role of state bank

8/3/2019 Role of State Bank

1/63

Role of State BankMajority of people think printing and circulating

the currency notes is the only responsibility ofState Bank. They think central banks role islimited to these two functions. It is not like that atall, rather currency printing and circulating aresome of many primary functions that centralbank undertakes to perform.State Bank functions have direct and indirect

impact on whole economy as it deals with PriceStability and Overall Growth which is achievedthrough Monetary Policy formulation andimplementation in the country. Monetary policymanages money supply such that neitherexcess nor shortage of money prevails in theeconomy. In order to control money, interest rateis used as a tool. Again interest rate hasrelationship with inflation, investment, productionand overall economic development in thecountry. Hence, Central bank controls economicindicator of interest rate which has direct andindirect impact on all other major economic

indicators.Therefore SBP has to remain very cautious allthe time and take into account all stakeholdersof the economy while formulating monetarypolicy.


2/63

As a regulator of economy, SBP regulates thefinancial system and is advisory body toGovernment of Pakistan on various financial

issues. It manages countrys foreign reserves,foreign and domestic debts and their debtservicing, all on behalf of Government ofPakistan.Government of Pakistan does financing from aswell as through SBP. It works as Lender of LastResort for financial institutions when they areshort of funds or near to default.If we look at the performance of SBP throughouthistory, we would come to know about greaterachievements of central bank. Achievementsinclude liberalisation of financial system byprivatising public sector banks, encouraging

private sector banks and increasing competitionamong all banks through its policies. Itsuccessfully terminated monopoly of publicsector banks and more than 40 private banksare operating with branch network of around2500 in the country at present. Pakistanibanking industry is considered sound, robust

and efficient in the Asian region, credit for thisagain goes to efficient and sound regulation ofthe central bank. A survey, for regulatory bodies(including SBP, SECP, PEMRA, NEPRA etc)was conducted in the country for role, efficiency,


3/63

and performance of the regulators. As a result ofthe survey, SBP ranked first.Political and economic stabilities are directly

linked and play major role for developingcountries like Pakistan. State Bank is aprofessional financial regulator and preparesMonetary Policy that ensures economicdevelopment. Owing to fiscal indiscipline andnon coordination with monetary policy, all effortsof SBP go in vain. Political instability and fiscalindiscipline deteriorate economic stability verybadly. Therefore, if Ministry of Finance ensuresfiscal coordination, SBP is firm towardsstabilising economy and consequently leadingtowards stable and prosperous Pakistan.

Reserve Ratio

What Does Reserve Ratio Mean?The portion (expressed as a percent) ofdepositors' balances banks must have on handas cash. This is a requirement determined bythe country's central bank, which in the U.S. is

the Federal Reserve. The reserve ratio affectsthe money supply in a country.

This is also referred to as the "cash reserveratio" (CRR).


4/63

Read

more: http://www.investopedia.com/terms/r/reserveratio.asp#ixzz1cBaUqj5X

required reserve ratioBanking

ratio of banks reserves to depositsthe proportion of a bank's deposits that must be kept

in reserve.

In the United Kingdom and in certain European countries, there is no compulsory ratio,although banks will have their own internal measures and targets to be able to repay

customer deposits as they forecast they will be required. In the United States, specified

percentages of depositsestablished by the Federal Reserve Boardmust be kept by

banks in a non-interest-bearing account at one of the twelve Federal Reserve Banks

located throughout the country.

In Europe, the reserve requirement of an institution is calculated by multiplying the

reserve ratio for each category of items in the reserve base, set by the European Central

Bank, with the amount of those items in the institution's balance sheets. These figures

vary according to the institution.

The required reserve ratio in the United States is set by federal law, and depends on the

amount of checkable deposits a bank holds. Up to $9.3M the required reserve ratio is

0%, from $9.3M to $43.9M it is 3% and above $43.9M it is 10%. These breakpoints are

reviewed annually in accordance with money supply growth. No reserves are required

against certificates of deposit or savings accounts.

The reserve ratio requirement limits a bank's lending to a certain fraction of its demand

deposits. The current rule allows a bank to issue loans in an amount equal to 90% of

such deposits, holding 10% in reserve. The reserves can be held in any combination oftill money and deposit at a Federal Reserve Bank.

Moral hazard
http://www.investopedia.com/terms/r/reserveratio.asp#ixzz1cBaUqj5Xhttp://www.investopedia.com/terms/r/reserveratio.asp#ixzz1cBaUqj5Xhttp://www.investopedia.com/terms/r/reserveratio.asp#ixzz1cBaUqj5Xhttp://www.investopedia.com/terms/r/reserveratio.asp#ixzz1cBaUqj5X


5/63

In economic theory, moral hazard is a situationin which a party insulated from risk behavesdifferently from how it would behave if it were

fully exposed to the risk.Moral hazard arises because an individual orinstitution does not take the full consequencesand responsibilities of its actions, and thereforehas a tendency to act less carefully than itotherwise would, leaving another party to hold

some responsibility for the consequences ofthose actions. For example, a person withinsurance against automobile theft may be lesscautious about locking his or her car, becausethe negative consequences of vehicle theft are(partially) the responsibility of the insurancecompany.

Economists explain moral hazard as a specialcase ofinformation asymmetry, a situation inwhich one party in a transaction has moreinformation than another. In particular, moralhazard may occur if a party that is insulated fromrisk has more information about its actions and

intentions than the party paying for the negativeconsequences of the risk. More broadly, moralhazard occurs when the party with moreinformation about its actions or intentions has atendency or incentive to behave inappropriately
http://en.wikipedia.org/wiki/Economic_theoryhttp://en.wikipedia.org/wiki/Riskhttp://en.wikipedia.org/wiki/Information_asymmetryhttp://en.wikipedia.org/wiki/Economic_theoryhttp://en.wikipedia.org/wiki/Riskhttp://en.wikipedia.org/wiki/Information_asymmetry


6/63

from the perspective of the party with lessinformation.

Moral hazard also arises in a principal-agent

problem, where one party, called an agent, actson behalf of another party, called the principal.The agent usually has more information abouthis or her actions or intentions than the principaldoes, because the principal usually cannotcompletely monitor the agent. The agent may

have an incentive to act inappropriately (fromthe viewpoint of the principal) if the interests ofthe agent and the principal are not aligned

In insurance

In insurance markets, moral hazard occurs whenthe behavior of the insured party changes in away that raises costs for the insurer, since theinsured party no longer bears the full costs ofthat behavior. Because individuals no longerbear the cost of medical services, they have anadded incentive to ask for pricier and more

elaborate medical servicewhich wouldotherwise not be necessary. In these instances,individuals have an incentive to over consume,simply because they no longer bear the full costof medical services.
http://en.wikipedia.org/wiki/Principal-agent_problemhttp://en.wikipedia.org/wiki/Principal-agent_problemhttp://en.wikipedia.org/wiki/Principal-agent_problemhttp://en.wikipedia.org/wiki/Principal-agent_problem


7/63

Two types of behavior can change. One type isthe risky behavior itself, resulting in a before theeventmoral hazard. In this case, insured parties

behave in a more risky manner, resulting inmore negative consequences that the insurermust pay for. For example, after purchasingautomobile insurance, some may tend to be lesscareful about locking the automobile or chooseto drive more, thereby increasing the risk of theftor an accident for the insurer. After purchasingfire insurance, some may tend to be less carefulabout preventing fires (say, by smoking in bed orneglecting to replace the batteries in firealarms).[citation needed]

A second type of behavior that may change isthe reaction to the negative consequences of

risk, once they have occurred and onceinsurance is provided to cover their costs. Thismay be called ex post(after the event) moralhazard. In this case, insured parties do notbehave in a more risky manner that results inmore negative consequences, but they do askan insurer to pay for more of the negativeconsequences from risk as insurance coverageincreases. For example, without medicalinsurance, some may forgo medical treatmentdue to its costs and simply deal with
http://en.wikipedia.org/wiki/Ex_antehttp://en.wikipedia.org/wiki/Ex_antehttp://en.wikipedia.org/wiki/Wikipedia:Citation_neededhttp://en.wikipedia.org/wiki/Wikipedia:Citation_neededhttp://en.wikipedia.org/wiki/Wikipedia:Citation_neededhttp://en.wikipedia.org/wiki/Ex_antehttp://en.wikipedia.org/wiki/Ex_antehttp://en.wikipedia.org/wiki/Wikipedia:Citation_needed


8/63

substandard health. But after medical insurancebecomes available, some may ask an insuranceprovider to pay for the cost of medical treatment

that would not have occurred otherwise.Sometimes moral hazard is so severe it makesinsurance policies impossible. Coinsurance, co-payments, and deductibles reduce the risk ofmoral hazard by increasing the out-of-pocketspending of consumers, which decreases their

incentive to consume. Thus, the insured have afinancial incentive to avoid making a claim.

Moral hazard has been studied by insurers[4] andacademics. See works by Kenneth Arrow,[5][6][7]Tom Baker,[8] and John Nyman.

John Nyman suggests that two types of moralhazard exist: efficient and inefficient moralhazard. Efficient moral hazard is the viewpointthat the over consumption of medical carebrought forth by insurance does not alwaysproduce a welfare loss to society. Rather,individuals attain better health through theincreased consumption of medial care, making

them more productive and netting an overallbenefit to societal welfare. Also, Nymansuggests that individuals purchase insurance toobtain an income transfer when they become ill,
http://en.wikipedia.org/wiki/Moral_hazard#cite_note-3http://en.wikipedia.org/wiki/Kenneth_Arrowhttp://en.wikipedia.org/wiki/Moral_hazard#cite_note-4http://en.wikipedia.org/wiki/Moral_hazard#cite_note-5http://en.wikipedia.org/wiki/Moral_hazard#cite_note-6http://en.wikipedia.org/wiki/Moral_hazard#cite_note-Baker1996-7http://en.wikipedia.org/wiki/Moral_hazard#cite_note-3http://en.wikipedia.org/wiki/Kenneth_Arrowhttp://en.wikipedia.org/wiki/Moral_hazard#cite_note-4http://en.wikipedia.org/wiki/Moral_hazard#cite_note-5http://en.wikipedia.org/wiki/Moral_hazard#cite_note-6http://en.wikipedia.org/wiki/Moral_hazard#cite_note-Baker1996-7


9/63

as opposed to the traditionalist stance thatindividuals diversify risk via insurance.

Insurance analysts sometimes distinguish moral

hazard from a related concept they call moralehazard.[edit]

Economic Indicators: Consumer Credit Report

The Consumer Credit Reportis a monthly

release from the Federal Reserve Board thatestimates changes in the dollar amounts ofoutstanding loans to individuals, funds which aremainly used to purchase consumer goods.Loans backed by real estate, such as homeequity lines of credit (HELOCs), are not includedin the survey. The two classes of credit covered

are revolving and non-revolving credit; revolvingcredit can be increased by the consumer up to alimit without contacting the creditor (as in creditcards), while non-revolving terms are fixed at thetime the loan (as with an auto loan).

Both classes are segmented into the categoriesbelow. The Consumer Credit Reportshows theoutstanding balances for each:

Commercial banks
http://en.wikipedia.org/wiki/Morale_hazardhttp://en.wikipedia.org/wiki/Morale_hazardhttp://en.wikipedia.org/w/index.php?title=Moral_hazard&action=edit&section=3http://www.investopedia.com/terms/l/loan.asphttp://www.investopedia.com/terms/h/homeequitylineofcredit.asphttp://www.investopedia.com/terms/h/homeequitylineofcredit.asphttp://www.investopedia.com/terms/r/revolvingcredit.asphttp://en.wikipedia.org/wiki/Morale_hazardhttp://en.wikipedia.org/wiki/Morale_hazardhttp://en.wikipedia.org/w/index.php?title=Moral_hazard&action=edit&section=3http://www.investopedia.com/terms/l/loan.asphttp://www.investopedia.com/terms/h/homeequitylineofcredit.asphttp://www.investopedia.com/terms/h/homeequitylineofcredit.asphttp://www.investopedia.com/terms/r/revolvingcredit.asp


10/63

Finance companies Credit unions Federal government & Sallie Mae Savings institutions Non-financial businesses Securitized asset pools

Average interest rates are shown for many typesof consumer debt, such as auto loans, creditcards and bank loans, collectively showinginvestors the overall "credit quality" ofconsumers and where the highest rates ofgrowth are occurring.

Data is collected through surveys of banks,finance companies, retail sales outfits and credit

unions, among others. Each release will showthe three previous months' results, including anyrevisions to recent periods, if they haveoccurred. (For related reading, see ConsumerCredit Report: What's On It.)

What it Means for Investors

Consumer credit is considered a good indicatorof the potential future spending levels seen inthe Personal Consumption and RetailSales reports, and shows the extent to whichbenchmark interest rates such as the fed funds
http://www.investopedia.com/articles/04/122304.asphttp://www.investopedia.com/articles/04/122304.asphttp://www.investopedia.com/terms/f/federalfundsrate.asphttp://www.investopedia.com/articles/04/122304.asphttp://www.investopedia.com/articles/04/122304.asphttp://www.investopedia.com/terms/f/federalfundsrate.asp


11/63

rate and prime rate have manifested themselvesat the consumer level (it can take six months toa year for macro interest rates to work their way

down to consumers).

The headline stats of this release will betotal consumer debt (expressed in trillions andseasonally adjusted), the current annual runrate of growth or decline, and the totalpercentage of credit card delinquencies. Thedelinquencies are studied because suddenspikes may lead to fears that consumers areoverextended in their debt levels. Someeconomists will try to compare the defaultpercentages seen in the most recent recessionas a breakpoint - if current default levels

approach it, they will look for a recessionarytrend to show itself in other economic indicators.

These factors are important when investorsconsider that consumers make up more thantwo-thirds of total GDP consumption. Ifconsumers stop spending or face a credit

crunch, GDP will not be able to growmuch. Investors in consumer cyclical stocksshould be keenly interested in consumers' abilityto spend more in the future.
http://www.investopedia.com/terms/f/federalfundsrate.asphttp://www.investopedia.com/terms/p/primerate.asphttp://www.investopedia.com/terms/c/consumercredit.asphttp://www.investopedia.com/terms/r/runrate.asphttp://www.investopedia.com/terms/r/runrate.asphttp://www.investopedia.com/terms/d/default2.asphttp://www.investopedia.com/terms/f/federalfundsrate.asphttp://www.investopedia.com/terms/p/primerate.asphttp://www.investopedia.com/terms/c/consumercredit.asphttp://www.investopedia.com/terms/r/runrate.asphttp://www.investopedia.com/terms/r/runrate.asphttp://www.investopedia.com/terms/d/default2.asp


12/63

Consumer credit figures have a lot of seasonaland inherent volatility, so investors shouldalways review the current report for adjustmentsto prior periods, paying particular attention torevisions to year-over-year growth. Long-termtrends are the most studied portion of the report,both in the total outstanding balances as well asthe change in overall interest rates beingcharged.

The Conference Board has tapped consumercredit as a lagging indicator, and uses a ratio ofconsumer credit to personal income as a

component of its Index of LaggingIndicators. The Fed operates on the theory thatconsumers will not significantly increase theirborrowing levels until their personal incomesincreases enough to justify the higher debt load.As such, borrowing may show the largestincreases when the economy is already coming

out of a recession, rather than during the worstof it.

Strengths:
http://www.investopedia.com/terms/c/conferenceboard.asphttp://www.investopedia.com/terms/l/laggingindicator.asphttp://www.investopedia.com/terms/c/cilagi.asphttp://www.investopedia.com/terms/c/cilagi.asphttp://www.investopedia.com/terms/c/conferenceboard.asphttp://www.investopedia.com/terms/l/laggingindicator.asphttp://www.investopedia.com/terms/c/cilagi.asphttp://www.investopedia.com/terms/c/cilagi.asp


13/63

Contains detailed breakdown of auto loanfigures, such as average maturity andprevailing interest rates

Data is provided with and without seasonaladjustments.

Release shows comparisons againstprevious month, previous year, and alsoagainst results from the last five years

Weaknesses: Only total growth in outstanding loans is

shown; there is no way of knowing ifconsumer payments have fallen off or if newloan growth has slowed based on a fallingconsumer credit number (and vice versa).

Absence of home-equity debt provides foran incomplete picture.

Because it comes out after the consumerconfidence report and retail sales reports forthe month, some analysts will not look asintently at the consumer credit figures monthto month, instead reviewing multi-periodtrends once or twice a year

The Closing LineThe Consumer Credit Reportwill not be a bigcatalyst in the markets because of earlier-released indicators, but it remains a goodlagging indicator, especially when examined in


14/63

conjunction with personal wage growth andinterest rates. If prevailing rates are moderateand incomes are rising, consumer credit can

grow in step without causing elevated fears inthe market.

U.S. Consumer Credit Card Debt May CrashEconomy

U.S. Consumer Credit Card Debt May Crash

Economy

Readmore: http://www.foxnews.com/story/0,2933,143037,00.html#ixzz1cBcox3cS

Christmas has come and gone and the biggest

buying spree of the year is over. Now comes thetough part: cutting up those credit cards to keepfrom charging anymore in the new year.American consumers could keep EdwardScissorhands (search) busy for months justslicing up our maxed-out credit cards.Remember him? In the 1990 Tim Burton movie

about the young man with scissors for hands,Johnny Depp mainly used his talentedappendages to trim hair and create whimsicalcreatures out of shrubbery.
http://www.foxnews.com/story/0,2933,143037,00.html#ixzz1cBcox3cShttp://www.foxnews.com/story/0,2933,143037,00.html#ixzz1cBcox3cShttp://www.foxnews.com/story/0,2933,143037,00.htmlhttp://www.foxnews.com/story/0,2933,143037,00.htmlhttp://search.foxnews.com/info.foxnws/redirs_all.htm?pgtarg=wbsdogpile&qcat=web&qkw=Edward%20Scissorhandshttp://www.foxnews.com/story/0,2933,143037,00.html#ixzz1cBcox3cShttp://www.foxnews.com/story/0,2933,143037,00.html#ixzz1cBcox3cShttp://www.foxnews.com/story/0,2933,143037,00.htmlhttp://search.foxnews.com/info.foxnws/redirs_all.htm?pgtarg=wbsdogpile&qcat=web&qkw=Edward%20Scissorhands


15/63

But what's in order now is much morestraightforward snipping. Sure, everyone spendsa little more than they mean to at this time of

year; the problem is that we've all beenspending a little more than we mean tothroughout the last few years. Let's take a lookat our credit habits as a nation to get an idea ofwhy we may be in over our heads.About a month ago, the New York Timesexamined how the use of credit has taken offdramatically in the United States since 1990.While the number of people holding chargecards grew about 75 percent from 82 million in1990 to 144 million in 2003 the amount theycharged during that period grew by a muchlarger percentage: approximately 350 percent,

from $338 billion to $1.5 trillion.Since the number of chargers is growing at aslower rate than the amounts being charged,you can guess what that means. Yes, themonthly revolving balances have been growingby leaps and bounds. In 1990, the Timesreports, the average was about $2,550 for those

households that carried a balance. At the end of2003, that balance averaged about $7,520 anincrease of nearly 200 percent!And that's only credit cards. The average U.S.household owes mortgage debt, student loans


16/63

and automobile loans, in addition to credit carddebt.Why are we doing this to ourselves? Personal

income certainly hasn't risen 350 percent overthe past 13 years. Let's see, that would meanthat a person who depended on salary as theonly source of income would have had to see a$35,000 salary in 1990 grow to $122,500 in2003 to keep up with the extra credit load.Possible, but not too likely.In fact, on the national level, the Bureau of LaborStatistics shows that aggregate U.S. personalincome in 1990 was $4.9 trillion. In 2003, it was$9.2 trillion. The rate of growth? 188 percent pretty far off from the 350 percent growth incredit card charges.

This give-me-credit-or-give-me-death behaviorreminds me of that America Online commercialwhere hapless souls say things like: "I'd reallylike all my files to be obliterated by a virus." Byanalogy, you might think that consumers weresaying, "I'd really like to owe a whole bunchmore than I can afford, so that I can

declare bankruptcy."There's a reason you don't hear anyone sayingthose words: It's because credit has been soeasy to get that we've managed to inflatepeople's ideas of how much they can take on.
http://www.foxnews.com/story/0,2933,143037,00.htmlhttp://www.foxnews.com/story/0,2933,143037,00.html


17/63

Barron's recently published a striking chart,showing Household Surplus (more income thanoutgo) compared with Household Deficit.

Throughout the 1960s, '70s, '80s, and '90s,households showed a surplus of varyingdegrees. It wasn't until 1999 for the first timein about 50 years that U.S. householdsstarted spending more than they took in. Whatstarted as a small deficit of about $50 billionamong households quickly spiked to a deficit ofmore than $350 billion in the second quarter ofthis year.There's an axiom known by every economistunder the sun: because consumers make uptwo-thirds of the economy, they must keepspending to keep the economy healthy. It's easy

to see that we've been acting on cue for theeconomy, to the point that we've kept spendingwell beyond our means.And perhaps we've been doing it because as anation, we're basically optimistic optimistic thatwe will always make more money to cover ourdebts, optimistic that the value of the houses we

live in will continue to go up, optimistic that theeconomy will improve and create manynewjobs.In fact, that same optimism is implicit in thecredit inflation bestowed on us by the Federal
http://www.foxnews.com/story/0,2933,143037,00.htmlhttp://www.foxnews.com/story/0,2933,143037,00.html


18/63

Reserve (search). To keep the economypumped up, Greenspan & Co. have had to keepthe consumer pumped up. And easy credit is the

easiest path to take.The only problem with this strategy is that atsome point, consumers do reach a limit whenthey recognize that they're overstretched. It maybe the day they notice that they've maxed out alltheir cards. Or the day they apply for anothercard and are turned down. Or the day the creditcard company doubles their interest rate if theymiss one payment.But the day will also come when consumers willwish they'd played the role of EdwardScissorhands themselves. As consumeroptimism slips away, it will affect the credit card

companies, all of whom have been making akilling in their pretax profits, as well as on the feerevenues they collect from consumers. Pretaxprofits for the credit card companies as reportedby the Times have grown 360 percent from 1990to 2003, while fee revenues have grown 250percent. Do you think they want to close that

spigot voluntarily?The larger picture is even grimmer somethingwe haven't faced as a nation in a long time. Asfewer people have the ability to buy stuff, pricesmust come down. And when prices come down
http://search.foxnews.com/info.foxnws/redirs_all.htm?pgtarg=wbsdogpile&qcat=web&qkw=Federal%20Reservehttp://search.foxnews.com/info.foxnws/redirs_all.htm?pgtarg=wbsdogpile&qcat=web&qkw=Federal%20Reserve


19/63

across the board, what the Fed will be facing isnot that old enemy, inflation, but a new-oldenemy, deflation.

Readmore: http://www.foxnews.com/story/0,2933,143037,00.html#ixzz1cBdI6Mxr

The Economics of Consumer Credit

Academic research and policy discussionsof credit markets usually focus onborrowing by firms and producers ratherthan by households, which are typicallyanalyzed in terms of their savings andportfolio choices. The Economics ofConsumer Creditbrings together leadinginternational researchers to focusspecifically on consumer debt, presentingcurrent empirical and theoretical researchcrucial to ongoing policy debates on suchtopics as privacy rules, the regulation of

contractual responsibilities, financialstability, and overindebtedness.

The rapidly developing consumer credit
http://www.foxnews.com/story/0,2933,143037,00.html#ixzz1cBdI6Mxrhttp://www.foxnews.com/story/0,2933,143037,00.html#ixzz1cBdI6Mxrhttp://www.foxnews.com/story/0,2933,143037,00.html#ixzz1cBdI6Mxrhttp://www.foxnews.com/story/0,2933,143037,00.html#ixzz1cBdI6Mxr


20/63

industry in the United States is mirroredby that in Europe, and this volume isnoteworthy for its cross-national

perspective. Several chapters compare theuse of credit markets by households indifferent countries, while others focus onsingle country case studiesincludingconsumer credit dynamics in Italy, therole of housing expenditure in the cyclical

pattern of borrowing in the UnitedKingdom, and the use of credit cards byU.S. consumersto illustrate generalinsights. Other chapters draw policylessons from the U.S. experience withbankruptcy regulation and the

development of the credit counselingindustry. Finally, the book reviewshistorical, theoretical, and empiricalaspects of information sharing, ofparticular interest in light of theintegration of European Union creditmarkets.

Contributors: Carol C. Bertaut, GiuseppeBertola, Sarah Bridges, Luca Casolaro,Jonathan Crook, Richard Disney, Leonardo


21/63

Gambacorta, Charles Grant, Luigi Guiso,Michael Haliassos, Andrew Henley, RobertM. Hunt, Tullio Jappelli, Nicola Jentzsch,

Marco Pagano, Amparo San Jos Riestra,Michael Staten, Michelle J. White

About the Editors

Giuseppe Bertola is Professor of

Economics at the University of Turin andScientific Coordinator at Finance andConsumption, European UniversityInstitute, Florence.

Richard Disney is Professor of Economics

at the University of Nottingham andResearch Fellow at the Institute of FiscalStudies, London.

Charles Grant is a Lecturer at ReadingUniversity and a visiting ResearchAssociate at Finance and Consumption,European University Institute, Florence.

The Board of Governors

of the Federal Reserve System


22/63

Appointments to the Board

The seven members of the Board of Governors areappointed by the President and confirmed by theSenate to serve 14-year terms of office. Membersmay serve only one full term, but a member who has

been appointed to complete an unexpired term maybe reappointed to a full term. The President

designates, and the Senate confirms, two members ofthe Board to be Chairman and Vice Chairman, forfour-year terms.

RepresentationOnly one member of the Board may be selected from

any one of the twelve Federal Reserve Districts. Inmaking appointments, the President is directed bylaw to select a "fair representation of the financial,agricultural, industrial, and commercial interests andgeographical divisions of the country." These aspectsof selection are intended to ensure representation ofregional interests and the interests of various sectors

of the public.

ResponsibilitiesThe primary responsibility of the Board members is
http://www.federalreserve.gov/bios/default.htmhttp://www.federalreserve.gov/otherfrb.htmhttp://www.federalreserve.gov/bios/default.htmhttp://www.federalreserve.gov/otherfrb.htm


23/63

the formulation of monetary policy. The seven Boardmembers constitute a majority of the 12-memberFederal Open Market Committee (FOMC),

the group that makes the key decisions affecting thecost and availability of money and credit in theeconomy. The other five members of the FOMC areReserve Bank presidents, one of whom is the

president of the Federal Reserve Bank of New York.The other Bank presidents serve one-year terms on arotating basis. By statute the FOMC determines itsown organization, and by tradition it elects theChairman of the Board of Governors as its Chairmanand the President of the New York Bank as its ViceChairman.

The Board sets reserve requirements and shares theresponsibility with the Reserve Banks for discountrate policy. These two functions plus open marketoperations constitute the monetary policytools of theFederal Reserve System.

In addition to monetary policy responsibilities, theFederal Reserve Board has regulatory andsupervisory responsibilitiesover banks that are

members of the System, bank holding companies,international banking facilities in the United States,Edge Act and agreement corporations, foreignactivities of member banks, and the U.S. activities of
http://www.federalreserve.gov/fomchttp://www.federalreserve.gov/policy.htmhttp://www.federalreserve.gov/banknreg.htmhttp://www.federalreserve.gov/banknreg.htmhttp://www.federalreserve.gov/fomchttp://www.federalreserve.gov/policy.htmhttp://www.federalreserve.gov/banknreg.htmhttp://www.federalreserve.gov/banknreg.htm


24/63

foreign-owned banks. The Board also sets marginrequirements, which limit the use of credit for

purchasing or carrying securities.

In addition, the Board plays a key role in assuring thesmooth functioning and continued development ofthe nation's vast payments system[see Fedwire and Payment System Risk Policy].

Another area of Board responsibility is the

development and administration of regulations thatimplement major federal laws governing consumercredit such as the Truth in Lending Act, the EqualCredit Opportunity Act, the Home MortgageDisclosure Act and the Truth in Savings Act[see Consumer Information and CommunityDevelopment].

MeetingsThe Board usually meets several times aweek.Meetingsare conducted in compliance with theGovernment in the Sunshine Act, and many meetingsare open to the public. If the Board has convened to

consider confidential financial information, however,the sessions are closed to public observation.
http://www.federalreserve.gov/regulations/regref.htm#thttp://www.federalreserve.gov/regulations/regref.htm#thttp://www.federalreserve.gov/paymentsystems/fedwire/http://www.federalreserve.gov/paymentsystems/psr/http://www.federalreserve.gov/consumers.htmhttp://www.federalreserve.gov/community.htmhttp://www.federalreserve.gov/community.htmhttp://www.federalreserve.gov/boarddocs/meetings/http://www.federalreserve.gov/regulations/regref.htm#thttp://www.federalreserve.gov/regulations/regref.htm#thttp://www.federalreserve.gov/paymentsystems/fedwire/http://www.federalreserve.gov/paymentsystems/psr/http://www.federalreserve.gov/consumers.htmhttp://www.federalreserve.gov/community.htmhttp://www.federalreserve.gov/community.htmhttp://www.federalreserve.gov/boarddocs/meetings/


25/63

Contacts within GovernmentAs they carry out their duties, members of the Boardroutinely confer with officials of other government

agencies, representatives of banking industry groups,officials of the central banks of other countries,members of Congress and academicians. Forexample, they meet frequently with Treasury officialsand the Council of Economic Advisers to helpevaluate the economic climate and to discussobjectives for the nation's economy. Governors alsodiscuss the international monetary system withcentral bankers of other countries and are in closecontact with the heads of the U.S. agencies that makeforeign loans and conduct foreign financialtransactions.

How to control trade deficit

In the short term, an increase in the trade deficit

is an indication of economic recovery but this will

affect the balance of payment, foreign currency

reserves and exchange rates, and destabilise

macro finances.

Trade deficit control proven ineffective

According to the General Statistics Office (GSO),export turnover in April was estimated to reach
http://www.federalreserve.gov/general.htmhttp://www.federalreserve.gov/general.htm


26/63

US$5.7 billion, up 1.9 percent over March whileimports were likely to hit US$6.95 billion, up 3

percent. The trade deficit in April jumped to

approximately US$1.25 billion compared to US$1.16 billion in March, US$1.33 billion in Februaryand US$ 945 million in January. The total tradedeficit in the past four months was US$4.65 billion,accounting for 23 percent of export turnover, andsurpassing the National Assemblys set target ofkeeping the trade deficit below 20 percent.

Trade deficit control has proved ineffective asimports rose 32.6 percent but products subject toimport controls such as seafood, fruits andvegetables, steel, gems and precious metal increased59 percent. Products subject to import limits(consumer goods, fully-assembled cars seating undernine people and motorbikes rose 41 percent.

Mr. Phi Dang Minh from the Foreign CurrencyManagement Department says that the exchange rate

policy is not a main cause of the trade deficit inVietnam as the State Bank of Vietnam (SBV) hasadjusted the exchange rates in line with the

Governments monetary policy and economicgrowth.


27/63

Mr Minh attributes the recent increased trade deficitto a trade imbalance which has led to exportsincreasing faster than imports.

In addition, Vietnam has to reduce its import tax ratesunder international commitments while price hikesfor materials in the global market has sent importvalues soaring. The global economic downturn hasalso limited Vietnams import capacity.

China, Vietnams largest trade partner, has seen ayear-on-year increase of 40 percent in two-way tradeturnover. The figure reached more than US$21.65

billion in 2008, US$20.751 billion in 2009 andUS$5.37 billion at the end of the first quarter of thisyear, up 37.8 percent over the same period last year.

It is worth noting that Vietnams import surplus fromChina is increasing in value and accounts for themajority of Vietnams total trade deficit (73.7 percentin 2007, 69.8 percent in 2008, 97.1 percent in 2009and an estimated at 94.9 percent in 2010).

While Vietnams exports exceed imports from theUS, UK, Germany, and Australia, its imports fromChina increasingly surpass exports. Vietnam needsmore effective measures to deal with its trade deficitwith China.


28/63

In addition, the recent devaluation of the US dollaragainst the VN dong has caused a fall in the price ofimported goods which is considered a competitive

advantage in the domestic market.

On the other hand, FDI disbursement has seen asignificant increase over the past few months,reaching US$3.4 billion in April, according to theForeign Investment Agency. FDI sector imports havealso risen sharply since the beginning of 2010.

The export and import of gold is another factor thataffects Vietnams international trade relations and themass importation of gold in 2010 has led to a tradeimbalance.

Trade deficit and solution to the problem

Dr. Nguyen Mai, Vice-Chairman of the StateCommittee for Cooperation and Investment (SCCI),says the trade deficit should be viewed not only interms of figures but also its structure.

It is a positive sign if we import more machinery,equipment and materials because it shows a recovery

in production, says Dr. Mai.

In fact, businesses have spent US$2.73 billion toimport cotton and other materials for textiles andgarments and footwear, much more than last year.


29/63

Meanwhile, US$4 billion has been spent onmachinery and equipment and US$1.5 billion on

plastic and plastic products, bringing the import value

of these three groups to US$8 billion, accounting forone third of the countrys imports.

According to Mai, the increase in imports also meansan increase in production and business activities. Inthe first four months of 2010, Vietnamese businessesearned more than US$3 billion from garment and

textile exports and US$1.35 billion from footwearexports.

One solution for reducing the trade gap is to developsupport industries to encourage domestic productioninstead of importing. This would also to help thecountry deal with other problems like inflation

http://www.cpv.org.vn/cpv/Modules/News_English/News_Detail_E.aspx?CN_ID=401473&CO_ID=30113

Controlling trade deficit

By Agha Saiddain
http://www.cpv.org.vn/cpv/Modules/News_English/News_Detail_E.aspx?CN_ID=401473&CO_ID=30113http://www.cpv.org.vn/cpv/Modules/News_English/News_Detail_E.aspx?CN_ID=401473&CO_ID=30113http://www.cpv.org.vn/cpv/Modules/News_English/News_Detail_E.aspx?CN_ID=401473&CO_ID=30113http://www.cpv.org.vn/cpv/Modules/News_English/News_Detail_E.aspx?CN_ID=401473&CO_ID=30113http://www.cpv.org.vn/cpv/Modules/News_English/News_Detail_E.aspx?CN_ID=401473&CO_ID=30113http://www.cpv.org.vn/cpv/Modules/News_English/News_Detail_E.aspx?CN_ID=401473&CO_ID=30113


30/63

THE Federal Budget 2005 06 initiated measures tocontain the trade deficit and can be deemed as a

major step in this direction. For any developingeconomy foreign exchange inflow is vital forsustainable development because it acts as a cushionagainst unanticipated and sudden shocks and defenceagainst unforeseen emergencies.

Pakistan has experienced the adverse impact of heavyborrowing when our external debts and foreignexchange liabilities jumped from $10 billion in 1980to $20 billion in 1990 and just before the freezing offoreign exchange, the liabilities touched $42 billionin May 1998.

During this period, Pakistans debts reached to anunmanageable level due to large deficits and animprudent use of borrowed resources.

During the financial year 2004 05 the trade deficitcontinued to surge steadily, reaching $5.51 billionduring July May 2005. Dr Ishrat Hussain, governor

of the State Bank of Pakistan, does not rule out thetotal deficit touching US$ 6.5 billion for the financialyear-2005.

Now, the situation seems to be fairly under control


31/63

but the government needs to keep an eye on its tradedeficit. It has to take into consideration the foreigntrade statistics for the last 11 years. (table)

In 2000 2001 the level of SBP reserves wereinadequate, which were barely $1 billion and onlyadequate for three weeks imports when reformsintroduced by the present government brought a

positive change. On September 10, 2001, the foreignexchange reserves touched $3.2 billion and by June23, 2005 they reached U$12.406 billion.

In any developing economy, imports are always onthe higher side than exports, creating a trade deficit.In order to meet that deficit, the developingeconomies use their Current Account Surplus or

alternatively borrow from various internationalagencies like the IMF, World Bank and othercountries with trade surplus.

In the past, Pakistan experienced disaster byexcessive dollar borrowing, which to a large extenthas been taken care of now.

Borrowed dollars come with conditionalities andimpinge on indepedent decision making. Imprudent

borrowing also has a deleterious effect on economicgrowth which keeps foreign investors away.


32/63

On the other hand, foreign exchange generated byexports has manifold benefits like it provides

employment, accelerates economic growth,strengthens local currency, improves the standard ofliving, attracts foreign investment and bringseconomic stability.

The budget is no doubt export friendly as it aims tobring changes by reducing duties on inputs of variousindustries like leather, textile, surgical goods, andcarpets. The government may further exempt allmachinery, spares, and other inputs of said industriesfrom all duties and taxes in order to make our exportscompetitive in a global market. There is a dire needto close foreign trade gap.

For this, we have to identify bottlenecks andhindrances in exports and introduce export culture.Exporters must be given facilities to remaincompetitive in an international market which is only

possible through the availability of export refinanceat low rates, waiver of all taxes on imports, and

availability of utilities at cheaper rates. Thegovernment needs to remove tariffs on imports forthe export industry.

Another area where the government needs to pay


33/63

attention is the re-structuring of the Export PromotionBureau (EPB). It needs to identify problems faced bythe export industry and suggest remedies to remove

such problems in consultation with the exporters.

The export cities for various items should beintroduced as is the case in countries like China andIndia. We should develop cities like Sialkot for sportsand surgical goods and leather garments, Gujrat forcutlery and furniture and Faisalabad for textiles.

This concept may be expanded to other parts of thecountry keeping in view the potential of the area.Borrowing should be confined to development

projects like construction of dams, ports,infrastructure, electricity generation and human

resource development. The concept of borrowingneeds to be restricted to projects which are feasibleand have the ability to pay back loans.

The country needs to attract foreign investment forsuch projects. And it cannot be done simply byrequesting foreigners to invest. For this, feasibility

studies of various potential projects should beprepared. These reports should be distributed gratis torelevant investors during the foreign visits of the

president and the prime minister. Pre-feasibilitystudies should be conducted to facilitate foreign


34/63

investors for mega projects like construction of dams,exploration of oil, gas, minerals and other vitalinfrastructure projects.

The other major source of foreign exchange is theremittances by our expatriates. We must encouragethese expatriates to transfer their all savings home.Currently, most of the remittances were madethrough proper banking channel rather than thesystem of hundi.

Privatization of our major units is, no doubt, the rightstep to meet our foreign exchange requirements. Butthe country can only benefit if priority is given toexport sector, which to some extent has beenaddressed in the 2005-6 federal budget, and export

refinance is made available at a cheaper rate.http://archives.dawn.com/2005/07/04/ebr16.htm

TE Taylor Enterprisewww.variation.comReturn to Change-Point Analyzer Main Page
http://archives.dawn.com/2005/07/04/ebr16.htmhttp://www.variation.com/cpa/index.htmlhttp://archives.dawn.com/2005/07/04/ebr16.htmhttp://www.variation.com/cpa/index.html


35/63

Change-Point Analysis:

A Powerful New Tool For Detecting

Dr. Wayne A. Taylor

Change-point analysis is a powerful new tool for determining whcapable of detecting subtle changes missed by control charts. Furtdetected by providing confidence levels and confidence intervals. W

point analysis is not a replacement for control charting. But, becau

further information, the two methods can be used in a complementadata, especially when dealing with large data sets, change-point anaA change-point analysis is more powerful, better characterizes the cis robust to outliers, is more flexible and is simpler to use. Thchange-point analysis and demonstrates its capabilities through a nu

Table of ContentsIntroduction

Example 1: US Trade Deficit DChange-Point Analysis

Procedure for Performing a Change-PoiDetecting Changes in the Variation and Othe

Handling OutliersExample 2: Complaint Data

Example 3: Part StrengthConclusionReferences
http://www.variation.com/wayne.htmlhttp://www.variation.com/cpa/tech/changepoint.html#Introductionhttp://www.variation.com/cpa/tech/changepoint.html#Example%201http://www.variation.com/cpa/tech/changepoint.html#Change-Point%20Analysishttp://www.variation.com/cpa/tech/changepoint.html#Procedurehttp://www.variation.com/cpa/tech/changepoint.html#Detecting%20Changes%20in%20the%20Variationhttp://www.variation.com/cpa/tech/changepoint.html#Handling%20Outliershttp://www.variation.com/cpa/tech/changepoint.html#Example%202http://www.variation.com/cpa/tech/changepoint.html#Example%203http://www.variation.com/cpa/tech/changepoint.html#Conclusionhttp://www.variation.com/cpa/tech/changepoint.html#Referenceshttp://www.variation.com/wayne.htmlhttp://www.variation.com/cpa/tech/changepoint.html#Introductionhttp://www.variation.com/cpa/tech/changepoint.html#Example%201http://www.variation.com/cpa/tech/changepoint.html#Change-Point%20Analysishttp://www.variation.com/cpa/tech/changepoint.html#Procedurehttp://www.variation.com/cpa/tech/changepoint.html#Detecting%20Changes%20in%20the%20Variationhttp://www.variation.com/cpa/tech/changepoint.html#Handling%20Outliershttp://www.variation.com/cpa/tech/changepoint.html#Example%202http://www.variation.com/cpa/tech/changepoint.html#Example%203http://www.variation.com/cpa/tech/changepoint.html#Conclusionhttp://www.variation.com/cpa/tech/changepoint.html#References


36/63

troduction

d a change occur? Did more than one change occur? When did the e changes occur? All these questions and more can be answered bange-point analysis is capable of detecting multiple changes. For eacluding a confidence level indicating the likelihood that a change occhen the change occurred. The change-point analysis procedure prformed on all types of time ordered data including attribute dathaved data such as particle counts and complaint data, and data wrform a change-point analysis and highlights its potential applications

aditionally, control charts are used to detect changes. The major difntrol charting is that control charts can be updated following the colint analysis can only be performed once all the data is collected. Conlated abnormal points and at detecting a major change quickly while

anges frequently missed by control charts. The two methods can be u

hen analyzing historical data, especially when such data sets are largntrol charting the data. One benefit of a change-point analysis is thaesult, each change detected is likely to be real. Control charts contro

ousands of data points, numerous points can exceed the control limhange-point analysis offers many other benefits as well. This artic

ethods of detecting change.

ample 1: US Trade Deficit Data


37/63

fore providing the details on how to perform a change-point analysiows US trade deficit data from 1987 to 1988 in billions of dollars. Th

Table 1: US Trade Deficits 1987-1988

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov

1987

10.7 13.0 11.4 11.5 12.5 14.1 14.8 14.1 12.6 16.0 11.7

1988

10.0 11.4 7.9 9.5 8.0 11.8 10.5 11.2 9.2 10.1 10.4

plot of this data is shown in Figure 1. Take a moment and determine trade deficit change during this period of time? How did it change?

Figure 1: Plot of US Trade Defici

e trade deficit appears to be lower in 1988 than in 1987. To confirmta can be constructed. The correct type of control chart to use for thdividuals control chart of the trade deficit data is shown in Figure 2. Te called control limits. They represent the maximum range that the


38/63

change has occurred. Points outside the control limits indicate atober 1987 point is above the upper control limit indicates that a chan

Figure 2: Individuals Control Chart of US T

ut what type of change occurred? Look at Figure 2 and decide hotober 1987 an abnormally high month? Did the trade deficit graduadid the trade deficit suddenly shift downward at the end of 1987 or

n be used to detect changes, interpreting these changes can still be di

the change, a change-point analysis can be performed.

hange-Point Analysis

ere are numerous approaches to performing a change-point analysplemented in Taylor (2000a). This software was used to perform theocedure are provided later. First we will look at the results of perults of a change-point analysis for the trade deficit data in Table 1.

Table 2: Results of Change-Point Analysis on U


39/63

e analysis detects two changes. The first change is estimated to hapresents the first month following the change. The second changovember 1987. Associated with each change is a confidence level in

e change actually occurred. The first change occurred with 91% con0% confidence. We are much more confident about the second chang

so associated with each change is a confidence interval for the time the change has been pinpointed. 95% confidence is used for all co

e first change occurred between May and July of 1987. With 95% covember 1987. The fact that the confidence interval for the first chast change cannot be as accurately pinpointed as the second change.

ble 2 also gives additional information about each change. The tableerage trade deficit was 11.82 billion dollars while after the first changves a level associated with each change. The level is an indication ofange is the first change detected and that which is most visibly appare

e detected on a second pass through the data. Any number of levanges found.

gure 3 shows a graphical presentation of the results of the chanpresented by the shifts in the shaded background. The shaded ba


40/63

ntain all the values based on the current model that two changes ogion, this model fully explains the variation in the data. The contrplayed in Figure 3.

Figure 3: Plot of US Trade Deficit Data Showing C

hile the control chart barely detected any change had occurred (one ptected two changes. It also provided additional details including cois example illustrates two of the benefits of a change-point analysi

stained changes and it better characterizes such changes. When useanges, a change-point analysis provides far more useful information tproach is to perform a change-point analysis. However, this does notta. The results of both approaches can be displayed in a simple plot li

ocedure for Performing a Change-Point Analysis

e procedure used by Taylor (2000a) for performing a change-point mulative sum charts (CUSUM) and bootstrapping to detect the nstruction of the CUSUM chart shown in Figure 4.


41/63

Figure 4: CUSUM Chart of US Trade

USUM charts are constructed by calculating and plotting a cumulativ

4 represent the 24 data points. From this, the cumulative sums S0, ms are calculated as follows:

1. First calculate the average

2. Start the cumulative sum at zero by setting S0 = 0.

3. Calculate the other cumulative sums by adding the differencethe previous sum,i = 1, 2, , 24.

r the Trade Deficit data:

S0 = 0


42/63

is is the series plotted in Figure 4. The cumulative sums are not they are the cumulative sums of differences between the values and the e cumulative sum always ends at zero (S24=0).

erpreting a CUSUM chart requires some practice. Suppose that durove the overall average. Most of the values added to the cumulatiadily increase. A segment of the CUSUM chart with an upward slopbe above the overall average. Likewise a segment with a downwardlues tend to be below the overall average. A sudden change in directchange in the average. Periods where the CUSUM chart follows

here the average did not change.

oking at Figure 4, the CUSUM chart takes a sudden turn in directioat around this time, the average shifted. Before the change in directiat initially the trade deficits are above the two-year average. After thwnwards indicating that the trade deficits are below the two-year aatively straight indicating no subsequent change occurred. However,could be broken into two separate line segments. This gives an incurred.

e CUSUM chart in Figure 4 appears to indicate that at least one aoblem with CUSUM charts is that they require considerable skill to pese changes took place? A confidence level can be determined fotstrap analysis. Before performing the bootstrap analysis, an esti


43/63

quired. One choice, which works well regardless of the distribution a

Sdiff = Smax - Smin where

Smax =

Smin =

r the trade deficit data, Smin = -0.69583, Smax = 17.04583 making Sagnitude of the change has been selected, the bootstrap analysis crformed by:

1. Generate a bootstrap sample of 24 units, denoted X01, X0

2, 24 values. This is called sampling without replacement.

2. Based on the bootstrap sample, calculate the bootstrap CUSU

3. Calculate the maximum, minimum and difference of the boS0diff.

4. Determine whether the bootstrap difference S0diff is less than th

e idea behind bootstrapping is that the bootstrap samples represent re behavior of the CUSUM if no change has occurred. By performingn estimate how much Sdiff would vary if no change took place. Youculated from the data is its original order to determine if this value i


44/63

change occurred. Figure 5 shows the CUSUM chart in Figure 4 ofe CUSUM charts from 5 different bootstrap samples. The bootstrap an the CUSUM of the data in its original order. This leads one to susp

Figure 5: CUSUM Charts of Data In Original Order

bootstrap analysis consists of performing a large number of bootstrar which S0diff is less than Sdiff. Let N be the number of bootstrap samp

otstraps for which S0diff < Sdiff. Then the confidence level that a changlows:

Confidence Level = %

pically 90%, or 95% confidence is required before one states that a

e five bootstraps shown in Figure 5, the values of S0diff are 7.0, 14.91these values are below Sdiff = 17.74167. Figure 6 shows a histogram

ut of 1,000 bootstraps, 995 had S0diff < Sdiff. This gives a confidence le


45/63

Confidence Level = = 99.5%

is is strong evidence that a change did in fact occur.

Figure 6: Histogram of S0difffor 1000 Boot

eally, rather than bootstrapping, one would like to determine the

orderings of the data. However, this is generally not feasible. The tral number of possible reorderings is 24! = 6.2 1023. This is more same bootstrap analysis randomly selected 1000 of these possible retribution of S0diff. A better estimate can be obtained by increasing th00 bootstraps is sufficient for most purposes. Repeating the above nfidence levels: 99.6%, 99.2%, 99.3%, 99.2%, 99.4%, 99.7%, 99

alysis performed in this article are based on 1,000 bootstrap samples.

otstrapping results in a distribution free approach with only a singleucture. Both control charting and change-point analysis are basedpresent the data in time order. The mean-shift model can be written a


46/63

Xi =i +i

herei is the average at time i. Generallyi =i-1 except for a smaints. i is the random error associated with the i-th value. It is aeans of zero. Taylor (2000b) provides a procedure for detecting a propriate for a change-point analysis and control charting include auces. Control charts also make the assumption that thei are idenditional assumptions as well, for example, that thei are normally di

nce a change has been detected, an estimate of when the change occe CUSUM estimator. Let m be such that:

|Sm| =

is the point furthest from zero in the CUSUM chart. The point

curred. The point m+1 estimates the first point after the change. Fint on the CUSUM chart is the November 1987 (S11). The best tween November 1987 and December 1987.

second estimator of when the change occurred is the mean squarefined as:

where and


47/63

e MSE error estimator is based on the idea of splitting the data inimating the average of each segment, and then seeing how well thelue of m that minimizes MSE(m) is the best estimator of the last poi

+1 estimates the first point after the change. For the trade deficit da. The best estimate of when the change occurred is between Octobonth earlier than predicted by the CUSUM estimator. In Table 2, thmes of the changes and the months reported are the first months follow

nce a change has been detected, the data can be broken into two segd the analysis repeated for each segment. For each additional signgments in two. In this manner multiple changes can be detected. Flits the data into two segments, January 1987 through October 19888. Repeating the analysis on each segment results in the detecticurred between May 1987 and June 1987. This changed occurredely to have occurred as the first change. However, there is still strongure 7 shows the CUSUM chart from Figure 4 with the significant ch

at the slope of the CUSUM chart changed twice. These correspond to

hile the above procedure generally works well, it can be improved. the above procedure as a set of candidate change-points. Once this

eir confidence levels are reestimated. A backward elimination procat no longer test significant. When a point is eliminated, the surrouth their significance levels. This reduces the rate of false detections.


48/63

Figure 7: CUSUM Chart of US Trade Def

Significant Changes Shown in t

ble 2 also gives confidence intervals for the time of the change. Mervals are provided in Efron and Tibshirani (1993). Change-point anginally stated in Page (1955, 1957). Hinkley (1971) and Pettitt (arts. Hinkley and Schechtman, (1987) suggest the application of bombines these two approaches in a novel fashion. It also adds an iterabe detected. All the above references are restricted to the problem of

tecting Changes in the Variation and Other Characteristics

e change-point analysis procedure can be applied to any characteristie trade deficit example, it was applied to individual values. It cviations, ranges, defect levels, counts and many other types of data. alysis can be used to detect a change in the autocorrelation structure o

r the trade deficit data we would also like to determine if the variationcause we have only a single observation per month so we cannot calc


49/63

se, it is still possible to analyze the variation by performing the changnsecutive points. Let X1, X2, , X24 represent the 24 data points. Fr, D12 are calculated as follows:

Di = | X2i X2i-1 | for i=1,2,,12

gure 8 shows a CUSUM chart of the trade deficit variation (differens chart indicates that no significant change in the trade deficit variatio

Figure 8: CUSUM Chart of US Trade De

is important to only use differences that do not have points in commve the point X2 in common and are therefore correlated. This violquired by a change-point analysis. While the difference between cood estimate of the short-term variation in the process, at least some oshifts or changes in the average. Such shifts can create outlier or ex

e next section, the change-point analysis procedure is robust to such o

andling Outliers


50/63

hat if the data contains an outlier? As an example of what can happenanged to 25. Figure 9 shows an individuals control chart for this mod

ange-point analysis shown in the background. The outlier is clearly eriation in the data making it more difficult to detect a change. Despittects the second change, which was the change with the greatest effec

Figure 9: Individuals Chart of US Trade Defic

e change-point analysis procedure is reasonably robust to such outlalyzing the ranks of the values instead of the values themselves. Fints. The largest value is given a rank of 24, the second largest vallue is given a rank of 1. Analyzing the ranks instead of the actual vaffected by outliers. Table 3 and Figure 10 show the results of p

nks. The same results are obtained as before without the outlier includ

Table 3: Results of Change-Point Analysis on Ranks W


51/63

e outlier in Figure 10 falls outside the shaded region indicating it stilalysis and should be treated as an isolated abnormal point. This exaould be supplemented with an individuals chart (or other control cnormal points. If such points exist, the change-point analysis shou

nks are analyzed, the change-point analysis can detect smaller sustaintliers.

Figure 10: Plot of US Trade Deficit Data With Ou

ample 2: Complaint Data

further illustrate the benefits of a change-point analysis, two moample involves a set of complaint data. The complaint rate had inuse. Figure 11 shows the number of complaints for the last eight mo


52/63

ose to zero as in the first month. It appears that the complaint rate std then gradually started to decline. A team was put together to ideonths to cause the increase in complaints.

Figure 11: Plot of Number of Complain

mplaints are recorded by lot number. There were approximately 30nuary, lots 31-60 correspond to February 2 and so on. The lots arntrol chart of the number of complaints by lot number. Only an uppe bounded below by zero. There are 24 points above the control limiore than one change occur? One way of answering this question ntrol limits each time a change is detected. When did the first chang

mits is lot 75 (middle of March). Is this the best guess of when the termine how many changes occurred and when each change occurred


53/63

Figure 12: Control Chart of Number of Co

get a better understanding of the number and timing of the changee results are shown in Table 4. The level 1 change occurred around evel 2 change occurred around lot 42 (middle of February). The num6 during the first change and then up to 27.7 following the second mes the magnitude of the first change. Complaints are still being receonths. The other four changes all occur toward the end of the data sed may be caused by the fact that complaints are still being received.

Table 4: Results of Change-Point Analysis o


54/63

gure 13 shows the CUSUM chart of complaints with the changes anges represent the onset of the problem. The relative straightness o

anges indicates the changes were fairly sudden.

Figure 13: CUSUM of Number of C

e change-point analysis has yielded several important clues. The per a 3-month period, seems to be the result of two changes, the secon

curred around lot 74. The confidence interval indicates that with 95tween lots 72 and 86. This corresponds to 2-week period of time. Fus to cause a smaller number of complaints. Since the process has 5 lir the complaint problem was piloted on one line around lot 42 (firsther 4 lines around lot 74 (second change). This led to the identificatie clues and proved to be the cause of the problem.

is example illustrates several of the benefits of change-point analalysis provided much more detailed information about the number onthly summary in Figure 11 or the control chart in Figure 12. Detey to solving the problem. Second, the change point analysis determin


55/63

. This change was missed by the control chart, which didnt have a cond change). Change-point analysis can detect changes missed balysis is more flexible. The same analysis worked on the complaint

fferent type of control chart was required. An individuals control cgure 2 while a u-chart was used in Figure 12 for the complaint datrform a change-point analysis. Finally, a change-point analysis is sis with multiple changes. Using a control chart requires calculatio

der to detect further changes. This is complicated by the need to esthat data should be used to calculate the new control limits. Chanocess, again making it easier to train others to do the analysis. Wharly a change-point analysis is preferable to control charting.

ample 3: Part Strength

e next example involves the online collection of data on part strength

ntrol chart. Control limits were calculated using historical data. Eveults immediately added to the control chart. Figure 14 shows the rint plotted falls outside the control limits. At this point the operatortermine the cause of the problem as well as take whatever interim od product. Generally, this investigation focuses on what occurred oblem. A change-point analysis of the first 25 points might provide ferator to determine the root cause.


56/63

Figure 14: Control Chart of Part S

ble 5 shows the results of performing a change-point analysis on

alysis indicates two changes in fact occurred. The first change occucurred between the points 22 and 23. This is several points before

mits.

Table 5: Results of Change-Point Analysis o

n investigation into what events occurred around points 13 and 23 det

roll changes of the raw material. This lead to a change in the roll chbility. Figure 15 summarizes the results of the change-point analysis


57/63

Figure 15: Plot of Part Strengths Showing Results o

is third example illustrates how change-point analysis can complealysis is a powerful problem investigation tool. It can detect chanovide better estimates as to the timing of the changes. It just mightution of a problem that would otherwise go unsolved.

onclusion

hange-point analysis is not a replacement for control charting. Contrta is collected and plotted on an hour-by-hour basis such as on a proalysis can be use to complement the control charts. First, a changmulated data once a week to try and detect more subtle changes misntrol chart detects a change or problem, a change-point analysis ca

ming and nature of the change as part of the problem solving process.

hen performing a one-time analysis on historical data, a change-poinpecially when you are dealing with large data sets. Sometimes, suchincrease in complaints. This could trigger the review of the compla


58/63

hen the problem started. Other times such analyses are done to gaight decide to review accident data to determine if there are anyriodically reviewed to gain the big picture. Each year as part o

rformed on inventory turns to see whether there has been an improvange-point analysis has numerous advantages over a control chart inc

1. It is more powerful at detecting smaller sustained changes. Tindividuals control chart.

2. It better characterizes such changes including detection oconfidence levels, and providing confidence intervals for the time

3. It reduces the number of false detections by controlling the chchange detected there is high confidence that it is real. Control cwhen used with large data sets, can produce numerous false detec

4. It is robust to outliers and can be made even more robust by ranks.

5. It is more flexible. The same procedure works for all types values, counts, averages and standard deviations. Control charttype of data. Further, a change-point analysis can be performed complaint data, which do not follow any of the traditional cont

numerous outliers. Only one assumption is made which can easi

6. It is simpler to use and interpret, especially for large daoccurred. On a control chart, a single change can cause numeroDetermining the number of changes and exactly when the chan


59/63

way of accomplishing this with control charts is to establish neorder to be able to detect further changes. This is complicatechanges to determine what data should be used to establish n

automates this difficult process.

spite its numerous advantages, change-point analysis has two shortnormal points. A change-point analysis should be supplemented wints are of concern. Box and Luceo (1997) demonstrate that Shewlated abnormal points while CUSUM charts are optimal at detectinth both types of changes, both procedures can be used to complement

cond, the bootstrapping approach will not produce identical results eae random selection of the bootstrap samples. This last shortcoming icreasing the number of bootstraps results in increasingly more precisamples are recommended as the minimum number. For small data seperformed in a reasonable amount of time.

hange-point analysis represents a powerful new tool that complementntrol charts when analyzing historical data. However, possibly the gxibility and simplicity. When implemented in a software package, thribute versus variables data, normality assumptions, distributions, anart. This can result in software that is easier to use than are most uld automatically verify the assumption of independent errors and c

ftware could notify the user of the problem and guide them on whaalysis an extremely safe approach, even for a novice. Hopefully, thisalysis of time ordered data a much more common practice.


60/63

ferences

x, George and Luceo, Alberto (1997), Statistical Control by Mon

wYork.

ron, Bradley and Tibshirani, Robert (1993), An introduction to the Bo

nkley, D. V. (1971), Inference about the change-point from cumulat

nkley, David and Schechtman, Edna (1987), Conditional bootsometrika, 74 1, 85-93.

ge, E. S. (1955), A test for a change in a parameter occurring at an u

ge, E. S. (1957), On problems in which a change in parameter occ8-252.

ttitt, A. N. (1980), A simple cumulative sum type statistic for servations, Biometrika, 67 1, 79-84.

ylor, Wayne (2000a), Change-Point Analyzer 2.0 shareware programeb: http://www.variation.com/cpa

ylor, Wayne (2000b), A Pattern Test for Distinguishing Betweebmitted to Journal of Quality Technologies.

heeler, Donald (1993), Understanding Variation The Key to Mnnessee.


61/63

ey Words: Change-Point Analysis, Control Chart, Individuals Chart,

tation: Taylor, Wayne A. (2000), "Change-Point Analysis: A PowEB: http://www.variation.com/cpa/tech/changepoint.html.

Copyright 1997-2011 Taylor Enterprises, Inc.Last modified: March 15, 2011
http://www.variation.com/cpa/tech/pattern.htmlhttp://www.variation.com/cpa/tech/pattern.html


62/63

Trade deficit hit $6.8 billion this month.Nhu Ngoc | vir.com.vn | Sep 24, 2011 17:00 pm

Share

Vietnamis positively controlling its trade deficit. TheGeneral Statistics Office (GSO) today reported thetrade deficit for September was estimated at $6.8

billion, equivalent to 9.8 per cent of total exportturnover. This level is much lower than 15 per centcap set by the government.

Control of the trade deficit ispositive because of export expansion in line withmeasures limiting import operations, GSO said inthe report.Export turnover this month is estimated to be $8.3

billion, up 33.6 per cent year-on-year, raising total

turnover since January to September to $70 billion.Meanwhile, import turnover is estimated at $9.3 percent, up 31 per cent year-on-year. During the pastnine months, import turnover was $76.9 billion.
http://www.vir.com.vn/news/business/trade-deficit-hit-$68-billion-this-month.htmlmailto:?subject=Trade%20deficit%20hit%20$6.8%20billion%20this%20month.&body=From%20vir.com.vn%0A%0ATrade%20deficit%20hit%20$6.8%20billion%20this%20month.%0A%0AVietnamis%20positively%20controlling%20its%20trade%20deficit.%20The%20General%20Statistics%20Office%20(GSO)%20today%20reported%20the%20trade%20deficit%20for%20September%20was%20estimated%20at%20$6.8%20billion,%20equivalent%20to%209.8%20per%20cent%20of%20total%20export%20turnover.%20This%20level%20is%20much%20lower%20than%2015%20per%20cent%20cap%20set%20by%20the%20government.%0A%0AYou%20can%20click%20or%20copy%20and%20paste%20this%20link%20into%20your%20browser:%20%0Ahttp://www.vir.com.vn/news/business/trade-deficit-hit-$68-billion-this-month.htmlhttp://digg.com/submit?phase=2&url=http://www.vir.com.vn/news/business/trade-deficit-hit-$68-billion-this-month.html&title=Trade%20deficit%20hit%20$6.8%20billion%20this%20month.http://twitter.com/home?status=Trade%20deficit%20hit%20$6.8%20billion%20this%20month.%20%7C%20http://www.vir.com.vn/news/business/trade-deficit-hit-$68-billion-this-month.htmlhttp://www.vir.com.vn/news/business/trade-deficit-hit-$68-billion-this-month.html


63/63

The low trade deficit will ease pressure on the dongand maintain Vietnams balance of payment surplusin 2011. According to State Bank, the balance of

payment may be $2.5-$4.5 billion this year.

role of state bank

Documents