President Reagan appointed Edwin J. Gray to the Bank Board in early 1983 with the expectation that he would become chairman when Pratt resigned a few months later. Gray was a longtime friend of the president, and had served as his press secretary in California and then as a public-relations officer for a large San Diego S&L run by Gordon Luce, a member of the President’s “kitchen cabinet.” Gray served in the White House as a domestic policy advisor for President Reagan. He was valedictorian of his high school and received his BA in journalism. Like the president, he was a Democrat who became a conservative Republican. Gray led an antitax initiative in San Diego. He loved the president and his policies. Gray believed in deregulation. He freely told people that the president appointed him to make the league happy—it considered him a patsy.
Gray soon took positions that put him in opposition to the president he loved, his party, and his own philosophical views. He had nothing to gain from this transformation. He did not welcome it. He knew it would ruin his career and that he would lose what he most treasured, Ronald and Nancy Reagan’s support and friendship. He knew that the counterattack would be savage, personal, and effective. He would make too many powerful enemies and he would have few effective allies.
Some people fight because they enjoy it, others because that is how they were trained; some because they are thin skinned, others because they are bullies; some because they are irrational, and others because they are cornered and have no other hope of survival. Soldiers fight mostly for their close comrades. Young males fight because they are desperately afraid of being afraid. Drug sellers fight because if they do not intimidate, they die. Most of the reasons people fight are bad ones. Gray did not fight for any of these reasons. Gray was not cornered: if he were unwilling to support S&L deregulation, he could have gone home to San Diego.
Gray was not naturally brave. His hands shook, he couldn’t sleep, he chain-smoked, and his face became pallid. He shrank from confrontation. He was not calm under pressure. He was so nervous that he could not concentrate. He projected nervousness, making others uncomfortable. He did not have “the right stuff.” John Glenn, the astronaut who exemplified the right stuff, disdained Gray. Glenn became one of the Keating Five (the five U.S. senators who tried, at Charles Keating’s behest, to intimidate Gray), and Gray refused to back down to their pressure. Perhaps we should reexamine our definition of “the right stuff.” It is how far Gray had to move philosophically, at what personal cost, and in a manner so at odds with his basic personality that makes him both the most unlikely of heroes and so heroic. But that is not how Gray’s chairmanship began.
Gray continued Pratt’s policies during his first five months as chairman. He imposed a hiring freeze. He adopted a creative RAP rule. He approved disastrous acquisitions by men who would become infamous control frauds. He gave speeches lauding deregulation and urging S&Ls to make use of the new powers. He fought the last war, constantly reminding S&Ls of the need to reduce interest rate risk. He slowed the rate of S&L closures. Like Pratt, he capped FSLIC spending at a ridiculously inadequate level in order to show a small increase in the FSLIC fund at the end of 1983 ($6.4 billion). He approved goodwill mergers.
He made no improvement in the moribund enforcement efforts or in the dysfunctional examination and supervision systems. Indeed, he harmed supervision by agreeing to a proposal to move the headquarters of the 9th District from Little Rock, Arkansas, to Dallas, Texas. The new FHLB-Dallas decided to save money by being cheap when paying for relocation. The overwhelming majority of its supervisors resigned in late 1983.
Gray did not clean up the “change of control” system that allowed almost anyone to acquire an S&L. By continuing Pratt’s policies, Gray increased the damage the control frauds later caused.
Donald Regan was the head of Merrill Lynch before he became treasury secretary during President Reagan’s first term. He made Merrill Lynch the dominant deposit broker.1
Deposit brokers seek S&Ls and banks offering the highest interest rate on insured deposits. Deregulation allowed banks and S&Ls to compete on interest rates. Insured depositors had no risk of loss or even inconvenience. In order to maintain public confidence, if the FSLIC liquidated an S&L, it paid the depositors in full within days. An insolvent S&L that was closed on Friday afternoon would reopen on Monday morning. Deposit brokers fueled the growth that made control frauds ideal Ponzi schemes. An S&L could raise hundreds of millions of dollars in as little as three days through the brokers.
The Bank Board and the FDIC simultaneously adopted rules in March 1984 restricting insurance for accounts placed by brokers. The rules could not work. S&Ls could grow almost as rapidly by advertising a slightly higher interest rate and then “dialing for dollars.” S&Ls created phone banks to call prospective depositors and solicit deposits. Any S&L could raise tens of millions of dollars in a week without using brokers.
The Bank Board’s authority to restrict deposit insurance for brokered deposits was weak. My predecessor as Bank Board litigation director, Harvey Simon, advised Gray that the brokers would likely win their lawsuit if he adopted the rule. Simon proved correct.
Treasury Secretary Regan opposed the rule. He believed in deregulation, and he thought that brokers helped customers. Regan was Mehle’s boss when he testified that insolvent S&Ls should grow rapidly and that the industry’s fast growth showed that it was sound. Regan was convinced that President Reagan shared his views. Gray made a powerful enemy. Regan sought to drive him to resign.
The deposit broker rule was a mistake. It would not have worked even if we had won the case. We had more effective ways of limiting growth and the clear statutory authority to use them. The rule diverted resources we could have used more profitably against the control frauds.2
Empire Savings made famous the “land flip.” The S&L paid the small fry (which it called “minnows” or “guppies”) up to $50,000 to lend their name to this fraud. The guppies sat at a long table. A big fish started the land flip by selling a piece of undeveloped (“raw”) land to Guppy One for $2,000 an acre. Guppy One sells it to Guppy Two for $4,000 an acre. This continued until another big fish agreed to pay $50,000 an acre. The S&L would then loan money to the big fish to purchase the raw land at this price. In five minutes, the “market” value of the land rose twenty-five-fold. The big fish was a residential real estate developer who was purportedly going to build thousands of townhouses.
It was impossible, of course, for such a developer to create a profitable project. The developer was part of a fraud, and he made money in several different ways. The S&L paid him an up-front “profit.” The developer hired related entities at generous rates to do the construction. The developer gained directly by having a hidden interest in the construction group or through kickbacks for awarding contracts. The construction group charged high rates but provided inferior work and pocketed the difference. One of the scams that caused losses at Empire Savings occurred when contractors used one-fifth of the concrete needed to construct a road properly. The roads soon buckled. Empire Savings also funded the construction loans in full without tying disbursements to completion of construction.
The result was a disaster of epic proportions. Because other control frauds joined with Empire Savings to run the same scam and because they all (semi-) built their projects in the same area (along the “I-30 corridor” in Garland, Texas, near Dallas), the disaster was concentrated and stretched for miles.
The Bank Board examiners hired an old-time Texas real estate expert to help them figure out how bad the mess was. He decided that you had to see it to believe it. He videotaped a tour of the I-30 corridor. There were thousands of partially completed units left open to the elements and arsonists. Many units had no construction, just the concrete pad. We dubbed them “Martian landing pads.” The images on the tape were so revolting that they created horror in everyone who saw them. He shot some of it from a small plane, which provided panoramic views of devastation. His narration increased the effect. It was so disturbing because it was artless. Imagine a strong Texas accent that remains calm and matter-of-fact while presenting a nightmare. Gray watched the film on March 14, 1984 (Day 1993, 156–157). He told me he wanted to throw up when he saw it.
He showed the tape to as many senior officials as possible, including Paul Volcker (Day 1993, 162). Kathleen Day interviewed one of the key people Gray showed the tape to, the House Banking Committee chief of staff, Paul Nelson. Whereas Volcker was horrified, Nelson responded, “Gray’s the regulator, not us…. That’s his job to stop this. Why am I watching this?” (ibid.) The smaller incidents are often the most revealing. Paul Nelson’s reaction illustrates why congressional oversight failed throughout the critical years of the debacle. Gray was stunned, outraged, and determined to prevent any future Empire Savings from spawning similar horrors. Nelson and House Banking Committee chairman St Germain did not see that their deregulation had created the environment that produced the wave of control fraud. They were annoyed with the messenger.
Notably, he did not try to show the film to President Reagan. Gray and the president (who became famous as “the great communicator”) neither met nor talked while he was chairman. Even when Regan was pushing for his resignation, Gray never tried to use his friendship with the president to have him call off the dogs. Many of us asked him why. His answer was always the same: “You don’t know him. You can’t do it; it’s impossible.” The president was an extraordinarily hands-off manager. He did not ask people like Gray for advice about how to deal with the debacle. Treasury and the OMB talked to him. They had two contradictory messages throughout his presidency—there is no crisis, and our top priority is making sure that the public does not learn of the crisis. President Reagan’s autobiography ignored the debacle, at the time the largest financial scandal in U.S. history.
The mythic part of the story about the Empire Savings videotape is that it produced an instantaneous transformation in Gray: he rode in as a deregulator, the truths in the film knocked him off his horse, and he awoke as the Great Reregulator. Gray had begun transforming himself into a reregulator before he saw the film. Further, Gray did not complete his transformation until years after he saw the film. The film was, however, important in speeding Gray’s transformation and the intensity of the war against the control frauds.
Gray’s first effective counterattack against the control frauds assaulted the de novos. Taggart (the California Department of Savings & Loans commissioner) approved 200 new California charters by early 1984. Texas and Florida approved dozens of new charters. Gray adopted a rule in November 1983 requiring de novos to have 7 percent capital—over twice the requirement for other S&Ls—in order to receive FSLIC insurance.
Gray’s second act was more decisive. He refused to approve FSLIC insurance for de novos from California, Texas, and Florida unless they improved their regulation. This was one of the most effective actions possible against control frauds. There would have been hundreds of additional control frauds but for Gray’s moratorium. However, it enraged all three state commissioners and state and federal politicians representing California, Texas, and Florida, states with powerful congressional and state officials. The list of Gray’s powerful opponents grew.
The related problem was that Gray had no clear legal basis for imposing a moratorium. His action was logical: none of the states had remotely adequate regulatory resources to deal with their existing charters. It was irresponsible of them to increase the number of charters. Gray tried to convince them to stop the approvals until they could build up their staffs. The three commissioners refused. Gray’s action was desirable; but was it legal?3
Gray proposed the net worth and direct investment rules in early 1984. He focused, however, on the deposit broker rule. He saw it as a direct way to limit growth, which he considered the central problem. When the courts struck down the deposit broker rule, these other two rules became his top priority. The Bank Board designed them to stop any future Empire Savings.
The de novo rule and Gray’s moratorium on granting FSLIC insurance to new charters in California, Texas, and Florida had substantial support within the industry; they restricted future competitors. The deposit broker rule had some support within the traditional S&Ls. The proposed direct investment rule (which limited such investments to 10 percent of total assets) did not affect many S&Ls.4 Gray did not expect the direct investment rule to spur major opposition. Gray designed the net worth rule to end extreme growth. He knew that the industry would oppose it because so many S&Ls were trying to “grow out of their problems.”
The idea of growing out of their problems was seductive to many S&Ls. The amount of growth required, however, was enormous. When interest rates are high, home sales slow and it is hard for a mortgage lender to grow, even slowly. The solution was to grow rapidly by investing in much higher-yield (and higher-risk) assets. The higher the yield of the new assets, the less growth an S&L needed in order to grow out of its problems. Deregulation made this solution possible.
One of the standard logical fallacies is that of composition. Logicians warn that the fact that something is true in a particular case does not prove that it will hold true when applied simultaneously to many cases. An individual S&L might be able to grow out of its problems, but an industry cannot. Indeed, it is a prescription for disaster to try.
Some critics claim that Pratt expressly encouraged S&Ls to grow out of their problems. Pratt was generally careful to avoid going that far. Nevertheless, he gave the industry the ability to grow massively, and he took no action to stop the large number of S&Ls that began to grow rapidly. The results were, predictably, disastrous.
The proposed net worth rule restricted growth by increasing capital requirements for faster growing S&Ls. Gray proposed to reverse the Bank Board’s prior policy, five-year averaging, which reduced the capital requirement for fast growers. Gray acted because of the tendency of fast growers to become the worst failures. The agency did not understand the acquisition, development, and construction (ADC) Ponzi scheme fully when it drafted the rule. This was ironic because the rule struck the control frauds’ Achilles’ heel and proved the most critical act of reregulation. A Ponzi scheme operates by growing rapidly and using a portion of the new money brought in to pay off the old investors while the fraudster pockets the remainder. Pilots say, “Speed is life.” For a Ponzi, growth is life.
S&L Ponzis invested in assets that had no readily ascertainable market value and that allowed S&Ls to treat (fictitious) noncash income as real for GAAP purposes. Professionals, like appraisers, determine value when there is no clear market value for an asset. Every S&L control fraud was able to get top professionals to overstate asset values massively and to get a clean opinion from a Big 8 auditor blessing its fake financial statements. The advantage of using noncash “income” as a fraud vehicle was that it guaranteed production of the income.
The “acquisition” in ADC refers to buying raw land. “Development” means adding improvements like sewers, utility lines, and roads. “Construction” meant creating a commercial building or a multifamily residential project. Control frauds called their ADC transactions loans and structured them as loans, but the typical ADC loan was really a direct investment. The law, the fields of economics and finance, and accounting all agree: if the economic reality is that the lender is taking an equity risk, then the transaction is a direct or equity investment, not a loan. If the success of the underlying project (such as the commercial office building being constructed) determines whether the borrower repays the S&L, then the lender is assuming an equity risk.
Big 8 auditors—despite three attempts by the profession to prevent the abuse—consistently blessed the accounting of ADC transactions as loans. Treating ADC deals as loans created extraordinary (fraudulent) income and hid enormous real losses. S&Ls combined accounting fraud with deregulation and massive growth to create the ideal Ponzi. If the auditors had required the control frauds to account for their ADCs as direct investments, they could not have recognized the fictitious income.
The typical control fraud made ADC loans with the following characteristics.
• There was no down payment.
• There were substantial up-front points and fees.
• All of those points and fees were self-funded: the S&L loaned the buyer the money to pay the S&L the points and fees.
• The term of the loan was two or three years.
• There was no repayment of principal on the loan prior to maturity. (The loan was interest-only, or nonamortizing.)
• The S&L self-funded all of the interest payments. The S&L paid itself the interest when it came due out of an interest reserve. The S&L increased the borrower’s debt by an amount equal to the interest reserve.
• The interest rate on the loans was considerably above prime.
• The borrower had no personal liability on the debt (the loan was nonrecourse). His construction company was indebted, but the developer was not. He did not provide any meaningful personal guarantee of his company’s debt.
• The developer pledged the real estate project (the collateral) to secure the loan.
• The loan amount equaled the (purported) value of the collateral.
• It was common for the borrower to receive at closing a developer’s profit that could represent up to 2 percent of the loan balance.
• It was common for borrowers to give S&Ls an equity kicker, an interest in the developer’s net profits on the project. At first, these kickers often exceeded 50 percent. After the accounting profession issued a “notice to practitioners” that said a 50 percent equity kicker was evidence that the transaction was not a true loan, it became common to give a 49 percent interest. (For the characteristics in this list, see NCFIRRE 1993a; Black 1993b; Lowy 1991; O’Shea 1991; Strunk and Case 1988; Mayer 1990; Pizzo, Fricker, and Muolo 1989; Calavita, Pontell, and Tillman 1997.)
The implications of these characteristics are not obvious. One has to understand the fraud mechanism, Bank Board regulation, and a bit about accounting to see how elegant a Ponzi the control frauds created. The first implication is that one can see why the S&Ls were taking an equity risk, not making a loan. The real estate developers were not personally liable to repay the loan. Their companies were not really on the hook either: they used shell companies with no assets to sign the note. The developer would not repay the S&L unless the real estate project succeeded. Indeed, it had to succeed fully because the loan was 100 percent of the projected value; the S&L would lose money if the appraiser inflated that value even slightly. These were not close calls: the typical ADC Ponzi loan was clearly an equity investment. This makes their audit partners’ consistent treatment of them as true loans all the more disturbing.
Cash moved in one direction in an ADC Ponzi scheme: out of the S&L to the developer. The developer typically did not provide an S&L control fraud with any cash—no down payment, no fees or points, and no interest payment. The developer was only required to pay cash when the loan matured, and as I’ll explain, they rarely paid cash even then.
This may seem to be a foolish way to run a fraud, but it is actually clever. The first advantage is that an ADC loan can never default prior to maturity. The Bank Board based its supervisory system on rates of loan defaults and delinquencies. Those are very good leading indicators of failure in conventional home mortgage lending. Supervisors considered S&Ls with low delinquency rates to be safe and sound. ADC lending looked safe and highly profitable. That was, of course, too good to be true, but few supervisors thought like white-collar criminologists in the early years. Pratt praised the new entrepreneurs’ managerial skills. Managerial expertise brings profits. A Big 8 audit firm blessed it. It must be true.
The second implication is that an S&L running an ADC Ponzi was certain to report extraordinarily high “income” as long as the auditor allowed it to classify the deals as loans instead of as investments. Every control fraud was able to get its auditor to misclassify its ADC deals as loans, so that caveat had no effect. It was simply a matter of math (and some additional abusive accounting). ADC Ponzis booked most of the points and fees that they self-funded, i.e., paid themselves as immediate income. Roughly 3 percent of the total loan amount instantly became income.5
ADC Ponzis recognized the self-funded interest as income every month or quarter when interest came due, crediting the borrower for paying interest to the S&L from the interest reserve. Because of the high fees and interest rates, all of it self-funded, a control fraud was certain to report high profits. The extremely rapid growth rate of the ADC Ponzis guaranteed that these profits would be extraordinary. Any S&L that grew rapidly and made primarily ADC loans was mathematically guaranteed to report extraordinary profits. This is why the worst control frauds invariably reported that they were among the most profitable S&Ls. Indeed, one of the best ways to spot control frauds was to concentrate examination efforts on the S&Ls reporting the highest profits. This also demonstrates how fully the frauds suborned their Big 8 audit partners. They were routinely able to get clean opinions attesting that they were among the most profitable S&Ls while in fact being massively insolvent.
The consistent patterns used to optimize a control fraud made it easier for the agency to spot them. The frauds’ first requisite was to make their auditors their allies. After that, the three keys were growth rate, the dollar amount of ADC loans, and the magnitude of the interest rate, fees, and equity kicker. ADC Ponzis frequently grew more than 100 percent in a year; they typically grew by more than 50 percent; some grew by more than 1,000 percent. Deregulation was a major cause of the debacle not because it permitted far greater investment in risky assets, but because the permitted investment assets were optimal for creating fictitious income and hiding real losses. Deregulation combined with weak supervision and deposit insurance to create the ideal Ponzi. The worst control frauds were Texas- and California-chartered S&Ls because they could put 100 percent of their assets in investments that were the best vehicles for fraud—typically, ADC loans. (The auditors’ failure to treat ADC loans as investments was harmful. Texas S&Ls could put only a small percentage of their assets in equity investments, but 100 percent of their assets into ADC loans.) Control frauds routinely made ADC loans that were substantially greater than the market value of even the completed project, because the total dollar amount of the ADC loans drove income. The rules prohibited this practice, so control frauds engaged in “appraiser shopping” to secure grossly inflated appraised values for the real estate.
Control frauds sought to make loans to uncreditworthy developers because they were willing to pay the highest interest rates and fees and to provide the largest equity kickers (Pierce 1993; Akerlof and Romer 1993; Calavita, Pontell, and Tillman 1997; O’Shea 1991; Lowy 1991; Pizzo, Fricker, and Muolo 1989; Mayer 1990). Deliberately seeking to make loans to the worst developers strikes most people as absurd. Surely, it is in the interest of control frauds to make profitable loans so they can stay in business and continue to loot the corporation? Most people (fortunately) do not think like white-collar criminals. Control frauds operate by different, perverse rules that harm society, but that does not mean they are irrational.
A Ponzi scheme invariably collapses, but that does not mean that the fraud fails: the corrupt CEO can make a great deal of money looting the company. The purported profitability of the Ponzi schemes, combined with deposit insurance, meant that S&Ls could grow rapidly for several years before they collapsed. Such CEOs did not have the option of making money honestly, as my discussion about the inability to grow out of your problem illustrated. ADC accounting frauds created fictitious income that allowed CEOs to convert firm assets to personal use through normal and seemingly legitimate corporate mechanisms: raises, bonuses, stock options, and perks.6
Poor-quality real estate developers had no meaningful downside to agreeing to pay high fees, interest rates, and equity participations to the control fraud. They took no financial risk. They did not actually pay high interest rates and fees. The ADC loans did not require bad developers to pay a penny of their own money or to put a penny of their own money at risk. All their payments were actually made by the S&L; there was no down payment; and they did not have to guarantee the loan personally. When the loan defaulted, they could walk away from it.
A good developer could have also walked away from the default without any direct financial loss, but he would have suffered a loss of reputation. A reputation as a high-quality developer has substantial indirect financial value. A developer who had no favorable reputation, or even a poor reputation, however, would suffer little or no loss of reputation when the loan defaulted (as it was very likely to do given the typical inflated appraisal and the poor business acumen of the developer). Developers who had poor reputations could not get similar loans from a reputable lender. This maximized the fraudulent S&Ls’ leverage over them.7
Everyone agrees that underwriting by the S&L high fliers was pervasively horrific (Patriarca 1987, 3–5; GAO Thrift Failures 1989, 31–38). They frequently made massive ADC loans to individuals without conducting credit checks or appraising the value of the real estate. This maximized adverse selection (and losses). It was a perversely rational practice for a control fraud precisely because it maximized adverse selection. Moreover, if the initial appraisal would have shown a large loss (or the credit check would have shown the developer to be uncredit-worthy), it was better to make the loan without the damning appraisal or credit check in the file, where the Bank Board examiners could find it and use it to prove that the loans were unsafe and unsound.
The cover-up phase of the ADC Ponzis made bad developers (and informal alliances with other control frauds) critical. Adverse selection and the perverse incentive of control frauds to increase their ADC lending in the teeth of a glut of commercial real estate meant that the ADC projects were likely to fail and the loans would default at maturity. Control frauds hid these defaults and turned them into new sources of fraudulent income and new means of deceiving the regulators. In increasing order of elegance: S&Ls would refinance their ADC loans, engage in “cash for trash” deals with bad borrowers, trade “my dead horse for your dead cow” with other control frauds, and perform intricate transactions with “daisy chains” of control frauds. The fundamental gambit was to remove the real loss and create fictitious income through fraudulent loans and sales. Control frauds used equity kickers to create fictitious profits from sham sales.
The crudest way to prevent default was to refinance the ADC loan. The Ponzi would start the whole game again by self-funding the fees and interest. The problem was that this was easy for auditors and examiners to spot and should have raised warning flags. Remarkably, Arthur Young & Company gave clean opinions to two of the most notorious Texas control frauds, Vernon Savings and Western Savings, even though they routinely refinanced all their ADC loans.8
Cash-for-trash deals are harder to spot because they create seemingly arm’s-length sales. The best evidence of market value is a recent arm’s-length sale. A good rule, except when there is fraud. A cash-for-trash deal started with an uncreditworthy borrower asking the S&L for a $3 million ADC loan. The loan officer responds that the S&L will not loan him $3 million, but it will loan him $33 million. The catch is that the borrower has to use $30 million of the loan proceeds (the “cash”) to buy a particular property (the “trash”). The property is a soon-to-fail real estate project that the S&L funded with an ADC loan.9
“I’ll trade you my dead horse for your dead cow” works the same accounting alchemy. “Mustang” S&L buys a troubled project from one of “Longhorn” S&L’s ADC borrowers that is about to default, and Longhorn S&L reciprocates. Both S&Ls pay well above book value, allowing both S&Ls to avoid default and loss recognition while booking a substantial gain. The examiners are very unlikely to spot this because the control frauds hide the linkage. However, it is possible that examiners will review both deals.
The most elegant scam was impossible to spot by examining an S&L. “Daisy chains” of control frauds operated in some parts of the country. This was not some vast, directed conspiracy, but a large mutual aid society. Daisy chains allowed S&L A to buy S&L C’s problem asset, while S&L B bought A’s and C bought B’s. No examiner, no manner how suspicious, could find from the records of A, B, or C that the purchases were linked, because no single S&L’s records could demonstrate the link.
This discussion explains the first two reasons that ADC Ponzis produced such large losses. First, making bad loans to bad developers maximized control fraud profits.
Second, because control frauds used adverse selection to attract the worst borrowers, they increased the risk that junior officers and the borrower would engage in independent fraud schemes.10 One reason control fraud through an ADC Ponzi was so elegant is that the CEO never had to openly direct an officer or employee to act fraudulently. All the CEO had to do was to emphasize that the business plan was to grow extremely rapidly and invest in ADC loans, that career advancement and bonuses were dependent on ADC volumes, and that he valued aggressive team players who were creative. It became immediately apparent, as the CEO approved one insane ADC loan after another, that credit quality was irrelevant.
ADC Ponzis were particularly susceptible to insider and borrower fraud because of the nature of construction. As I explained, fraud is always a major problem in that industry. Such fraud sometimes meant, as at Empire Savings, that it was cheaper to raze the construction than to repair it.11
Control frauds produced catastrophic real estate losses because they maximized their fraudulent take by making an enormous number of wildly inflated ADC loans to the worst borrowers for unneeded projects. Surging vacancy rates increased the need to make additional bad ADC loans and added to the record glut of commercial office space and multifamily housing. Those dynamics meant that an ADC Ponzi had to make progressively more and worse ADC loans to delay its collapse. (Remember, a Ponzi’s mantra is “growth is life.”)
Because the ADC Ponzi was the best available fraud mechanism, it made sense for other S&L (and bank) control frauds to follow the same strategy. The control frauds came in a wave: about 300 S&Ls (and scores of banks) were dumping ever greater amounts of unneeded projects into the same glutted markets. Worse, the fraudulent S&Ls were concentrated in the states with the greatest deregulation and the weakest state regulation, so they drove the Texas and Arizona real estate markets. This, together with the perverse incentive created by the 1981 Tax Act to build uneconomic real estate projects, produced first the bubble in Texas real estate values in the early 1980s and then the horrific crash (Akerlof and Romer 1993; NCFIRRE 1993a; Black 1993c). The fact that oil prices fell at the same time made the losses all the greater. Prominent Texas politicians intervened on behalf of the Texas control frauds to blunt Gray’s war against them. They premised their efforts on the view that the high fliers were innocent victims of a regional recession. In fact, they were the villains of the piece. The politicians added to the ultimate cost to the taxpayers, as I will soon explain.
Gray had three good ideas in proposing to raise capital requirements and to restrict growth and direct investments. He also had three problems. The problems immediately blocked reregulation. The elegance of the ADC Ponzi made it very difficult for the Bank Board to prove a need for regulation; his opponents were numerous, politically powerful, and skilled; and he led a dysfunctional agency without allies. ADC Ponzis looked superb: they reported far higher income, higher capital levels, and low default rates. S&Ls that grew faster and made more direct investments reported record profits. Any reregulation would have enraged the administration, but Gray’s proposals seemed irrational on their face.
This substantive problem made it easy for Gray’s critics to attack his motives and competence; this was his second problem. Critics had a ready explanation for the irrationality. Gray had worked for a large California S&L. The large traditional S&Ls (with inept managers) were losing the competition with the innovative, entrepreneurial entrants. The economic literature is full of examples of politically powerful businesses running to the legislature or the regulators to hamstring their superior competitors. The literature also shows that old-line firms typically disparage the morals of the new competitors and call them names like high fliers (Fischel 1995). Gray was falling for this old ad hominem assault on the new because he was dumb and an industry tool.
Gray’s opponents also made his substantive problems worse by hiring lawyers and economists, including Alan Greenspan, who explained why the agency had no statutory authority to adopt the rules and conducted studies proving that faster growth and greater direct investments produced higher profits and healthier S&Ls. Keating expanded the ranks and virulence of Gray’s opponents through a wide-ranging effort to build allies in the administration, the press, and Congress and among the state S&L commissioners.
The first two problems were grave, but the third was disabling. The Bank Board was dysfunctional for many reasons, and Gray often contributed to that problem. He was a terrible manager. Meetings never ran on time, agendas were never real. He wandered in and out of meetings.
There was no strategic planning. The Office of Policy and Economic Research (OPER) was in charge of developing policy, but had neither the expertise nor the orientation to do so. OPER faced the common problem of how to attract good economists when government pay caps made their salaries uncompetitive. OPER met that problem by hiring some more-prestigious economists on short-term appointments, who were attracted by the prospect of using agency data to push their research agendas. That was sensible, but it meant that the economists often had little institutional orientation and were interested almost exclusively in their own research. There were two telling results. First, only a small percentage of OPER studies had any direct tie to the problems presented by the high fliers. There was no attempt by OPER to determine whether they were gamblers or control frauds. I know of only one OPER study of fraud. It was finished in 1989 after the Bank Board had removed the last high flier. There was no relevant OPER research when Gray proposed the rules on capital and on direct investments. OPER remained focused on interest-rate risk into 1984, then lost focus as its staff pursued individual research agendas.
OPER controlled what information the industry reported to the Bank Board in the “Thrift Financial Reports.” Everyone agrees that ADC lending caused the worst losses, yet OPER never gathered data on ADC lending and never studied the ADC Ponzi schemes.12
OPER posed two other problems: its staff reflexively opposed re-regulation, and they were not very good at translating economics into English. That was unfortunate, because the lawyers who drafted the rules and, more importantly, the rationale for the rules, which the courts would review for sufficiency, were economically illiterate and did not know anything about examination, supervision, or fraud. The “regulations and legislation” (“Regs and Legs”) division could craft a convoluted rule, and had no problems with reregulating, but they could not communicate with the economists or the supervisors.
The supervisors were a mixed bag. They were generally former supervisory agents with one of the FHLBs. They meant well, and most of them came to favor reregulation, but they were rarely examiners, and they often had not been hands-on supervisors for many years. They often lagged in understanding fraud mechanisms and emerging patterns.
Altogether, reregulation was a very uneven contest. Gray had the best lawyers in the world against him; they had prestigious economists as experts who produced economic studies showing that the rules were harmful; the administration opposed the rules; the industry attacked him; and his staff was, by disposition or talent, poorly poised to support him. Eric Hemel, the OPER director, sent a copy of Lincoln Savings’ comments on the proposed direct investment rule to Norm Raiden, Gray’s general counsel, warning him that the Bank Board’s “preamble” had not made a strong case for the rule. Lincoln Savings was promising to sue to strike down the rules. Norm asked me, his new litigation director, to look at the preamble.13
This request transformed my career. It led me to become Gray’s principal aide for reregulation, and it changed the way the agency approached the war against the control frauds.14 The Bank Board had done no economic studies supporting either rule. It had not explained why economic theory supported both proposals. The preamble cited only one example of a direct investment causing problems: ADC loans at Empire Savings. It did not explain why ADC loans were really direct investments.
I advised Gray that he needed to repropose both the capital and the direct investment rules because the preambles had so little support that we would likely lose both rules to court challenges. One of the impressive things about Gray is that he took that advice and reproposed the rules.15 Gray tasked me with getting the rules in shape to survive the legal challenges. I asked the supervisory staff to write up case studies of our dismal experience with control frauds. They proved incapable of doing so. I had to assign the drafting to my litigation attorneys who were handling the lawsuits arising from the control frauds. They provided concrete examples of how excessive growth and direct investments led to serious losses. The supervisors reviewed the litigators’ case studies and agreed they were accurate. They signed the final memoranda and were supportive.
I asked our economists to carry out studies to examine the need for the rules. They were largely at a loss as to how to do so. In their view, economic theory suggested that the direct investment rule was unnecessary and harmful. They believed that increased capital requirements were desirable, but had no idea how to show that empirically. I ended up designing several economic studies, an unusual job for a litigation director.16
The Bank Board was able to adopt the reproposed rules in late 1984. We know that Keating had planned to sue to strike them down. The quality of the support for the rules caused his lawyers to advise him not to sue. This was a remarkable turnaround, and it gave Gray the confidence to proceed with additional reregulation. Gray’s reregulation did not prevent the debacle, but it saved the nation from disaster. Without the restriction on growth imposed by the rule on capital, the control frauds would have grown large enough to cause losses that would have threatened the general economy by the end of Gray’s term. Without the growth rule, the control frauds would have grown massively during Wall’s chairmanship, and their political power would have endangered the nation.
Because S&L control frauds were Ponzi schemes, the consequences of delay in stopping their growth were horrific. Without Gray’s war on the control frauds, their growth would have increased throughout his term. His successor, Danny Wall, would not have taken on the control frauds for reasons made clear later in this book. Indeed, he helped the most notorious control fraud escape regulatory control. Eventually, the expanding wave of control fraud would have caused such a massive bubble in real estate values that it would have collapsed. Since Japan’s real estate and stock market bubbles grew for a full decade during the 1980s without the growth advantages provided by deposit insurance, a U.S. bubble could have lasted for over a decade. Therefore, the wave of control fraud could have extended throughout the Reagan and Bush administrations had it not been for Gray’s desperate war against the control frauds.
There were over 300 control frauds. My study found that the 11th District (California, Arizona, and Nevada) had 58 control frauds (Black 1998). Texas had over 100. Together, California and Texas had over half of the control frauds and produced over half the total losses. Prosecutors convicted over 1,000 S&L insiders of felonies. (That was an unprecedented level of success, particularly given the grossly inadequate resources provided by the Justice Department.) Every detailed study of the most expensive S&L failures found a common pattern that included control fraud (Black, Calavita, and Pontell 1995; Black 1998).
The S&L industry grew by 18.6 percent in 1983, and that staggering rate of growth increased to 19.9 percent in 1984. The industry would roughly double in size every four years at that rate (because growth rates, like interest, compound). Nominal GNP grew at less than half that rate, and commercial banks grew at only one-fourth the rate of the S&L industry in 1984. The growth in the S&L industry was concentrated disproportionately in the worst S&Ls. The soon-to-fail group grew at 101 percent, over twice the industry rate, between 1982 and 1985 (White 1991, 100–101). But the high fliers lived up to their name: 74 of them grew by more than 400 percent during those same years. In 1984, more than 300 S&Ls grew by more than 50 percent (NCFIRRE 1993a, 52). Half of those ultrafast growers were located in Texas, California, and Florida (Strunk and Case 1988, 133). Over 700 S&Ls grew by more than 25 percent in the first half of 1984 (ibid., 132).
Growth was fastest in the states with the greatest deregulation and the weakest state supervision (White 1991, 100). Forty Texas S&Ls grew by more than 300 percent between 1982 and 1985. The percentage of assets best suited to creating fictitious accounting income increased dramatically at the high fliers between 1982 and 1985 (ibid., 102–103). The fastest growing S&Ls reported a return on assets that was nine times larger than the slowest (positive growth) S&Ls.
The number of high fliers also grew rapidly. Some of them were reactive control frauds; others acquired S&Ls that converted from mutual to stock form; the most destructive group acquired failed S&Ls through mergers; and over 100 entered as de novos. Without Gray’s late 1984 moratorium on new California, Texas, and Florida charters’ receiving FSLIC insurance, there would have been hundreds of new control frauds during his term.
The capital rule and other steps that Gray took to toughen supervision reduced growth dramatically. The industry rate of growth fell by more than half in 1985 to 9.5 percent (White 1991, 100). The cuts in growth were, of course, much greater for the control frauds that drove much of the ultrarapid growth in 1983 and 1984. Ponzis have to grow faster over time; Gray’s rule required them to cut their growth rates dramatically. Gray had found their Achilles’ heel, and his reregulation delivered a fatal blow. The Ponzis had only two options. They could drive Gray from office and get a replacement who would not enforce the rule, or they could violate the rule and use political interference to prevent the Bank Board from taking effective enforcement action. The growth rule did not simply impair the existing control frauds. It made the industry a far less attractive vehicle for control fraud and, in conjunction with Gray’s moratorium on new charters, largely ended the entry of new opportunistic control frauds.
So what would have happened in the absence of Gray’s war against the control frauds? There would have been no capital rule and no moratorium on de novos. The industry’s rate of growth would have increased substantially as the control frauds’ much higher growth rate and the entry of hundreds of additional control frauds increasingly drove overall industry growth. The major control frauds that Gray closed in 1986 and 1987 would have remained open and would have grown massively. Remember, without the capital rule’s restriction on growth, Ponzis could have used deposit insurance to grow extremely rapidly and could have continued to report record profits for many years. The Bank Board would not have closed control frauds that reported record profits. By the end of Gray’s term (June 30, 1987), the industry would have had total assets of well over $2 trillion. By the end of his next term (he would have been reappointed had he not reregulated) or Wall’s term, the industry would have had over $6 trillion in total assets. Control frauds would have held a majority of those assets.
The ADC Ponzis would have hyperinflated real estate markets throughout the nation, resulting in a staggering glut of commercial office space. There is no way of knowing in what year the real estate bubble would have collapsed. The only sure things are that it had to collapse and that the more the bubble inflated, the greater the devastation of the general economy. Japan, as I write, has not recovered from the collapse of its twin bubbles in 1990.
The best estimate of the cost of the S&L debacle (measured in 1993 dollars) is $150–175 billion (NCFIRRE 1993a, 4). (You may have read much larger estimates; they are misleading because they treat interest improperly.) An S&L industry with over $6 trillion in assets, and more than half those assets held by control frauds engaged in ADC Ponzi schemes, would have caused trillions of dollars in losses directly to the taxpayers and would have injured national real estate markets badly enough to gravely damage the national economy. If Ed Gray had not become our most unlikely of heroes, that would have been our fate.