What Made the Financial Crisis Systemic?, Cato Policy Analysis No. 693

8/2/2019 What Made the Financial Crisis Systemic?, Cato Policy Analysis No. 693

1/28

Executive Summary

The current narrative regarding the 2008 sys-temic financial system collapse is that numer-ous seemingly unrelated events occurred in un-regulated or underregulated markets, requiringwidespread bailouts of actors across the financialspectrum, from mortgage borrowers to investorsin money market funds. The Financial Crisis In-quiry Commission, created by the U.S. Congressto investigate the causes of the crisis, promotes

this politically convenient narrative, and the 2010Dodd-Frank Act operationalizes it by complet-ing the progressive extension of federal protec-tion and regulation of banking and finance thatbegan in the 1930s so that it now covers virtuallyall financial activities, including hedge funds andproprietary trading. The Dodd-Frank Act furthercharges the newly created Financial Stability Over-

sight Council, made up of politicians, bureau-crats, and university professors, with preventing asubsequent systemic crisis.

Markets can become unbalanced, but they gen-erally correct themselves before crises become sys-temic. Because of the accumulation of past politi-cal reactions to previous crises, this did not occurwith the most recent crisis. Public enterprises hadcrowded out private enterprises, and public pro-

tection and the associated prudential regulationhad trumped market discipline. Prudential regu-lation created moral hazard and public protectioninvited mission regulation, both of which under-mined prudential regulation itself. This eventuallyled to systemic failure. Politicians are responsiblefor both regulatory incompetence and mission-induced laxity.

What Made the Financial Crisis Systemic?by Patric H. Hendershott and Kevin Villani

No. 693 March 6, 2012

Patric Hendershott holds a chair in real estate economics and finance at the University of Aberdeen, Scotland. Previously he held tenure positions at Purdue University and Ohio State University. Hendershott is also aresearch associate at the National Bureau of Economic Research. Kevin Villani is a former chief economist forthe U.S. Department of Housing and Urban Development and Freddie Mac.


2/28

2

It wasnt the lackof regulatory

authority,but rather its

pervasivenessand widespreadfailure, that not

only allowedbut caused the

subprime lendingdebacle.

Introduction and Overview

The 2008 global financial collapse ema-nated from the U.S. subprime lending bubble.Economists generally resort to mass psychol-

ogy to explain how the bubble could inflateto such an extent, assuming that borrowerand lender behavior was irrational. The FraudEnforcement and Recovery Act of 2009 cre-ated the Financial Crisis Inquiry Commis-sion (FCIC) to investigate why the financialcrisis became globally systemic. The FCIC wascharged with conducting a comprehensive ex-amination of 22 specific and substantive areasof inquiry relating to various and seeminglyunrelated hypotheses advanced primarily bypoliticians, business executives, and univer-

sity professors. The final FCIC report (2011)found varying degrees of merit for all of thesehypotheses, blaming the financial crisis on aconfluence of generally independent events,such as recklessness of the financial industryand the abject failures of policymakers andregulators1 to regulate the essentially deregu-lated or never regulated parts of the mortgageand deposit markets. This justified the 2,300-page Dodd-Frank Acts regulatory approach.

The basic difference between the UnitedStates and other market economies since 1975

has been that U.S. mortgage and related mar-kets relied more on federally sponsored andregulated enterprises that were more perva-sivelythats not to say appropriatelyregu-lated. The conventional narrative is that theunregulated private-label securitizers (PLS)like Countrywide and Bear Stearns fundedthe subprime bubble, subsequently draggingdown the giant government-created mortgagefinancers Fannie Mae and Freddie Mac withthem. But all of the markets contributing tothe crisis were already subject to regulation

in one way or another. The FCIC Reportsconclusions succeed in diffusing the politicalresponsibility for making the crisis systemicand hence fail as a guide to avoiding futuresystemic crises.

The specific regulatory failures necessaryfor Fannie Mae and Freddie Mac to become soheavily entwined in the subprime bubble were

allowing them to bypass the primarymortgage insurers;

allowing them extreme leverage; and requiring them to finance at least half

of the subprime market.

The specific regulatory failures necessaryfor the private-label securitizers to becomelikewise heavily entwined in the subprimebubble were

the Securities and Exchange Commis-sion (SEC) designation of approvedcredit rating agencies without the neces-sary supervision;

the use of the credit rating agency des-ignations in risk-based capital require-

ments; the failure of bank regulators to pre-

vent the deterioration of underwritingguidelines, regulatory arbitrage, offbalance sheet funding, and the rise ofthe shadow banking market; and

woefully inadequate Securities and Ex-change Commission capital regulationsfor investment banks and account-ing rules that allowed the accelerationof income and delayed recognition ofexpense.

These will all be explained below. However, allof these failures should not obscure a moregeneral problem: federal regulators did notunderstand and mitigate systemic risk.

Regulation and Intervention

Public protection of private enterprises cre-ates a moral hazard wherein private enter-prises are more willing to take risk when they

know that they will be rescued if they get intotrouble. As a result, protected enterprises willtake excessive risks. Virtually all of the behaviorthat created the subprime lending debacle canbe explained by incentive distortions, mostlymoral hazard, and none of this behavior wasnew or irrational. It wasnt the lack of regula-tory authority, but rather its pervasiveness and


3/28

3

Market disciplincant be said tohave failedduring thesubprimelending bubble,because it didntexist.

widespread failure, that not only allowed butcaused the subprime lending debacle. Add-ing to this, purely political mission regula-tionspecifically, the government-embracedmission of expanding homeownership rates

regardless of riskpushed the bubble to sys-temic proportions.Politicians are responsible for regula-

tory oversight, but that oversight failed be-cause of both incompetence and incentiveconflict.2 Politicians and their regulatorsconsistently failed to understand how regu-lation would undermine and replace, ratherthan complement, market discipline. Marketdiscipline cant be said to have failed dur-ing the subprime lending bubble, becauseit didnt exist; market forces had been re-

placed by regulatory oversight. Only marketspeculation operated largely outside of thisregulatory regime, but politicians and theirregulators have always tried to limit and evencriminalize the stabilizing activity of specu-lators shorting the market.

Banking deregulation has often beenblamed for causing the financial crisis, butits unclear why that would be the case. Theso-called Reagan-era banking deregula-tion (which was actually signed into law byPresident Jimmy Carter after being passed

by a Democrat-controlled Congress) was thephase-out of deposit interest rate ceilings be-ginning in 1980, and that had nothing to dowith housing finance. The next piece of bank-ing deregulation was the 1994 eliminationof bank branching restrictions, which waspassed by a Democrat-controlled Congressand signed into law in 1994 by Democraticpresident Bill Clinton, and which, again, hadnothing to do with housing finance. The 1999banking reform, passed by a Republican Con-gress and signed by President Clinton, elimi-

nated the Glass-Steagall Acts forced separa-tion of investment and deposit banking. Butthat didnt seem to contribute to the financialcrisis: the institutions at the heart of the cri-sisBear Sterns, Lehman Brothers, Ameri-quest, Countrywide, AIG, and so onwerenot universal banks (though many of thoseinstitutions were later merged into universal

banks so as to stabilize them). These reformsall removed political distortions and strength-ened the financial system.

The credit allocation goals of mission regu-lation, in contrast, directly conflicted with pru-

dential regulation. In the case of Fannie Maeand Freddie Mac, political mission regulationand corruption explains most of the regula-tory failure. In the case of the banks, missionregulation also likely explains why regulatorsdidnt raise the credit standards for their loansor increase capital requirements.

How did the U.S. mortgage markets be-come so much more politicized than thoseof other market economies? The numerousfederally sponsored enterprises that insure de-posits and mortgages, or lend against or make

a market in mortgages, all trace their roots tothe response of federal politicians to the GreatDepression.3

Government-provided deposit insuranceintroduced in 1933 to prevent bank runs andprotect the payments mechanismarguablyhad a positive effect over its first four decadesof existence. But repressive government pro-hibitions on branching and paying interesteventually spawned the shadow banking sys-tem, a financial system outside the formalregulated system of banks and thrifts. Moral

hazard grew in the banking system as depositprotection became comprehensive, regulatorspromoted bank consolidation over competi-tion between banks, and large banks and otherfinancial firms became too-big-to-failthatis, so important to the U.S. economy that gov-ernment would rescue them if they got intotrouble. Moral hazard eventually affected theentire shadow banking system as well, whichsubsequently played a role in funding the sub-prime lending debacle.

Intervention in HousingMortgage market enterprises such as Fan-

nie Mae and Freddie Mac were introduced tomaintain liquidity of mortgage lenders andstimulate housing construction and jobs. In-stead of acting as traditional mortgage compa-nies that raise money from investors and thenuse that money to extend loans to individual


4/28

4

With eachsuccessive

economic crisis,politicians and

bureaucrats havedoubled downon regulation

to protect thefiction that

markets, ratherthan bureaucrats

and politicians,had failed.

homebuyers, these government enterprisespurchased from banks the mortgage contractsthat the banks had already made with home-buyers. This reduced the banks risk of defaultand provided them with cash to use for more

home loans, while the enterprises received thelong-term payments from the homebuyers.Originally, the enterprises financed the pur-

chases with money from Congress augmentedby the mortgage payments from existing bor-rowers. But in time, the enterprises began re-selling the mortgages, as large bundles of di-versified mortgages, to independent investorswho then receive the mortgage payments.

Ultimately, the two largest enterprises, Fan-nie and Freddie, were partly spun off from gov-ernment, becoming government-sponsored

enterprises (GSEs). Though they still had apublic mission to boost the housing marketand they were under special government over-sight, they also had private investors and couldhold mortgages and attempt to profit fromdoing so.

For approximately the next four decades,these enterprises did no obvious harm. Homemortgage lending to households financedlargely by private mutually chartered savingsand loan, savings bank, and life insurancecompany intermediariesthe backbone of the

U.S. housing finance system for over a centuryraised the U.S. homeownership rate from alevel of about 45 percent in 1945 to 55 percentby the early 1950s as returning veterans gotVeterans Administration loans, and then to 65percent by 1975, prior to the GSE era, wherethe rate remained for approximately the nextquarter century.

Restrictions on branch banking and de-posit rates caused regional capital shortagesfor banks by the 1970s, owing to huge demo-graphic migrations. As a result, banks came

to rely more heavily on wholesale fundingsources like Fannie and Freddie. Mortgagebonds of the type commonly used in otherdeveloped market economies date back to the1800s in the United States. But U.S. capitalmarkets were fragmented by multiple state ju-risdictions and overlapping state and federalregulation. Providing GSEs with regulatory

exemptions was politically more expedientthan streamlining securities laws or removingbranch banking restrictions. But politicianshave consistently muddled the distinction be-tween liquidity and marketability, as well

as the distinction between market-makingand trading, which can refer to both broker/dealer activities and speculation. The conse-quence was often that GSE policies justifiedas promoting liquidity often promoted onlymarketability and frequently only speculationMoreover, the subsequent privatization ofFannie and Freddie created a huge incentivedistortion of public risk for private profit, andtheir regulatory preferences that conveyedagency status allowed them to become publichousing banks that crowded out the private

market while inviting a public mission for po-litical cover.Politicians werent motivated to dramati-

cally reform the U.S. financial system, but at-tempts to use regulation to favor both borrow-ers and savers did disturb the systems balanceWith each successive economic crisis, politi-cians and bureaucrats have, not surprisinglydoubled down on regulation so as to protectthe fiction that markets, rather than bureau-crats and politicians, had failed and to main-tain and expand the federal regulation and

programs this fiction justified.During the Depression, the Franklin D

Roosevelt administration and bank regulatorswere much more concerned with the moralhazard consequences of regulation than Dodd-Frank is today, even though the New Dealersattempts to mitigate it with arms-length fed-eral sponsorship of insurers ultimately failedDodd-Frank will fail to achieve its goal ofmitigating systemic risk because it doesnt dis-tinguish between the deposit/money marketsthat Congress should haveand now must

comprehensively and appropriately regulate,and the investment markets, including mort-gage markets, that Congress didnt and canteffectively regulate.

The laws and regulations resulting in thesubprime lending debacle reflected the uniquehistorical evolution of federal intervention inU.S. deposit and mortgage markets. That evo-


5/28

5

In sharp contrasto the UnitedStates, bankruns werent

a problem inCanada, wherebanks branchedfreely and were,as a consequencwell diversified.

lution will be described in the rest of this paper.The next section traces the early origins of fed-eral regulations and enterprises from the in-troduction of the Federal Reserve to the land-mark Housing and Urban Development Act

of 1968. The following section describes thedevelopment of the GSEs and the subsequentexpansion of U.S. secondary markets in hous-ing finance through the end of the last centu-ry, resulting in substantial replacement of theprivate market. Following that is an explana-tion of how moral hazard created by the SECand deposit insurance further underminedmortgage market discipline in the 1990s. Thatis followed by a description of the origins anddevelopment of the subprime lending debacle.The paper concludes with policy implications

and recommendations.

The Great Depression to 1968Economists are still debating the relative

causes of the Great Depression, but the centralfindings are indisputable. It was not caused byinstability in unregulated financial markets.Unregulated private financial markets in theUnited States had previously produced regu-lar but smaller self-correcting financial crisesthat were blamedunfairlyfor the GreatDepression.4 Rather, politically induced dis-

tortions turned the recession into a depres-sion and made it great. For example, labormarket distortions relating to the Wagner Actand Davis-Bacon kept wage rates as much as40 percent above market clearing levels, andSmoot Hawley tariffs reduced internationaltrade by more than half.

The seeds of the Great Depression wereplanted two decades before the crash, withthe creation of the Federal Reserve and U.S.embrace of modern monetary policy. Whenthe Federal Reserve Act was passed in 1913

establishing the Federal Reserve, the Ameri-can Banker hypothesized that from this timeforward the financial disorders which havemarked the history of the past generation willpass away forever.5 But soon thereafter, theFed was accused of fueling the asset bubble ofthe 1920s. Housing production had boomed,fueled by the Feds easy credit policies, and

then fell by over 80 percent, from 753,000units in 1928 to 134,000 in 1932. According toSimon Johnson and James Kwak:

Not only did the Federal Reserves

System encourage excessive risk takingby bankers, the safety net, it turnedout, had gaping holes that could not befixed in the intense pressure of a crisis.The result was the Great Depression.6

But the Fed did more. The current Fedchairman Ben Bernanke publicly apologizedto Milton Friedman on behalf of the Fed,agreeing that the centralized monetary au-thority was the cause of the systemic deflationthat perpetuated the Depression.7

The Depression era spawned numerousfederal interventions in the financial system,including deposit insurance, mortgage insur-ance, and mortgage discount lending andmarket-making facilities.

Liquidity of Deposit InstitutionsMany banks had liquidity problems early

in the Depression and manybut not neces-sarily the same bankswere technically insol-vent. It was virtually impossible for depositorsto discern between insolvency (when a bank

has failed as an ongoing concern) and illiquid-ity (when a bank is temporarily short on cash).Hence, at any sign of trouble, depositorsrushed to the bank to withdraw their fundsbefore the bank ran out of money. These bankrushes turned many illiquid banks insolvent.

It fell to the Fed to make that distinction,providing sufficient liquidity to solvent banksonly. In sharp contrast to the United States,bank runs werent a problem in Canada, wherebanks branched freely and were, as a conse-quence, well diversified.8 Moreover, Canada

lacked deposit insurance, hence the lack ofinsurance couldnt have been a cause of bankruns in the United States.9

Contrary to the belief underlying Glass-Steagall, that universal banks were responsiblefor the Depression by duping depositors intomaking questionable investments, the largeU.S. money-center banks that engaged in both


6/28

6

The fear ofcentralizing

power in WallStreet money-

center banksmade branch

banking acrossstate linespolitically

problematic.

commercial and investment banking werediversified and generally remained both liq-uid and solvent. Combining commercial andinvestment bankingi.e., underwriting newsecurities sold through their sales force and

making a secondary market in these and othersecurities as a broker (matching buyers andsellers) or as a dealer (maintaining a modestinventory for sale), depending on the type ofsecurity and marketwas common in marketeconomies due to the potential synergies.

But Sen. Carter Glass and Rep. Henry B.Steagall had long been concerned with thepotential conflicts of interest between sellingsecurities and taking deposits and the con-centration of power on Wall Street. Congresscreated the Pecora Commission ostensibly to

determine the causes of the Great Depression,but really the commission was created specifi-cally to cast these universal banks as villainseven though there was no evidence that riskyinvestment banking activity was a contribut-ing cause of the financial collapse or GreatDepression.10 The negative press coverage ofthe bankers during the commission hearingswas sufficient to pass the Glass-Steagall Actof 1933, separating commercial banking anddeposit-taking from securities underwritingand market-making activity.

The fear of centralizing power in Wall Streetmoney-center banks made branch bankingacross state lines politically problematic. Thatwas unfortunate for U.S. depositors becausebranching and, hence, geographic diversifica-tion would have buttressed the banks, makingthem better able to withstand the droughtsand other local shocks that pushed many lo-cal banks into illiquidity and insolvency dur-ing the Depression. Glass-Steagall supportersdid not understand that and, not wanting towaste a crisis or the political momentum of the

Pecora Commission, they used the legislationto extend the prohibition against branchingacross state lines to federally chartered banks.

Of course, Glass-Steagall only exacerbatedthe problem with bank runs. In response,Congress established the Federal Deposit In-surance Corporation (FDIC). Both FDR andbank regulators had previously opposed pub-

lic deposit insurance because of concerns withmoral hazard, but the best they could do atthis point was to try to mitigate this risk bylimiting the insurance to small depositors andthe federal role to sponsoring a self-funding

enterprise.

11

Savings and loans (S&Ls) hadnot been subjected to the same runs as smallbanks because their deposits were not callableon demand, so they refused to join the FDICand pay the large insurance premiums neededto fund commercial bank losses. But they ac-quiesced two years later to their own federallysponsored insurer, the Federal Savings andLoan Insurance Corporation (FSLIC).

Whether it was deposit insurance thatstopped the bank runs is debatable, but bothfederally sponsored deposit insurers remained

solvent during their first 50 years. The strat-egy of mitigating moral hazard by limitingthe federal role to sponsorship was, how-ever, an immediate failure. As early as 1933,when FDR reopened the banks, the marketsperceived the federal backing of deposit in-surance as complete.12 Bank capital levels fellsteadily from over 16 percent of assets whendeposit insurance was first introduced to only5.5 percent by 1945, where they stayed for overfour decades before falling further in responseto risk-based capital requirements and off

balance sheet financing.

Housing and Mortgage Market DistressThe mostly mutual savings and loan

model of mortgage finance, patterned afterthe 200-year-old British system, had workedfairly well in the United States to this pointUnder this model, borrowers repaid prin-cipal by contributing monthly to a sinkingfundthe common practice of the timeef-fectively amortizing the mortgage principaland avoiding a balloon payment at maturity.

Payments into a sinking fund are used to re-tire the outstanding debt which funded theoriginal mortgage. Loans were rolled overevery 5 to 10 years, at which time they werere-priced to the current market interest rateSo long as the borrower was current and thelender remained solvent, rollover was relative-ly automatic. But no system could have sur-


7/28

7

Politicians triedto stimulatehousing demandby reducingdown payments

vived the systemic credit default debacle ofthe Great Depressionduring which three infour home borrowers defaultedunscathed.

S&Ls that faced liquidity problems hadrelied on their lines of credit with commer-

cial banks to keep them liquid, but these linesmostly disappeared when banks failed.13 TheS&Ls couldnt access credit through the Fed-eral Reserves discount window (a prohibitionthat would continue until 1989). Arguablyto address this problem, Hubert Hooversadministration established an independentfederally sponsored enterprise, the FederalHome Loan Bank (FHLB) System, in 1932 toprovide liquidity directly to S&Ls. The FHLBdid this by discounting home mortgages, thatis, providing loan advances against mortgage

collateral at less than par value (called a hair-cut) and with full recourse to the borrowinginstitution, which meant that the FHLB couldseize the S&Ls assets if the S&L failed to re-pay its loan. The FHLB advance program wasanalogous to the Fed discount window, with-out the discount windows stigma of signal-ing distress and with a more liberal collateralrequirement and significantly longer terms,both of which were at least partly intended topromote homebuilding and hence construc-tion jobs. The FHLB System had authority to

borrow up to $215 million from the U.S. Trea-sury in emergencies, but it generally relied oncapital market access for funding.

The demand for housing units fell muchfaster than supply because of doubling upconversion of single-unit dwellings to multiplesmaller units, for exampleso that the vacancyrate rose by over 60 percent from 8 percent to13 percent during a period of plummetinghousing starts.14 This caused housing pricesto fall by far more than the general monetarydeflation. Then as now, falling house prices

caused considerable economic pain, as home-owners found themselves owing more moneyon their houses than the houses were worth.

Politicians intervened to mitigate the im-pact of financial system distress on borrowersas well as to stimulate housing productionand jobs. The Homeowners Loan Corporation(HOLC) was established in 1933 and placed

under the FHLB System to implement thegovernments Depression-era forbearance pro-gram for distressed but salvageable borrowers,giving them a reprieve on their payments. Onecomponent was to refinance creditworthy

borrowers with a fixed-rate mortgage that re-placed the rollover provision with a long-termloan and the sinking fund with a more porta-ble amortization schedule. The HOLC markeddown the nominal loan balance to only 70 per-cent of the new lower house price, with lend-ers taking the loss, and offered an interest rateslightly above the Treasury borrowing cost.Still, 20 percent of the borrowers subsequentlyre-defaulted. The re-pricing of mortgages atrollover reduced the payment to reflect fallinginterest rates. The HOLC was profitable for

the U.S. government because it required bor-rowers to pay a fixed rate on their loans, butthe HOLC financed the program with debt atinterest rates that declined, creating the profitmargin. The HOLC was liquidated in 1951.15

Politicians next tried to stimulate hous-ing demand by reducing down payments.Congress enacted the National Housing Actof 1934 that established the Federal HousingAdministration (FHA) as an independent fed-erally sponsored mutual mortgage insurancefund authorized to insure only the long-term

(up to 20 years), fully amortizing fixed-ratemortgage (FRM) with a maximum loan-to- value (LTV) ratio of 80 percent. With suchinsurance, it was hoped that banks wouldbe more inclined to make loans to prospec-tive homebuyers. Private monoline mortgageinsurers had previously insured loans witha high LTV, but the Great Depression was asystemic risk to which they were inherently vulnerable due to their lack of diversifica-tion among other lines of insurance, and theyall failed as a consequence. As the National

Housing Acts stated intent was to promotehomebuilding and construction jobs,16 activ-ity was limited to new housing, and the initiallimit on the loan amount was more than threetimes the median house price. But portfoliolenders saw no need for default insurance onnewly originated loans with a 20 percent cashdown payment, so FHA activity was minimal.


8/28

8

Speculativetrading in GSE

mortgage-backedsecurities gave

rise to GSEagency status,

which droveout private

competition andincreased moral

hazard.

The National Housing Act also gave FHAthe authority to establish private nationalmortgage exchanges to make a market inthese FHA-insured FRMs and thereby pro-mote their use. But there were no private tak-

ers because lenders originated mortgages asa portfolio investment and S&Ls could nowdiscount mortgages at the FHLB. Neverthe-less Congress established the ReconstructionFinance Corporation (RFC) Mortgage Com-pany in 1935 with $10 million in capital to buyand sellFHA loans that financed new residen-tial construction, and later in February 1938Congress amended the National Housing Actto have the FHA create the National MortgageAssociation of Washington (later changed tothe Federal National Mortgage Association,

or Fannie Mae) to replace the RFC MortgageCompany. The restriction to finance only newconstruction was removed and the loan limitwas reduced to the median house price in1938.17 The RFC Mortgage Company was lat-er absorbed into the Reconstruction FinanceCorporation in 1947.

The private lending industry stronglyopposed the creation of what its membersrecognized as a government housing bankbecause Fannie Maes two main special as-sistance functions were financed directly

by the U.S. Treasury, which also provided anemergency liquidity backstop similar to thatof the Federal Home Loan Bank System. Toassuage these concerns, the third facilityex-plicitly limited by charter to a pure secondarymarket broker/dealer function that requiredselling in equal proportion to buyingwaslimited to FHA-insured loans, which at thetime had an insignificant market share. Inaddition, this dealer inventory had to befunded with private corporate debt, issuedonly with prior Treasury approval.

The conventional wisdom is that thesefederal interventions restored the hous-ing market. While there was little merit ina separate liquidity facility without centralbank access, the FHLB discount facility didno harm. It is unlikely that the FHLB, FHA,or Fannie Mae stimulated manyif anyconstruction jobs. And Fannie Mae did not

build a secondary market. By 1966, after al-most three decades in operation, the second-ary market dealer portfolio was estimatedat about $2.5 billion, reflecting the activitiesof the post-war Veterans Administration

whose loans were made eligible for purchaseafter the war, as well as the FHA inventory,and there was little if any turnover.

One later ex postrationalization is that theFHAs long-term amortizing FRM solved theproblem of rollover balloon mortgages. Butthis justification doesnt fit the facts. Lackof amortization wasnt a problem, becausemost mortgages had sinking funds, and therate adjustment at rollover wasnt a prob-lem, because rates were adjusting downwardduring the Depression, making fixed-rate

mortgages more expensive at the time thanrollover mortgages. Only the rollover provi-sion was potentially problematic. In theorysome current borrowers could have facedproblems rolling over their loans, but theborrowers were mutual owners as well andhence had an equal say in making rolloverpolicies, and liquid lenders had every incen-tive to roll over a loan for a current borrowerrather than foreclose on an unsalable houseEven in the event of lender insolvency, depos-itors and other creditors were better off with

a paying loan, so the problem, to the extent itexisted, was with the way in which insolventinstitutions were liquidated.

The Development of theGSE Mortgage Capital

Market System

The influence of the numerous Depres-sion-era federally sponsored enterprises, in-

consequential up to this point, grew in sig-nificance with the 1968 Housing Act thatshaped the GSEs and mortgage securitizationduring the last quarter of the 20th century.Speculative trading in GSE mortgage-backedsecurities (MBS) gave rise to GSE agency sta-tus, which drove out private competition andincreased moral hazard.


9/28

9

The motivationfor the newGinnie Maemortgage-backesecurities wasto bypass allconflicting stateand federal lawsand regulations.

Early Origins of GSE MortgageSecuritization

One of the most consequential aspectsof the 1968 Housing Act was the so-calledprivatization of Fannie Mae. This was driv-

en by nothing more than myopic political ex-pediency. President Lyndon Johnson wantedto wage war in Vietnam while simultaneouslyinstituting his Great Society domestic agen-da, but the federal budget could not accom-modate both. This ratcheted up the normalpressure for budget accounting gimmicks.One gimmick would be to make Fannie Maedisappear from the budget.

Politicians had previously tried to get Fan-nie Mae mortgages held as a consequence ofthe public housing bank functions off the

budget with the 1964 Housing Act. The actprovided authority for Fannie Mae to issueparticipation certificates, an early form ofMBS, on pools of mortgages by treating thesecuritization as a sale of assets rather than afinancing, but Fannie continued to borrowfrom the Treasury instead.18 The 1968 Hous-ing Act established the Government NationalMortgage Association, or Ginnie Mae, tomanage and liquidate the Treasury-financedFHA/VA-insured Fannie Mae portfolio ofabout $4 billion as quickly as feasible, getting

them off the budget to reduce the reporteddeficit.

Arguably, the newly chartered Ginnie Maeshould have been given Fannies $2.5 billioncorporately financed dealer portfolio to liq-uidate as well, but that was already off-budgetand would have to be carried on-budget untilit was sold, having the exact opposite effect ofwhat the act was trying to accomplish. It waspolitically more expedient to simply give thecompany to the mortgage bankersholdersof nominal stock issued in return for a fee

for using the secondary market facility. Thisstock had little value and became virtuallyworthless when Ginnie Mae introduced MBSbacked by pools of FHA-insured and VA-guar-anteed loansan activity that went well be-yond its management and liquidating char-tershortly after the newly privatized FannieMae rejected the securitization concept. In

retrospect, had the Ginnie Mae security beenanticipated and authorized by the 1968 act,this corporately funded broker-dealer facilitywould likely have been liquidated as well.

The motivation for the new Ginnie Mae

MBS was to bypass all conflicting state andfederal laws and regulations that would haverequired separate security registrations inall 50 states for each offering. This bypasswas easy because its securities were treatedas federal government issues with a federalpreemption of state and local laws. But Gin-nie Mae was not exempt from federal law; inparticular, it was not exempt from the taxa-tion of mortgage revenues as profits beforedistribution of interest to security holders. Itfound an old but limited grantor trust stat-

ute to avoid corporate tax because it was bylaw passive, that is, it passed through all of thecash flow from the mortgages, rather than ac-tively manage its income.19 The IRS allowedthe guaranteed advancement of the typicallydelayed FHA insurance reimbursements inthe event of borrower default, but the grant-or trust vehicle prohibited Ginnie Mae fromguaranteeing credit risk.

Soon after Ginnie Mae introduced theparticipation certificates, the newly priva-tized Fannie Mae became totally superfluous.

Rather than liquidate it, the mortgage bank-ing industry lobbied for new legislation pro-viding Fannie with conventional loan author-ity, but the charter restriction limiting activityto a broker-dealer function remained. Eventhough S&Ls had no use for such a secondarymarket facility, they did not want to lack foran authority given to mortgage bankers. Theyhad previously lobbied for and got the FederalHome Loan Mortgage Corporation (FreddieMac) to deal in conventional mortgages, with-out the explicit charter restrictions but with

the explicit promise that if this dealer func-tion was a commercial failure the corporationwould be liquidated. But the mortgage bank-ers immediately turned the Fannie Mae sec-ondary market facility into a portfolio-lendinggovernment-sponsored housing bank servingtheir interests in spite of the unambiguouscharter limitations to the contrary. Likewise,


10/28

10

Insuranceunderwritingwas based on

the premise thatborrowers wouldpay if they could.

when Freddie Macs broker dealer operationproved to be a commercial failure, the cor-poration reneged on its pledge to liquidateand followed the Fannie Mae lead as a feder-ally sponsored quasi-publicsubsequently

privatizedhousing bank engaged in fundingmortgages.

Mortgage Insurance, Private and PublicHousing banks are notoriously political

and prone to actuarial (albeit often opaque)failure. The saving grace of the U.S. systemwas reliance on mortgage insuranceFHA forGinnie Mae securities and private mortgageinsurers (PMIs) for conventional loanstomitigate credit risk.

PMIs, bankrupted by the Great Depression

but reincarnated in the 1950s, became the pri-mary credit risk filter for conventional loanswith less than 20 percent down payments forthe originate-to-sell system of funding, just asFHA was for qualifying loans in its market seg-ment.20 The down payment requirements forinsured conventional loans remained relative-ly constant between 5 percent and 10 percent;the FHA minimum down payment was low-ered from the initial 20 percent to 5 percentin 1950 and 3 percent in 1961. (Beginning in1995, the minimum down payment percent-

age increased for loans above $50,000).21The FHA and the PMIs both assure risks

with ex ante risk-mitigation measures to mini-mize the potential for moral hazard associatedwith insuring borrowers with little or no equityat stake and insure remaining risks throughdiversification. The fundamental principleof insurance is that the remaining credit riskcan be diversified and actuarially priced be-cause of the uncorrelated nature of defaultrisk among the individual loans in a pool. TheFHA covered a lenders entire loss, while the

VA and PMI insured the top 20 and 25 percentrespectively, but these differences were gener-ally insignificant as loans not fraudulentlyunderwritten rarely resulted in losses greaterthan 2025 percent. Further, both the FHA/VA and the PMIs maintained a local under-writing presence and rigorous underwritingguidelines, and the PMIs attempted to avoid

correlated risks such as widespread economicdownturns and falling house prices.

Insurance underwriting was based on thepremise that borrowers would pay if theycould, which they had always done before

to protect their down payment investmentother assets, and credit reputation as well asin response to societal expectations. But un-like market-oriented systems in other coun-tries, recourse to the borrowers income andother assets in the event of default is out-lawed in 27 states and not enforced in manyof the rest. Hence, when borrower equityevaporates due to minimal down paymentfalling house prices, or both, insurance in theUnited States relies on the strength of the fu-ture borrowing penalty and societal expecta-

tions to limit default.Moral hazard is inherent in the originate-to-sell model, but the FHA and Ginnie Maeminimized it in three ways: first, the FHAmaintained local underwriting offices; sec-ond, Ginnie Mae required an excessive servic-ing fee, postponing some of the originationprofit to the end of the loan, which was lostin the event of default due to foreclosure ex-pense borne by the servicer;22 and third, Gin-nie Mae had full recourse, which cross-collat-eralized all securitizations, thereby putting an

MBS originators entire profitable loan servic-ing business and capital at risk for a failure toperform on any individual pool. Freddie Macwas historically more protected against thismoral hazard by dealing with better capital-ized portfolio lenders rather than mortgagebrokers and bankers.

The 1968 act designated the U.S. Depart-ment of Housing and Urban Development(HUD) the regulator for the FHA, the Veterans Administration regulated the VA, and statinsurance commissioners regulated the PMIs

As HUD was a new agency with a social mis-sion without other prudential regulatory re-sponsibilities, conflicts with the FHA shouldhave been expected. Soon enough, conflictsarose when the Carter administration FHAReport recommended that the agency serveunderserved markets. That recommenda-tion came despite the fact that the prior early


11/28

11

The capitalmarket accessrationale becamobsolete by

1990 with theelimination ofbank branchingrestrictions.

1970s failure of the FHAs special risk insur-ance fund was enough to convince FHA actu-aries that insurance was not an appropriate vehicle for delivering subsidies to high-riskborrowers because adverse selection would

prevent actuarially sound pricing. (Adverseselection occurs when lenders attempt to ac-cept more risk, with higher losses paid forby charging all borrowers a higher rate pre-mium.) Unfortunately, the better credit bor-rowers find cheaper loans elsewhere, leavinginadequate premiums to cover losses on theinsured loans. This cross-subsidization of riskonly works if the insurance is mandatory orthe government has a monopoly. The GinnieMae participation certificates protected theFHA insurance fund from adverse selection

by giving FHA a monopoly pricing advantageand virtually a 100 percent market share in itsqualifying loan market by charging only 6 ba-sis points annually for the agency status con-ferred on Ginnie Mae MBS.

Increased borrower equity caused by priceinflation in the housing market during the1970s protected the insurers, but credit riskbecame a concern in the 1980s when pricesstagnated. The PMIs raised premiums numer-ous times, but in order to prevent what wouldhave otherwise been overwhelming adverse

selection, they also significantly tightened un-derwriting guidelines to exclude several typesof loans: investor loans, loans with cash outrefinancing, loans with deep buydowns, andloans in regions with a weak economy due toa systemic risk factor, such as the oil patch.23Even these steps didnt save all the PMIs fromadverse selection, but the industry survivedthe decade and the private investors losses dueto PMI failure were minimal.

In contrast, the FHA did not mitigate ad-verse selection. They did not exclude investor

loans, they eased qualification standards, andhalf of their loans endorsed in 198889 wereto those with LTV above 95 percent or to in-vestorsup from a third in 198283. The FHAwas arguably technically insolvent and clear-ly not actuarially sound,24 so the ironicallynamed Cranston-Gonzalez National Afford-able Housing Act of 1990 restored actuarial

soundness by charging new borrowers moreto subsidize existing borrowers. While theFHA did worse than private mortgage insur-ers, it did much better than public insurers inother countries have typically done, probably

reflecting the FHAs reliance on private loanservicers and investors to evaluate risk. Unlikedebt financing, securitization passes on the in-terest rate risk to investors so GSE securitiza-tion exposed Fannie and Freddie only to theresidual pool insurance risk. As a result, thepublics risk exposure to securitization in the1970s through the 1980s was minimal. GSEsecuritization benefited mortgage borrowersby alleviating regional credit shortages due tobranching restrictions, increasing the overallavailability of funds. This reduced rates some-

what in credit-short areas while raising themin credit-surplus areas. Borrowers of conform-ing conventional loans paid a quarter percent-age point less than borrowers of otherwisecomparable jumbo loans,25 but whether thisreflected the nonagency cost of capital marketaccess or the ability of GSEs to borrow at ratesslightly above that of the U.S. Treasury is un-clear. In any event, the capital market access ra-tionale became obsolete by 1990 with the elim-ination of bank branching restrictions and thelegal and regulatory obstacles to private capi-

tal market access, including new Real EstateMortgage Investment Conduit (REMIC) legis-lation of 1986 that enabled private-label MBS.

Speculative Trading, Agency Status andWall Street Greed

The conventional ex post rationalizationfor securitization is that it converted inher-ently illiquid mortgage loans into liquidMBS. The reality is that GSE MBS becamea trading vehicle for speculationmuch of itotherwise illegaland was complicit in con-

verting investment banks into highly lever-aged hedge funds issuing bank-like depositsin the shadow banking system.

The enduring political fiction that securi-tization and secondary market trading couldconvert intrinsically illiquid mortgages intoliquid securities continues to be the primarypolitical and economic GSE rationale. The


12/28

12

Wall Streetloved trading

GSE mortgaged-backed securities

because theywere risky and

thus inherentlyilliquid.

financial press has always used the termsliquiditythe ability to sell quickly forcash at par valueinterchangeably with theterm marketabilitythe ability to sell witha low bid-ask spread at whatever investors

think the security is worth. Cash managersbuy government securities to invest cashbalances and sell (or issue short-term debtsuch as commercial paper) when they needcash. Marketability, as measured by the bid-ask spread, does improve with the volumeof trades, but this had never before been aconcern to long-term investors in bonds ormortgages and hence debt rarely traded.

Unlike government bonds or highly ratedcorporate bonds, the cash flows of pre-pay-able and, at the time, assumable fixed-rate

home mortgages were extremely difficult topredict. Hence there was a reason to tradeMBS based on different prepayment and as-sumption views and changes in projectionsover time. As interest rates became more vol-atile, there were both premium and discountpools to trade. Wall Street loved trading GSEMBS because they were risky and thus inher-ently illiquid. There is nothing inherentlywrong with speculative trading, which argu-ably made the options price more efficient.But the implicit option premium is taken

into income currently with no regard for theresidual risk, which is called tail risk be-cause it often comes at the end.26 The tailscould be quite long, and the perception ofhigher yields created by the options premi-um encourages speculation, especially if thetail risk will be borne by others.

Speculating in tail risk was particularly at-tractive to money managers, to whom suchspeculation was typically prohibited. Becausethey trade so frequently, speculators are oftencalled traders or more specifically propri-

etary traders. The political fiction that GSEsecurities were liquid because they tradedessentially allowed regulated investors other-wise authorized only to invest in liquid gov-ernment securities to engage in speculativeGSE MBS trades, and the investment banks allcame up with competing strategies as to howto speculate. These trading strategies generally

allowed traders to write out of the moneyoptions using GSE and derivative securities inan opaque way, treating the entire trading rev-enue in the form of option premium as profit.By the 1980s these strategies were marketed

mostly to S&Ls as part of their go-for-brokesurvival strategy, often camouflaged as hedg-esthat is, using instruments and strategiesto reduce portfolio risk. After the S&Ls wentbroke, the bankers turned to cash managers ofstate and local governments (see box).

Investment banks historically had two trad-ing (market-making) desks: government andcorporate. As GSE MBS issuance and specula-tion-driven trading volume skyrocketed in themid-1970s, it was the lawyers at and advisersto the Wall Street trading firms that made a

judgment that Ginnie Mae, Fannie Mae, andFreddie Mac securities could all be traded onthe government desk. The judgment that suchsecurities would be backed by the governmentin the event of defaultin spite of the specificdisclosures to the contraryreflected their fed-eral sponsorship with the regulatory and taxexemptions of a public entity.27 As volumessoared, the market itself became too big tofail and, as with deposit insurance, there wasno denying the implicit government backing,removing any pretense of market discipline.

Whereas speculation had historically beenreserved to individuals and hedge funds,speculating in GSE MBSs enticed investmentbanks to do so for their own account, estab-lishing proprietary trading desks, essentiallyin-house hedge funds. Hedge fund managersgenerally keep 20 to 25 percent of the returnover a benchmark as a management bonuswith no downside risk, creating an incentiveconflict to load up on tail risk. Hedge fundmanagers are also expected to contribute theirown personal funds to mitigate moral hazard.

Similarly, proprietary trading began at invest-ment banks when they were all partnershipswhere partners waited a lifetime to enjoy thefruits of their greed.

But proprietary traders eventually demand-ed and received annual cash bonuses just likehedge fund managers, often on unrealizedprofits. This undermined the partnership


13/28

13

Repos wereprovided withGSE mortgage-backed securitieas collateralon essentially

the same basisas for liquidgovernmentsecuritiesbased on theirmarketability.

structure of traditional investment banking.When Salomon Brothersthe premier fixed-

income trading housecashed out by sellingthe firm to the commodities trading firmPhibro in 1979, the lesson was not lost on thepartners of other firms. Virtually all the WallStreet firms had sold out or converted to stockby 2000 (with only Goldman Sachs retainingsome partner equity). Because all the firmswere now owned by shareholders and manyan investment bank CEO now rose from the(proprietary) trading rankslimiting or de-ferring bonuses wasnt a competitive strategy.

By the end of the last century, the largest

investment banks had portfolios approachingor exceeding a trillion dollars of mostly hedgefundproprietary trading accountbalancesand some private equity fund assets. Thesebanks were able to fund their portfolios byborrowing extremely short-term using a vari-ety of instruments such as commercial paperand overnight repos. (A repurchase agreement,

or repo, is essentially an overnight loan at aslight discount to the trading or market value

of the collateral.) Repos were provided withGSE MBS as collateral on essentially the samebasis as for liquid government securities basedon their marketability. In fact, their entirefunding structure assumed that marketabledealer inventory and internal hedge fund as-sets were liquid, ignoring the distinctionbetween immutable liquidity and ephemeralmarketability. The too-big-to-fail investmentbanks had essentially had commercial bank,hedge fund, and private equity fund assets andnear-money liabilities without bank regula-

tion or a liquidity backstop.The separation of commercial from invest-

ment banking required by the Glass-SteagallAct was gradually phased out through regula-tory forbearance in the 1980s and 1990s andeliminated with the repeal of the act signed byPresident Clinton in 1999. The FCIC reportschapter 4, Expansion of Banking Activities:

Speculating in GSE Securities:The Case of Orange County

In 1994 Orange County, California, one of the highest family-income counties in

history, was forced to declare bankruptcy. Its investment manager, Robert Citron,had collected all the cash accounts that numerous local governments held in theirlocal bank accounts to meet the public payroll and deposited them at Merrill Lynch.He then leveraged them with re-purchase agreements, and invested them directly insupposedly liquid risk-free GSE securities. However, they were actually derivative se-curities employed in risk-controlled arbitrage strategies largely designed by Mer-rill Lynch. Citron was considered a hero for years as the higher earnings from thisspeculation allowed local politicians to keep taxes down. He was essentially playingthe yield curve by investing in long-term securities as well as speculating by earningexcess quoted yield that reflected not higher expected returns but rather the op-tion premium for prepayment risk and other derivative trading strategies, a formof tail risk. When GSE MBS prices subsequently plummeted as interest rates rose,

past gains were wiped out, bankrupting Orange County and severely wounding SanDiego County finances.Orange County blamed its demise on Wall Street greed. Citron and other cash

managers obviously had no business transferring taxpayer bank accounts to theshadow banking system and then speculating with options trading. Whose respon-sibility was it to stop them? The answer in this case was that politicians who provid-ed oversight took responsibility when the bets paid off, and Wall Street investmentbanks took the blame when they didnt, a political lesson that didnt go unnoticed.


14/28

14

By the early 1980sthe GSEs were

both the biggestcustomers of

and competitorswith the Wall

Street investmenthouses.

Shatter of Glass Steagall, follows the conven-tional wisdom that this deregulation was amajor cause of the financial crisis. But the cri-sis had nothing to do with commercial banksunderwriting or broker/dealer activitiesthe

original separation required by Glass-Steagall.Funding hedge funds with near money wassomething altogether different.

Agency Status Crowds Out Private LendersOnce privatized, the GSEs viewed both

portfolio lenders and private-label securitiz-ers as competitors and, consequently, spentlavishly on politicians to maintain their legaland regulatory advantages. Ultimately theyprevailed against both.

Thrift institutionsS&Ls and savings

bankshad been the backbone of the mort-gage lending industry for the century prior toGSE securitization. Deposit rate regulation in-troduced for demand deposits at commercialbanks in the 1950s (Regulation Q) was extend-ed to S&Ls in the 1960s. Thrifts were grantedregulatory authority to engage in traditionalbanking functions by offering money marketaccounts in the 1970s, and they had a competi-tive advantage of being allowed to pay interest.But rates were still capped, and when marketrates rose, Merrill Lynch pioneered the cash

management account that paid market inter-est rates while providing check-cashing privi-leges, extending the payments mechanism fur-ther to investment banks. Money market fundsbecame a big industry when interest rates roseand deposit institutions were disintermedi-ated by deposit outflows. Advances from theFHLB helped maintain thrift liquidity, but de-posit rate ceilings channeled a significant shareof mortgage funding to the GSEs.

The mortgage banking industry volun-tarily had Fannie Mae borrow short-term debt

to get the highest price for their loans. As a re-sult of the sharp run-up in short-term interestrates in the late 1970s and early 1980s, FannieMae became technically insolvent just as thethrift industry did. By the end of the century,the thrift industry was largely gone, replacedby the GSEs and a few large thrifts that wereindistinguishable from commercial banks.

Meanwhile, the shadow banking industryspawned by deposit rate ceilings had grown insize to rival the commercial banking industry,and was a source of funds for mortgages.

With thrift competitors largely out of the

way, the GSEs focused on eliminating com-petition from private-label securitizers. Bythe early 1980s the GSEs were both the big-gest customers of and competitors with theWall Street investment houses. The GSEs hadmonopoly power like the U.S. Treasury andcould negotiate low underwriting fees, butWall Street firms earned lucrative underwrit-ing fees on sheer volume. Similarly, the bid-ask spread was much narrower for GSE thancorporate securities, but this could be madeup with speculation-driving trading volume

The underwriting spreads were more lucrativefor private-label securitizers, but the invest-ment banks were leery of losing a sure thingwith the GSEs for the potential profits fromprivate securitization.

In 1983 First Boston purchased a firm withthe technology to originate mortgages straightinto private-label securities. The GSEs strong-ly opposed extending their regulatory exemp-tions to private-label competitors with all theirpolitical mightFannie Mae in particular wasknown to be quite vindictive regarding poten-

tial Wall Street competition. But at the end ofthe day, agency status still trumped private-label securitization, especially if exploited withrisky strategies. While there is some merit toGSE assertions that their subsequent interestrate risks were hedged, the trillion-dollar arbi-trage strategy of using debt to fund their buy-ing of MBS was obviously risky. The only issueis the extent to which this arbitrage reflected aspeculative interest rate or options play.

Prudential Regulation,Market Discipline, and

Mission Regulation

During the 1990s, private-label securitiz-ers found ways to exploit the credit rating de-pendent risk-based regulations and the moral


15/28

15

That politicaldistortion of theindustry wouldeventually butinevitably leadto systemicfailure had beenwidely predictedin numerous

congressionallymandated studycommissionsspanning fourdecades.

The 1989 S&L Crisis

The conventional explanation for why the S&L industry collapsed in 1989 is muchthe same as that for subprime lenders several decades later: the product of deregulation

and greed. Of course moral hazard was a problem, as the industry was highly leveraged,and many economists argued that this caused the industry to go for broke as a result.Some economists even argued that the cause of the systemic failure was that they pur-posely tried to go broke, with management looting the already failed and sure-to-be-closed thrifts.28 Charles Keating of Lincoln Savings and Loan in Phoenix was the posterboy of the decade and jailed for looting as a result of the failure of his institution.But the government spent an unprecedented amount on looting prosecutions in thesearch for scapegoats, with little to show for the effort.29

By the early 1980s, virtually the entire S&L industry was underwater, and the rea-son was entirely political. Federally chartered S&Ls were forced into an interest ratematurity mismatch by politicians who refused to allow them to invest in anythingother than fixed-rate mortgages. State-chartered S&Ls had issued mostly adjustable

rate loans, minimizing the maturity mismatch. But William Proxmire, chairman ofthe Senate Banking, Housing and Urban Affairs Committee refused to allow feder-ally charted S&Ls to do so.

Of course the industry went for broke, but this, too, was at least partially caused byexplicit federal policy. The deposit insurance coverage maximum was increased from$40,000 to $100,000 and the Garn St.Germaine Act of 1982 gave thrifts new powersto invest in both risky commercial real estate and high-yield bond investments, afterthey were already technically insolvent. And utilize them the S&Ls did: the share ofS&L assets in home mortgages plunged from 73 percent in 1981 to 57 percent in1985. This shift and the 1981 tax act with its incredibly generous tax depreciationallowances caused massive overbuilding of commercial real estate and large losses onthe new thrift investments.30

That political distortion of the industry would eventually but inevitably lead tosystemic failure had been widely predicted in numerous congressionally mandatedstudy commissions spanning four decades, and so it did.31 Two political forces cul-minated in the passage of the Financial Institutions Regulatory Reform and Enforce-ment Act (FIRREA) of 1989 that phased out thrifts. First, bank regulators at the Fedand FDIC didnt like the idea of S&Ls with money-like liabilities regulated by theFHLB, which also had a housing support mission. The regulators essentially won abureaucratic turf war, as the passage of FIRREA eliminated the prudential regula-tion duties of the FHLB and consolidated the Office of Thrift Supervision (OTS)into Treasury (and, in 2011, the Controller of the Currency within Treasury). Second,with both the S&L industry and Fannie Mae technically insolvent in the late 1980s,Congress could only let one increase market share, at the expense of the other, to

grow out of its problem. It chose the politically too-big-to-fail Fannie Mae, for whichlawmakers could not dodge accountability, by signaling markets that the govern-ment would stand behind Fannies debts no matter how great the losses. Politiciansnevertheless claimed they didnt see this failure coming and blamed greedy thriftowners (in spite of the fact that most thrift institutions were nonprofit mutual in-stitutions) and their managers (of which courts subsequently found virtually no evi-dence of guilt).


16/28

16

The first bigsubprime

mortgage lendingboom occurred in

the mid-1990s.

hazard of deposit insurance expanded acrossthe financial markets, eliminating market dis-cipline.32 Mission regulation was expanded tojustify public protection of banks and GSEs,but ended up competing with and ultimately

undermining prudential regulation.

Trumping Market DisciplineBeginning in 1975 with the SEC adoption

of Nationally Recognized Statistical RatingOrganizations (NRSROs), the risk regulatorsbegan moving away from what was prudentand toward greater reliance on risk assess-ments by credit ratings agencies Moodys,Standard & Poors, and Fitchthe only ratersso recognized at the time.33 In the late 1970s,Michael Milken at the investment firm Drexel

Burnham Lambert started issuing junk-ratedbonds and replaced bank loans with bond fi-nancing. These events converted publishedopinions into a regulatory sanctioned ap-proval and created a two-tiered new-issue mar-ket of investment grade and junk.

This SEC adoption had an enormous im-pact on the way markets evaluated risk. Whilethe credit ratings agencies had been ratingcorporate bonds since the early part of the lastcentury, by 1970 only about 10 percent of cor-porate bonds were publicly registered and rat-

ed, with the rest privately placed, mostly withlife insurance companies. Moreover, corporatebond opinions were fairly transparent, as in-vestors could generally check up on them byreviewing a simple transparent balance sheetof a big company. But bank regulators becameincreasingly reliant on the ratingssubse-quently embedding them in Basel I risk-basedcapital rules governing commercial bank capi-taland GSE regulators followed suit.34

By the end of this era, virtually all securi-ties were publicly placed and rated. Investors

in investment grade securities ceased doingindependent due diligence, depending en-tirely on the judgment of the ratings agencies.In addition, investment grade securities werepriced on the basis of their ratings and theirregulatory status at a slight discount to Trea-sury securities, reflecting only the difference inthe rating. In addition, prior to the SEC des-

ignation, the rating agencies didnt rate weakcompanies or those without a sufficient trackrecord as below investment grade (BBB) andas a result, such companies typically relied onbanks for funding.

The result was the development of privatelabel securitization. Investors were now able tocreate pools of mortgages, bundle them into aseries of both investment grade and below in-vestment grade tranches. They could sell virtu-ally all the parts, which allowed the investorsto treat the securitization as a sale of assets forregulatory accounting purposes. The creditrisk evaluation of these securities was directlydelegated to the ratings agencies, and pricingfollowed the ratings.

The first big subprime mortgage lending

boom occurred in the mid-1990s. Loans wereprovided to people with generally bad credit butsubstantial down payments, initially 20 percentto 30 percent. The loans typically were not eli-gible for sale to the GSEs because of the borrow-ers low credit scores, but the loans didnt requireprivate mortgage insurance due to the highdown payments. Originators chose private se-curitization over internal bank funding becausethe rating agencies dramatically underestimatedthe default risk and loss severity, enabling exces-sive amounts to be financed in the investment-

grade tranches with only a small retained eq-uity strip. In addition, following present valueGenerally Accepted Accounting Principles dic-tated by the SEC, these firms booked large cur-rent profits based on projected lifetime revenueof the residual interests discounted at a relativelylow interest rate, again specified by the SEC.

So banks spun off their mortgage bankingdivisions as finance companies or free-stand-ing mortgage banks that went public basedon these reported profits. The SEC-inflatedreported profits allowed lenders to raise both

equity and debt in the high-yield (junk) bondmarket relatively cheaply to fund residual in-terests of only 12 percent of the pool, therebyachieving about 1001 leverage. Some of theselenders converted to Real Estate InvestmentTrust (REIT) status to avoid paying taxes on theinvestment earnings of the retained strips (andpotentially on the operating profits as well).35


17/28

17

While prudentiaregulation by thdeposit insurerswas technically

limited tocommercialbanks andthrifts, the morahazard extendedwell beyondtheir portfolioactivities to thosthey funded.

Not all investors were fooled by SEC ac-counting. Many recognized the Ponzi scheme,profiting both on bubble speculation and sub-sequent shorting of the market. Within a fewyears, realized credit losses proved the fallacy

of SEC-dictated financial disclosure and virtu-ally all the publicly traded companies filed forbankruptcy.

While prudential regulation by the depositinsurers was technically limited to commercialbanks and thrifts, the moral hazard extendedwell beyond their portfolio activities to thosethey funded, particularly to hedge funds and

investment banks (which had their own inter-nal hedge funds). The failure of Long-TermCapital Management (LTCM) and its sub-sequent bailout in 1998 proved the systemicscope of moral hazard by the end of the last

century. If a hedge fund failed, it could bringdown at least one too-big-to-fail investmentbank. If just one investment bank failed, itcould bring down one or several too-big-to-fail commercial banks, and their failure couldbring down the global financial system. Sosays the current political conventional wis-dom, at any rate.

Long-Term Capital Management:The Extension of Moral Hazard

Long-Term Capital Management was a hedge fund established by formerSalomon Brothers proprietary trader John Meriwether to take advantage of whathe perceived to be small anomalies in the prices of fixed-income securities. Todo this, LTCM employed massive leverage, exceeding 1,000:1, mostly with bank-funded repos. Despite having two Nobel Prize winners in finance on its board andits renown for massive modeling capacity, LTCM simply played the tail risk andfailed spectacularly in 1998.

Timothy Geithner, then-CEO of the New York Federal Reserve Board (and cur-rent secretary of the treasury) organized a bailout by arm-twisting the too-big-to-fail banks and investment banks to socialize the loss by committing $300 million

each of their firms capital to the bailout. Lehman Brothers, the weak sister at thetime, was only asked to pony up $100 million, and Bear Stearns, which was notexposed, notably refused. The CEOs, some of whom may have been conflicted asthey had personally invested significant sums in LTCM, all agreed to contributetheir independent stockholders capital.

In the wake of the LTCM crisis, banks significantly raised haircuts and stoppedfinancing much of their repo activity. This forced a widespread de-leveraging byhedge funds that depressed asset prices and slowed asset-backed securitizationuntil significant de-leveraging had occurred. The Feds role in the aftermath wasto flood the market with liquidity and keep interest rates extremely low.

The LTCM bailout and subsequent comments about the Feds role in the af-

termath of a bubble created the perception of a Greenspan put, a bailout in theevent of trouble that many believe exacerbated subsequent moral hazard.36 Essen-tially, the Fed thought that an immediate hit to the banking systems capital wasbetter than the prolonged uncertainty of a bankruptcy proceeding.37 The tradeoffis the long-term economic cost of subsequent moral hazard behavior engenderedby a short-term bailout. The bailout of LTCM arguably prevented a more systemicproblem to avoid the economic consequences of the aftermath, but whether theprecedent was worth the moral hazard created is questionable.


18/28

18

If banks or GSEswere forced to

lend at ratesbelow actuarially

sound levels, thensomeone had

to subsidize thelosses.

Mission Regulation Competes withPrudential Regulation

Current social lending goals for housingin the United States date back to the HomeMortgage Disclosure Act (HMDA) of 1975

and the Community Reinvestment Act (CRA)of 1977. Purportedly, the HMDA and CRAreflected a concern that bankers were notlending enough in their local communities orneighborhoods, which were typically charac-terized by minority ethnic and/or racial con-centrations. Lending goals for Fannie Maewere introduced about the same time and os-tensibly for the same purpose.

Economists have provided ex post marketfailure rationalizations for housing goals fo-cused primarily on racial discrimination. Be-

cause older inner city neighborhoods often hada much higher percentage of African Ameri-cansand later, other racial minoritiestheimplicit concern of HMDA and CRA was withillegal racial discrimination. The theory behindthese goals was that there was a sufficient sup-ply of creditworthy borrowers in those areas,but that lenders were blinded by prejudiceand would not extend credit. Because incomeswere also generally much lower in older innercity neighborhoods and the risk of a systemicdecline in property values much greater, it was

generally difficult to distinguish illegal racialprofiling from legal credit discrimination.

HUD was charged with enforcing lawsprohibiting racial discrimination in housingfinance. Moreover, while the governments di-rect lending programs had been scaled back toavoid a budget impact, HUD had the capacityto direct credit to worthy borrowers who werediscriminated against through FHA, which italso administers. In this case, pricing addition-al credit risk was both politically problematicand actuarially difficult because adverse selec-

tion was a major obstacle to raising borrowerrates to cover the extra risk.

Lending goals became really serious in 1992with the adoption of the (ironically named)Federal Housing Enterprises Financial Safetyand Soundness Act. Part 2, Subpart B of theact required Fannie and Freddie to provideongoing assistance to the secondary market

for residential mortgages (including activitiesrelating to mortgages on housing for low- andmoderate-income families involving a reason-able economic return that may be less than thereturn earned on other activities). Politicians

and bureaucrats understood that it might notbe prejudice, but a sound appraisal of risks,that limited such lending, and thus the billslanguage, may be less than the return earnedon other activities, implies an acknowledge-ment of higher credit losses not actuariallypaid for with higher mortgage coupons.

In 1994 President Bill Clinton directedHUD to boost the homeownership rate to anall time high by the end of the century, andHUD secretary Henry Cisneros set a goal of a70 percent homeownership rate in the Nation-

al Homeownership Strategy of 1995. The U.Shomeownership rate had stabilized at about65 percent for the prior two decades in spite ofthe tremendous expansion of the GSEs. Thatis, even with mortgage credit generally avail-able with low or no down payment and oftenunderwritten at a below-market teaser interestrate and with no evidence of qualified bor-rowers systematically being denied credit, thehomeownership rate had not risen.

The origin of these lending goalsessen-tially quotaswas the belief that federal sub-

sidies (deposit insurance and GSE status) werea benefit for which a political price could beextracted. If banks or GSEs were forced tolend at rates below actuarially sound levelsor, more likely, make essentially uninsurableloansthen someone had to subsidize thelosses: probably shareholders. But these costscould also be passed on to more qualified borrowers or hidden in higher deposit insurancepremiums, artificially low deposit rates, po-tential taxpayer bailouts, or by other means ofopaquely providing subsidies through finance

The problem was not so much with themorality of extracting a user fee for federalsponsorship as it was the distortions and con-flict it created for regulators that could com-promise their primary prudential missionCommunity groups like ACORN used theirleverage with bureaucrats to extort subsidiesin return for discretionary approval of branch-


19/28

19

Most subprimeborrowers put inlittle or no cash.

ing and merger applications, which was par-ticularly ironic because just such restrictionshad led to deposit insurance and prudentialregulation in the 1930s. Reaching Cisneross70 percent target seemed to require major out-

reach and presumably large subsidies. HUDdemanded the outreach by requiring that theGSEs direct at least 30 percent of their loanfinancing to at- or below-area median incomehouseholds, but didnt request congressionalappropriations to fund the subsidies.

Making the FDIC responsible for enforc-ing social goals at banks to extract rents wasclearly in conflict with its primary mission ofprudential regulation. The conflict was evengreater at HUD. By 2000, given that HUD haddirected the GSEs to increase outreach to 42

percent of their loan financing, it is hard toargue that the GSEs were just making mort-gages liquid. But the GSEs could only fundthe implied subsidies by increasing risk, whichwas the responsibility of HUD as the pruden-tial regulator to prevent. Making HUDanagency with a social missionresponsible forprudential regulation as well as mission regu-lation was not only a conflict of interest, butsignaled a clear political intent that the latterwould eventually come to dominate.

Regulated Investors Fund theSubprime Lending Debacle

The Federal Reserves loose money policyduring the first half of the last decade alloweda credit boom to develop somewhere, andthe GSEs helped channel it to housing. But asubprime lending bubble in the United Stateskept the housing boom going for about threeadditional years, from mid-2004 throughmid-2007, and it is this bubble that caused the

collapse of the global financial system. Theconventional view is that unregulated privatelenders financed weak or victimized borrow-ers, eventually dragging the GSEs down withthem. The truth is that essentially the sameinvestors that funded private label securitizersalso funded GSE securities, and for essentiallythe same reasons.

Specifically, the investors reasons for fund-ing the GSEs were as follows:

1. Market discipline had long since beenreplaced by prudential regulation.

2. Prudential regulation conveyed agencystatus on the GSEs, essentially givingthem too-big-to-fail status, and that sta-tus was extended to non-GSE investmentgrade private-label securitizers.

3. Prudential regulation was by now politi-cally trumped by mission regulation, espe-cially for the GSEs. Speculatorsthe last re-maining source of market disciplinemayhave been able to prick the bubble severalyears earlier but for the continued bubbleinflation provided by the GSEs.

TheEx AnteCredit Risk of SubprimeLending

Peter Wallison provides a detailed discus-sion of credit risks in subprime lending.38One stunning fact reveals much about thecreditworthiness of subprime borrowers dur-ing this bubble: about one in five loans wasdelinquent within the first six months.39

In spite of all the concern over predatorylending, most of the subprime borrowers hadit pretty good. Most subprime borrowers put

in little or no cash. Most couldnt afford thefull monthly payment ex ante, obtaining a teas-er rate for several years. Many early subprimeborrowers were able to refinance, extending theteaser period for several more years. Some evengot to take out cash when refinancing that theyhadnt put in initially. Many lived rent-free fortwo, three, or even more years as foreclosureswere delayed by political pressure. The ex postconsequences suggest that borrowers andmortgage lenders were preying on investors,and indirectly on taxpayers.

Whether the borrowers were predatorsor prey, one thing is certain: historically theywould have been required to get mortgage in-surance, but not during the past decade. Themarket share of FHA was halved from 200103 to 200507, and the share of conventionalhigh LTV loans insured privately likely felleven more. The reason for this is simple: most


20/28

20

Only the GSEscould have kept

the bubbleinflated.

subprime borrowers could not afford the nor-mal insurance premium or, more importantly,qualify for insurance. Moreover, a much high-er premium would have been required, but ad-verse selection would have been extreme. Put

differently, the FHA and PMIs couldnt haveinsured these loans even if their risk modelsignored the bubble in house prices, which theydidnt, and as a consequence the vast majorityof subprime loans were not insured, but wereinstead funded by nontraditional means.

In place of insurance was the credit riskevaluation of the SEC-designated NRSROs,which evaluated risks differently. Their ratingsof second mortgage securities that substituteddirectly for insurance proved to be most in-strumental in obtaining funding. These rat-

ings proved wildly optimistic for two reasons.First, the rating agencies treated the first mort-gages on homes backed by second mortgagesthe same as if they had cash down paymentsor private mortgage insurance, even thoughthey had default rates that were approximatelyfive times greater.40 Second, they rated poolsof second mortgages as if they were home eq-uity loans. Home equity loans became popu-lar after the interest deduction was removedfor consumer credit in the 1980s, and the de-fault experience had been good because house

prices kept rising. But this was not the casewith the piggyback second mortgages thatwere replacing private mortgage insurance.Nevertheless, these piggyback seconds weresubsequently securitized with almost as muchleverage as the first mortgages.41 The exces-sively favorable ratings resulted in much lesscapital than regulated mortgage insurers wererequired to maintain to cover the same risk.

The conventional interpretation is thatthe credit rating agency models were seriouslyflawed, but it enabled the agencies that created

them to make more profit in the last decadethan in the previous century, mostly by ratingprivate-label securities.42 Having granted theagencies an incredibly valuable franchise, theSEC did nothing to limit the exploitation of thisfranchise or to regulate the adequacy of theirratings. Bypassing the mortgage insurers was anecessary condition for the subprime lending

debacle. Mission regulation compromised pru-dential regulators by enabling origination ofjunk mortgages, and SEC regulations regardingthe credit raters enabled the junk to be financed.

Leverage Ratios and Regulatory ArbitragePortfolio lenders subject to market disciplinewould face the same limitations as mortgage in-surers because the amount of leverage that mar-kets would allow for subprime credit risk wouldbe comparable to that of a finance companymaybe 4:1and the underlying loans could notearn the required expected return on equity. Pru-dential regulators charged with mitigating moralhazard should have substantially increased therisk-based required capital levels to reflect thisrisk. Had they done so, equity investors in private-

label securities would not have put up the financ-ing because of the insufficient expected returnGinnie Mae doesnt require equity, but the FHAscapital exposure would be transparent. GSE eq-uity investors would face the same facts as private-label securities investors, but the existing investorsmay have put up additional equity if required tohad they extrapolated their past returns to agencystatus without further due diligence. So only theGSEs could have kept the bubble inflated, but atleast taxpayer losses would have been lower withhigher capital requirements. In any event, all pru-

dential regulators did the exact opposite, and inthe case of the GSEs this reflected the explicit po-litical recognition that the higher cost of capitalwould preclude pursuit of their political quotas.

The GSEs regulator, the Office of FederalHousing Enterprise Oversight, required Fannieand Freddie to hold a mere 2.5 percent capitalagainst their debt-funded portfolio of wholeloans and only 0.45 percent capital for the MBS-funded portfolio. Fannie and Freddies averagecapital ratio for combined MBS and debt-fund-ed assets was 1 percent or less during the bubble,

reflecting the debt/MBS mix. But this overstatesrequired capital because the GSEs were allowedto hold half their capital in the form of pre-ferred stock, for which the risk-based capital re-quirements that commercial banks applied wasthe same 1.6 percent as applied to agency MBS,with the rest funded by government-insured de-posits. Hence, in the extreme, the governments


21/28

21

While it is truethat the shadowbanking marketwas not highly

regulated priorto the financialcrisis, thats notbecause of a lackof regulatoryauthority.

risk exposure could be almost double the statedbook leverage. This potential 200:1 combinedleverage is remarkable in that the historical ra-tional for such extreme leveragethat the GSEswere only pool insurers back-stopping the

PMIsno longer applied.Commercial bank risk-based capital regula-tions required only 8 percent total capital, witha 50 percent risk weighting for whole mortgageloans, a 20 percent risk weighting for AAA andAA securities, and a 100 percent risk weightingfor BBB securities. But like the GSEs, bankscould issue capital in the form of preferred stock,in this case trust preferred securities (TPS). TPSmay reflect the extreme of regulatory arbitrage.Whereas they were considered debt for tax andaccounting purposes, the Fed allowed TPS to

account for as much as 25 percent of equitycapital. Moreover, most of these securities werere-securitized in collateralized debt obligations

and issued as mostly investment grade debt thatother banks could then purchase with a 20 per-cent risk weighting. This didnt result in net newcapital going into the banking system, so thenet capital backing government-insured depos-

its could be as little as 75 percent of the nomi-nal risk-based capital, that is, only 1.2 percentfor AAA/AA-rated securities. And commercialbank sponsored off-balance sheet StructuredInvestment Vehicles (SIVs) provided even moreleverage, as they could purchase virtually anyinvestment grade securities by holding a cashreserve, typically around 1 percent with a putback to the sponsoring commercial bank.

But the leverage of investment banks waseven greater than that of commercial banks. Just prior to the housing market moving

from boom to bubble, in April 2004, the SEC voted to designate the five biggest nonbankinvestment banks (Goldman Sachs, Morgan

The Role of the Shadow Banking SystemMoney market funds were initially invested entirely in liquid U.S. Treasury securi-

ties, reflecting the arbitrage between market rates on Treasury securities and regu-lated deposit rates while avoiding the minimum Treasury purchase requirements.This investment strategy eliminated the need for systemic liquidity support andconfidence-building necessary for bank deposits. But when this source of arbitrage

revenue was eliminated by deregulation, the funds eventually bought GSE securitiesand highly ratedtypically AAAcommercial paper. Additionally, they relied on thetheory that prices couldnt change very much based on credit deterioration over theshort (typically 30 day) lifespan of these securities.

Money markets provide a convenience for retail and small institutional custom-ers, but large corporate customers can fund and invest directly in the money markets,bypassing money market funds. Some operate on a significant cash deficit, relying onthe ability to roll over outstanding paper. Others invest directly in commercial paperof other firms, investment banks, and SIVs of commercial banks. They may also makedirect short-term repo loans collateralized by liquidthat is, marketablesecuritiesto obtain higher yields than available on bank deposits. Hence the shadow bankingmarket ultimately issued bank-like deposits and purchased bank-like investments,

but without bank regulation and systemic liquidity support.While it is true that the shadow banking market was not highly regulated prior to

the financial crisis, thats not because of a lack of regulatory authority. Bank regula-tors had total authority to regulate SIVs and repo lending, and the SEC had regulato-ry authority over money market funds, corporate cash management, and investmentbanks. Regulators simply didnt understand the danger of muddling the conceptsof marketability with liquidity and the consequent systemic risk when the shadowbanking system directly or indirectly funded illiquid investments.


22/28

22

The goal ofprivate-label

securitizers wasto issue as much

cheap investmentgrade debt as

possible.

Stanley, Lehman Brothers, Merrill Lynch, andBear Stearns) as consolidated supervised en-tities and lower their capital requirements43based on computer model simulations ofthe 1988 Basel I Capital Accords. As a conse-

quence, investment banks had dramaticallygreater and less transparent leverage duringthis bubble than in prior decades, as statedbook leverage ratios approximately doubled.By year-end 2007, the book capital-to-assetsratio for Goldman Sachs, Lehman Broth-ers, Merrill Lynch, and Morgan Stanley aver-aged 3.33 percent, far exceeding commercialbank leverage.44 Moreover, as with the GSEsand commercial banks, book leverage ratiosunderstated the extent of investment bankover-leveraging because various accounting

g

Documents

What Made the Financial Crisis Systemic?, Cato Policy Analysis No. 693