric Lander Bob Langer Phil Sharp Harold Varmus Francis Collins, Eric Lander, Bob Langer, Phil Sharp, Harold Varmus, Craig Venter; Warren Buffett, Bill Gates, Jacob Goldfield, Bob Merton, Jim Simons, George Soros, Bill Sharpe

I i MM h h ld i ti h Imagine 10MM households investing $3,000 each Imagine corporate pension funds, foundations,

d t i i i ti llendowments, insurance companies investing as well Imagine government tax incentives, credit

h t t (thi k F i M F ddi M !)enhancement, etc. (think Fannie Mae, Freddie Mac!)


Conclusion

Don’t Declare War On Put A Price Tag On ItsDon t Declare War On Cancer…

Put A Price Tag On Its Head Instead!

With Sufficient Scale, We Can Do Well By Doing Good


Conclusion

“Can We Afford It?”Not To Try?”– heart disease, Alzheimer’s, dementia,

diabetes, obesity, H1N1, energy crisis, climate change, solar flares

With the Proper

c ate c a ge, so a a es

With the Proper Financial Structure,Financial Structure, AnythingAnything Is Possible!


Thank You!Thank You!

andrew lo

Documents

rights reserved9212010

g p p p

senior trancheportfolioi

real home price index

real drop

bad statefor seniortranche

josemaria fernandez

financial engineering