The Savings and Loan Crisis: Causes, Cost and Cleanup

Key Takeaways

• The first loss was mechanical and required no misjudgment by anyone. Thrift net income went from $781 million in 1980 to negative $4.6 billion in 1981, and tangible capital fell from 5.3 percent of assets to 0.5 percent by the end of 1982.
• By year-end 1982, 415 thrifts holding $220 billion were insolvent on tangible net worth. Their insurer held $6.3 billion, roughly a quarter of the $25 billion a later commission estimated closing them would have cost.
• Instead of funding closures, the rules changed. Required net worth was cut twice, and in July 1982 the amortization limit on supervisory goodwill went from ten years to as many as forty. Goodwill rose from $7.9 billion in June 1982 to $22 billion by December 1983, by which point it was 67 percent of all regulatory capital.
• Deregulation of lending powers arrived after the industry was already insolvent, which is why it was so expensive. Thrift assets grew 56 percent between 1982 and 1985 against roughly 24 percent at banks, and mortgages fell from 78 percent of thrift assets in 1981 to 56 percent by 1986.
• Texas concentrated the damage. By year-end 1987 insolvent Texas thrifts held 44 percent of the assets in all regulatory-capital-insolvent thrifts nationwide, and unprofitable Texas thrifts accounted for 62 percent of industry losses.
• The final bill was $152.9 billion as of 31 December 1999, of which taxpayers paid $123.8 billion. That estimate fell every year after 1991, which is a fact about resolution accounting rather than about how bad the crisis was.

What Was a Savings and Loan, and Why Was Its Balance Sheet Built to Break?

A savings and loan association, also called a thrift, was a deposit institution created to turn local savings into local home mortgages. The Federal Reserve traces the first one to Pennsylvania in 1831. The federal scaffolding arrived in the 1930s, and one detail in it mattered enormously later: for commercial banks, chartering and deposit insurance sat in separate agencies, while for federally chartered thrifts the Federal Home Loan Bank Board did both. The body that promoted the industry's expansion was also the body that would have to absorb its failures.

By 1980 the industry was large and almost perfectly undiversified. The FDIC records approximately 4,000 FSLIC-insured associations holding $604 billion in assets, plus 590 more with $12.2 billion insured by state programs in Maryland, Massachusetts, North Carolina, Ohio and Pennsylvania. The Federal Reserve puts about $480 billion of the federally insured total in mortgage loans, roughly half of all home mortgages then outstanding in the United States. Four-fifths of these thrifts were mutually owned, so no equity market priced them and no share price could signal distress.

The mismatch was in the charter, not in a trading decision

Assets were thirty-year fixed-rate mortgages, with income locked in at whatever rate prevailed on the day they were written. Liabilities were deposits that could be withdrawn immediately and repriced continuously. No thrift manager chose that duration mismatch as a strategy: it was what the charter permitted them to hold and what the deposit franchise gave them to fund it with. A thirty-year mortgage written at a 1970s coupon carries the duration of a long bond while sitting in the loan book of an institution that never thought of itself as holding bonds, which is why bond duration explained describes the thrift asset side better than any lending manual of the period did. The bond price and yield to maturity calculator will price one of those mortgages at 1981 yields.

The structure survived only while deposits cost less than the mortgage book earned, and federal law held that in place. Regulation Q capped what depository institutions could pay savers: when the ceiling was finally addressed in 1980, the maximum rate on a bank savings account was 5.25 percent while short-term Treasury securities yielded over 12 percent. That gap is the whole setup. It kept thrift funding artificially cheap and gave savers an obvious reason to leave the moment an alternative existed, and the money market mutual fund was that alternative. Policymakers therefore faced a genuine dilemma rather than an oversight. Leave the ceiling and thrifts lose their deposits; lift it and thrifts pay market rates while still holding mortgages written at 1970s coupons. Both roads led to the same loss. The Depository Institutions Deregulation and Monetary Control Act, signed on 31 March 1980, took the second and phased out rate restrictions over six years.

How Did Rising Rates Erase the Industry's Net Worth?

The rate move that did the damage was a deliberate policy choice made for reasons that had nothing to do with thrifts. Inflation had reached 11 percent by June 1979. Paul Volcker became Federal Reserve chairman that August and shifted policy to target the money supply rather than interest rates, and in late 1980 and early 1981 the Fed let the federal funds rate approach 20 percent. Long rates followed: the ten-year Treasury yield rose from about 11 percent in October 1980 to more than 15 percent a year later.

For an institution funding thirty-year mortgages with deposits, that is not a market move to be weathered. It is a permanent transfer of the entire spread, because the mortgage book kept paying its contractual coupon while the deposit book repriced toward the market. The result appears in the FDIC's own industry table with no ambiguity.

Selected statistics for FSLIC-insured savings and loan associations, 1980 to 1989. Dollar figures in billions. Insolvency measured on tangible capital. Source: FDIC, History of the Eighties, Volume 1, Table 4.1.

Three things in that table deserve to be read slowly. The speed of the capital collapse: 5.3 percent of assets in 1980, 4.0 percent in 1981, 0.5 percent in 1982. The insolvency count over those same two years, from 43 institutions holding $0.4 billion to 415 holding $220 billion. And the column on the right, where FSLIC reserves sit between $4.6 billion and $6.5 billion throughout before going negative in 1986 and staying there. The insurer was not getting bigger while the problem was. For scale, in the first three years of the 1980s, 118 thrifts holding $43 billion in assets failed at an estimated cost of $3.5 billion; across the preceding forty-five years, 143 thrifts holding $4.5 billion had failed at a cost of $306 million.

Why no crash, no panic and no famous date

Readers arriving from 1929 or Black Monday will notice this episode has no headline day, no index level and no drawdown percentage. That absence is diagnostic. The loss sat inside deposit institutions, four-fifths of which had no shares to be marked down, and the accounting they reported to their regulator did not require them to recognize it. A depositor whose savings were federally insured had no reason to run, which is exactly what deposit insurance is for. So the crisis had no forcing mechanism. In a funding crisis the calendar is set by creditors and decisions get made in days; here nobody was owed an answer, and the answer took most of a decade. FDIC deposit insurance covers the modern guarantee, and rate-sensitive industries covers which business models still carry this exposure.

Why Did Regulators Not Simply Close the Insolvent Thrifts?

Because closing them cost money the insurer did not have, and appropriating that money was politically impossible in 1983. The National Commission that later investigated the debacle estimated that resolving the institutions insolvent in early 1983 would have cost the FSLIC roughly $25 billion. The FSLIC held $6.3 billion at year-end 1982. Closing what was already broken required about four times the fund's entire reserves, with no way to supply the difference short of an appropriation that would have shown up immediately in the federal deficit.

Running alongside the money problem was an intellectual one. Many officials believed the insolvencies were an artifact of an unprecedented rate spike that would reverse, so a thrift with negative net worth but positive cash flow was not failing, it was waiting. Rates did in fact fall: inflation was back to 5 percent by October 1982, the funds rate returned to around 9 percent, and industry net income turned positive again in 1983. By then the second decision had already been taken.

Redefining insolvency instead of funding it

Unable to pay for closures, the Federal Home Loan Bank Board changed what counted as capital. The changes were individually defensible and collectively devastating.

• The requirement was lowered twice. Required net worth went from 5 percent of insured accounts to 4 percent in November 1980, then to 3 percent in January 1982. The 1980 Act had already replaced the statutory floor with a range of 3 to 6 percent, leaving the choice to the Bank Board.
• A phase-in rule made the effective floor lower still. Associations under twenty years old faced a requirement below 3 percent, and the deposit base in the calculation was a five-year average rather than current deposits. Both rewarded new institutions and fast growth: the FDIC notes that $2.0 million of initial capital could support $1.3 billion of assets by the end of a first year.
• Losses could be spread forward. From September 1981 a thrift selling an asset at a rate-driven loss could recognize it over ten years, carrying the unrecognized portion as an asset. The same month, troubled institutions could issue income capital certificates that the FSLIC bought, usually with notes rather than cash, and count them as net worth.
• Premises could be revalued into capital. From late 1982 the increase in the market value of a thrift's own buildings counted toward reserves.
• Goodwill amortization was stretched fourfold. Effective July 1982 the Bank Board removed the ten-year limit on amortizing supervisory goodwill, allowing up to forty. Since the offsetting discount on the acquired mortgages was accreted to income over roughly ten, the arrangement produced reported earnings from the act of acquiring an insolvent institution.

The last item scaled. Goodwill on thrift balance sheets rose from $7.9 billion in June 1982 to $22 billion by December 1983, at which point it was 67 percent of all regulatory capital in the industry. Two-thirds of the capital standing between the industry and its insurer was an accounting entry created by mergers.

The trap the regulator built for itself. Enforcement action required an institution to be insolvent under regulatory accounting principles, not on tangible capital. Having redefined regulatory capital to include goodwill, deferred losses and appraised premises, the Bank Board had made most of its insolvent institutions technically solvent and therefore untouchable. At year-end 1984 the FSLIC held $5.6 billion against 71 regulatory-capital-insolvent institutions with $14.8 billion in assets, a figure that reached 225 institutions and $68.1 billion two years later. On the tangible measure in the table above, the 1984 count was 695 institutions holding $360.2 billion. Same industry, same day, two answers an order of magnitude apart.

It is worth recording what the Bank Board was working with. In 1984 the average FHLBB examiner earned $24,775, against $30,764 at the Office of the Comptroller of the Currency, $32,505 at the FDIC and $37,900 at the Federal Reserve Board. Examiners reported to Washington while supervisory agents worked for the regional Home Loan Banks, which were owned by the institutions they supervised. Not until 1987 did thrift examiners have authority to classify assets by likelihood of repayment or to compel timely loss reserves.

How Did an Interest Rate Loss Turn Into a Real Estate Loss?

This is the part of the savings and loan crisis with no analogue elsewhere in this library, and it is why the final bill ran to hundreds of billions rather than tens. The policy on offer was not merely forbearance, it was growth. If insolvent thrifts could not be closed and could not be recapitalized, the remaining hope was that they would earn their way back, which meant giving them something more profitable to do than write mortgages. The constraints came off in a consequential order.

Funding was made easy first. The 1980 Act raised federal deposit insurance from $40,000 to $100,000 per account to slow the outflow to money market funds. It also turned the insured deposit into a purchasable commodity: brokers could assemble fully insured funds in $100,000 increments and place them wherever the rate was highest, and the institution paying the highest rate was reliably the one with the least to lose. The National Commission later concluded that federal deposit insurance at institutions carrying substantial risk was a fundamental condition necessary for collapse, and that raising the limit made it worse. Brokered CDs describes the modern descendant of that channel.

Then the asset and ownership rules opened. The 1980 Act expanded federal thrift authority to make acquisition, development and construction loans. Garn-St Germain, signed on 15 October 1982, removed the statutory limit on loan-to-value ratios, so a thrift could lend a developer 100 percent of a project's appraised value; it also permitted nonresidential and variable-rate mortgages and wrote capital forbearance into statute, replacing a numeric reserve requirement with a requirement to hold reserves in a form satisfactory to the FSLIC. In April 1982 the Bank Board had already dropped the rules requiring a federally chartered stock association to have at least 400 shareholders, dispersed ownership and mostly local holders. One owner could now acquire a thrift with very little cash and run it at high leverage. The states then competed to be more permissive still: California's Nolan bill, effective 1 January 1983, gave state-chartered associations unlimited authority to invest in real estate and service corporations, while California's own supervisory staff fell from 178 people in 1978 to 44 in 1983. The FDIC calls this competition in laxity, driven by states unwilling to lose institutions and charter fees to the federal system.

What the industry did with the new freedom

It grew, fast, exactly as an insolvent institution funded by insured deposits would be expected to. Total thrift assets went from $686 billion at year-end 1982 to $1,068 billion at year-end 1985, a rise of 56 percent against roughly 24 percent at banks, with more than $120 billion of net new money arriving in 1983 and 1984 alone. Nearly 500 new charters were issued between 1980 and 1986, more than 200 of them in 1984 and 1985. Stock-form associations went from 21 percent of the industry in 1981 to 38 percent in 1986, controlling 64 percent of its assets, and mortgage loans fell from 78 percent of thrift assets in 1981 to 56 percent in 1986.

Where the losses came from is worth stating carefully, because the popular version gets it wrong. Thrifts did buy junk bonds and finance ski resorts and windmill farms, and those made better copy than what was really happening. The FDIC's own judgment is that high-risk development loans, and the mortgages written on the same properties, were most likely the principal cause of thrift failures after 1982. Two features compounded the danger. Such a loan could be structured so the interest reserve was funded out of the loan itself, meaning a project produced reported income for the lender before anyone knew whether it would be finished or leased. And because so many thrifts entered the same regional markets at once with the same 100 percent loan-to-value authority, they collectively financed the overbuilding whose collapse then destroyed the collateral. That is the pattern in the credit cycle and refinancing risk, running at speed in one asset class in a handful of states.

By 1983, even as falling rates returned traditional thrifts to profitability, 10 percent of the industry was insolvent on generally accepted accounting principles and institutions holding 35 percent of industry assets were insolvent on a tangible basis. Those were the ones permitted to grow fastest.

Why Did Texas Produce So Much of the Damage?

Texas is where all four ingredients arrived together: permissive state law, a regional economy that turned sharply, aggressive growers, and the thinnest examination coverage in the country. The last was partly an accident of logistics. When the Federal Home Loan Bank System's Ninth District relocated from Little Rock, Arkansas to Dallas in September 1983, examinations in the district fell by a third and stayed low through 1984 and 1985, the two years when thrift growth nationally peaked. The concentration in the outcome is stark. By year-end 1987, insolvent Texas thrifts held 44 percent of the assets in all regulatory-capital-insolvent thrifts nationwide, and unprofitable Texas thrifts accounted for 62 percent of all industry losses. In 1988, the peak year for FSLIC failures, more than 40 percent of thrift failures nationally were in Texas.

The Texas premium, and why a sick institution sets the price of money

The most portable idea in this case study comes out of what happened to Texas deposit rates. An institution that is already insolvent has an unusual funding calculus: paying above market for deposits costs it nothing real, because the extra loss lands on the insurance fund rather than on an owner with anything left to lose. A solvent competitor has no such freedom but still has to fund itself. So the sick institutions set the price. Once the condition of the Texas thrift industry was widely known, even well-capitalized Texas banks and thrifts paid what became known as the Texas premium, estimated at 50 basis points or more above rates elsewhere, simply to retain their own deposits. The FDIC quotes a Texas thrift executive describing the resulting bidding wars as just out of control. Forbearance was not a neutral holding pattern. It was actively taxing every solvent competitor in the state.

The premium came down when, and only when, the institutions were dealt with. It peaked in mid-1987 and fell as resolution activity picked up, and by year-end 1989 the average cost of deposits at Texas banks was only eight basis points above the rest of the United States. The Bank Board's Southwest Plan merged some of the highest rate payers into consolidated groups, conserving cash by paying acquirers with notes and future guarantees rather than money. It drew criticism for advertising the tax advantages of acquiring an insolvent thrift before that provision expired at the end of 1988, delivering substantial benefits to wealthy acquirers who contributed very little capital. Underneath it all sat the Economic Recovery Tax Act of 1981, which had made commercial real estate investment unusually attractive and whose withdrawal in 1986 removed that support.

What Did FIRREA and the Resolution Trust Corporation Actually Do?

The turning point was administrative rather than dramatic: on 31 December 1986 the FSLIC was reported insolvent, with negative reserves of $6.3 billion, which became negative $13.7 billion a year later and negative $75.0 billion by the end of 1988. From that date the question was no longer whether public money would be used, only how much and how late. Congress tried the smaller answer first, creating the Financing Corporation on 10 August 1987 to raise money for the FSLIC through long-term bonds serviced by the thrift industry. By the time real legislation arrived two years later, that vehicle had contributed $8.2 billion, against an industry whose tangibly insolvent members held over $290 billion in assets.

The Financial Institutions Reform, Recovery and Enforcement Act, enacted on 9 August 1989, rebuilt the architecture rather than patching it. It abolished the Federal Home Loan Bank Board and the FSLIC, created the Office of Thrift Supervision, moved thrift deposit insurance to the FDIC through a new Savings Association Insurance Fund, and created the Resolution Trust Corporation to dispose of the failed institutions. That separated the promotion of the industry from the insurance of it for the first time since 1934.

The cleanup took longer and cost more than the statute assumed

FIRREA gave the RTC $50 billion and a window covering thrifts placed into conservatorship or receivership between 1 January 1989 and 8 August 1992. Both assumptions failed. Congress legislated three more times, raising authorized funding for RTC losses to $105 billion, and extended the deadline twice, finally to 30 June 1995. The RTC wound up on 31 December 1995.

Thrift resolutions by year and by agency, 1986 to 1995. Assets in billions of dollars, net of valuation allowances on the books at the time of failure. Source: FDIC Banking Review, Curry and Shibut, Table 1.

The shape of the delay is visible across the two halves. The FSLIC resolved almost nothing in 1986 and 1987 relative to the size of the problem, then handled 185 institutions with $96.8 billion of assets in 1988 as the position became untenable. The RTC took 318 institutions in its first partial year and another 213 the next. Close to seventy percent of the entire cleanup by asset value went through between 1988 and 1990, six to eight years after the industry's tangible capital had first gone to nearly zero.

One number here needs care. The Federal Reserve states that the RTC closed 747 thrifts with assets of over $407 billion, while the FDIC's cost study gives $394.0 billion for the same 747 institutions. Both are right: the FDIC figure is explicitly net of valuation allowances already on the books at the time of failure, and the study notes that other published figures are gross of them. A headline asset total for a failed institution is an accounting choice before it is a fact.

The industry that emerged was half the size of the one that went in, falling from 3,234 federally insured institutions on 1 January 1986 to 1,645 at year-end 1995, though the FDIC notes some of that consolidation would have happened anyway. There was also a legal tail. FIRREA's five-year phaseout of supervisory goodwill created before April 1989 pushed thrifts that had been encouraged to acquire failing peers back into insolvency, and on 1 July 1996 the Supreme Court ruled for three of them in United States v. Winstar Corp. The government had changed the terms of the deal it used to get the mergers done.

What Did the Cleanup Cost, and Who Paid It?

The FDIC's own reconciliation, published in 2000 and measured as of 31 December 1999, is the most careful figure available and it is the one this page uses. Total direct and indirect losses from resolving failed thrifts between 1986 and 1995 came to $152.9 billion. United States taxpayers bore $123.8 billion of that, or 81 percent. The thrift industry bore $29.1 billion, or 19 percent.

Estimated cost of resolving the savings and loan crisis as of 31 December 1999, in billions of dollars. Source: FDIC Banking Review, Curry and Shibut.

The split between the two top rows is the price of the delay, in dollars. FSLIC-era resolutions were 35 percent industry-funded, because the industry was still solvent enough to be assessed. RTC-era resolutions were 91 percent public, because by then it was not. Every year that passed shifted the burden from the insurance fund's members to the general taxpayer, for the simple reason that the members had less left to give.

Why you will see a larger number, and why it is also correct

The FDIC's history of the decade, written in 1997, cites a General Accounting Office estimate of $160.1 billion in total resolution costs including $132 billion from federal taxpayers, drawn from the audit of the RTC's 1995 and 1994 financial statements. The Federal Reserve puts the ultimate taxpayer cost as high as $124 billion. All three figures are honest, being the same running estimate at three observation dates, and the direction of travel is instructive: it fell every year after 1991. The RTC had reserved conservatively against the assets it seized, valuing them for markets that were collapsing at the time. As the economy improved, losses came in below those reserves and the excess was recaptured. A crisis cost figure is not a measurement of damage. It is a running estimate of the residual after a disposal programme, and it moves with the market the disposals happen into.

Three absences from that number make it usable rather than rhetorical. It excludes the Texas premium paid for years by solvent institutions across the Southwest. It excludes the losses to commercial banks from the same overbuilt real estate markets. And it excludes the goodwill litigation, still unresolved when the study was written. The $152.9 billion is the deposit insurance bill, not the economic cost of the episode.

Which Signals Were Legible at the Time, and Which Only After the Cleanup?

This episode inverts the usual hindsight problem, and that inversion is the most useful thing in it. In 2008 the aggregate data was genuinely unavailable: no disclosure let an outsider estimate how much securitized mortgage risk sat on which balance sheet or how much overnight funding supported it. Here the aggregate data was published, unambiguous and stark from 1982 onward. What was missing was not information.

Thrift crisis signals classified by whether they were observable at the time and whether observing them helped.

The honest reading of that table is uncomfortable in a different way from the usual crisis retrospective. The savings and loan crisis was not a failure of prediction. Anyone reading the industry's tangible capital ratio at the end of 1982 knew that a large fraction of the industry was gone and that the insurance fund could not pay for it. The failure was that recognizing the loss required an appropriation nobody would make, so the definition of the loss was changed instead.

Hindsight check. Ask of any thrift-era warning sign: would acting on it have required new information, or only a willingness to accept a loss that was already published? For tangible capital, the FSLIC reserve balance and the Texas deposit premium, the answer is the second. For identifying specific fraudulent institutions and for the final cost, it is genuinely the first. Conflating them produces two opposite errors, one holding that crises are unknowable and one holding that they are obvious. Because the decisive thrift number was published, ignored, and then vindicated, this episode is unusually good raw material for the second error, and cognitive biases in trading covers the machinery behind it.

Common Myths About the Savings and Loan Crisis

"It cost half a trillion dollars." The FDIC's cost study opens by noting that published reports had put the figure anywhere from under $100 billion to as high as $500 billion, and settling that was its purpose. The reconciled answer as of 31 December 1999 is $152.9 billion total and $123.8 billion to taxpayers. Larger figures typically add decades of interest on the bonds issued to fund the cleanup, which is a defensible thing to calculate but a different quantity, and it is rarely labelled as such.

"Deregulation caused it." Only the second half. The first loss happened while the industry was tightly regulated, and because of that regulation: a mandated model of long fixed-rate mortgages funded by rate-capped deposits cannot survive rates near 20 percent. Deregulation arrived afterwards, as the remedy, and determined how much larger the bill became. The FDIC's formulation is more precise than the slogan: deregulation of asset powers was enacted without the deposit insurance reform and supervisory resources recommended alongside it.

"It was all fraud." Fraud was real and is what people remember. But the FDIC is explicit that while windmill farms and casinos made for interesting reading, high-risk development loans and the mortgages written on the same properties were most likely the principal cause of thrift failures after 1982.

"The insured depositors should have taken losses." The insurance limit was statutory and had been raised to $100,000 by Congress in 1980. What was genuinely open was not whether the guarantee would be honoured but who funded it and when, and delay moved the burden from the industry's own assessments to the general taxpayer. The cost table shows it: FSLIC-era resolutions were 35 percent industry-funded, RTC-era resolutions 9 percent.

"The thrift industry recovered." It did not recover, it halved, from 3,234 federally insured institutions at the start of 1986 to 1,645 by the end of 1995. Unlike an index, an industry has no obligation to return to its previous level. Recovery statistics of the kind that appear in the dot-com episode or the 2020 crash have no equivalent here.

"FIRREA fixed it, so the mechanism is gone." FIRREA abolished the institutions that failed and separated the chartering of thrifts from the insuring of them, removing the specific conflict at the heart of this episode. It did not repeal duration. A deposit institution holding long-dated fixed-rate assets still takes an economic loss when rates rise sharply, whether or not accounting rules oblige it to report one, and that exposure reappeared when rates rose again in the 2022 rate shock. Mortgage-backed securities covers how the modern version of that asset behaves.

What a Reader Can Actually Carry Forward

The savings and loan crisis is not a template for spotting the next crisis, because the institution at its centre no longer exists in that form. It is something more useful: a clean natural experiment in what happens between the moment a loss occurs and the moment somebody writes it down.

What generalizes

• A book value not adjusted for a rate move is not a solvency statement. The thrift industry reported positive regulatory capital while carrying mortgages worth far less than par. Wherever an entity holds long-dated fixed-rate assets and is not required to mark them, its stated capital describes accounting policy rather than ability to pay.
• Losses do not stop accruing while a decision is postponed. Between tangible insolvency in 1982 and FIRREA in 1989, the bill grew from something a $25 billion appropriation would have covered to $152.9 billion. Delay was not free optionality on a recovery. It was a position financed at an unfavourable rate.
• Guaranteed funding lets a broken institution outbid a sound one. This is the Texas premium generalized. Where a funding source is insensitive to the borrower's condition, the weakest participant sets the price and the strongest pay it. An elevated deposit rate at an institution that should not need one is a solvency signal.
• Growth is the tell, not the reassurance. An industry with almost no capital grew 56 percent in three years. Rapid growth funded by liabilities carrying no discipline is closer to a definition of the problem than to evidence against it. See risk management for the portfolio-level version.
• Accounting definitions are policy instruments and get used as such. Amortization periods, deferred loss recognition and what counts toward a capital ratio decide whether an institution is legally alive. When one of them changes during a period of stress, that is usually the most important news in the announcement.

What does not generalize

• The charter and its regulator. One agency both promoting an industry and insuring it, with underpaid examiners and a structure giving nobody direct responsibility for a troubled institution, is not the current arrangement.
• The rate path. A federal funds rate approaching 20 percent was the product of a specific and unrepeated disinflation, not a stress scenario to build around.
• The absence of a market signal. Four-fifths of thrifts in 1980 were mutually owned and had no share price. Most institutions carrying this exposure today are publicly traded, which changes how fast a problem becomes visible and how fast funding reacts.

The one question worth asking now

Rather than asking whether something you hold could suffer a thrift-style rate loss, ask the question that actually decided this episode: if the loss appeared, who would have to fund the recognition of it, and does that party have both the money and the incentive to act promptly? The savings and loan crisis was not expensive because the initial loss was large. That loss was arithmetic on a $480 billion mortgage book and would have been absorbed. It was expensive because the only party that could act had neither the reserves nor the appetite, and the intervening years were spent building a much bigger loss on top of the first one. That question is answerable today, from published accounts and public statute, and it requires forecasting nothing.

References

Every figure on this page was verified against the following sources, each retrieved on 26 August 2026:

• Federal Reserve History: Savings and Loan Crisis: the 1831 Pennsylvania origin; the 1980 industry size and its share of home mortgages outstanding; the 1983 estimate of $25 billion against FSLIC reserves of $6 billion; the deposit insurance increase to $100,000; the 56 percent asset growth from 1982 to 1985 against 24 percent at banks; 1988 as the peak FSLIC failure year with more than 40 percent of failures in Texas; the FIRREA reforms; the RTC closing 747 thrifts with assets of over $407 billion and terminating on 31 December 1995; and taxpayer cost estimated as high as $124 billion.
• FDIC: History of the Eighties, Volume 1, Chapter 4, The Savings and Loan Crisis and Its Relationship to Banking: the industry statistics table reproduced above; the 1980 institution and asset totals; the failure counts for 1980 to 1982 against the preceding 45 years; every capital and accounting change described here; the examiner salary comparison and the regulatory-capital insolvency counts; the growth, charter and portfolio-composition figures; the California Nolan bill and supervisory staffing; the Ninth District relocation; the Texas concentration figures, the Texas premium and the Southwest Plan; the NCFIRRE and GAO cost estimates; and the 1996 Winstar decision.
• FDIC Banking Review: The Cost of the Savings and Loan Crisis, Truth and Consequences: the resolution table reproduced above; the 1,043 institutions with $519 billion in assets; the decline in federally insured thrifts from 3,234 to 1,645; the year-end 1986 insolvency of the FSLIC; FICO; the FIRREA date, the RTC's funding and its extended window; the cost breakdown and the $152.9 billion total; the range of published estimates; the 1986 withdrawal of the 1981 federal tax provisions that had favoured commercial real estate; and why the estimate declined after 1991.
• Federal Reserve History: Depository Institutions Deregulation and Monetary Control Act of 1980: the signing date of 31 March 1980, the Regulation Q ceiling of 5.25 percent against short-term Treasury yields over 12 percent, and the expansion of deposit insurance from $40,000 to $100,000.
• Federal Reserve History: Garn-St Germain Depository Institutions Act of 1982: the signing date of 15 October 1982, the new authority for nonresidential and variable-rate mortgages, and the account of weak thrifts using expanded powers to gamble for recovery.
• Federal Reserve History: Recession of 1981-82: inflation reaching 11 percent in June 1979, the federal funds rate approaching 20 percent in late 1980 and early 1981, the ten-year Treasury yield rising from about 11 percent in October 1980 to more than 15 percent a year later, and inflation falling to 5 percent by October 1982.

Figures deliberately not stated. This page gives no peak level for the three-month Treasury bill, no mortgage rate, no count of criminal prosecutions, no cost figure in present-day dollars or as a share of gross domestic product, and no state-level failure counts beyond the percentages quoted, because no source verified in this session supplied them. It names no individual institution or executive. Each omission above is a source check that came back empty rather than an oversight: the thrift-era mechanism is described either way, but the figure is left blank instead of being filled in from a secondary retelling.

Method note: the tables reproduce published FDIC tables without recalculation, other than converting the resolution table's assets from millions to billions rounded to one decimal. Dollar figures are nominal and not adjusted for inflation. Where two official sources give different values for the same quantity, both are shown and the difference explained rather than reconciled by preference.

This is educational content about the thrift crisis of the 1980s and its cleanup. It is not investment advice, it is not a forecast, and nothing here should be read as a claim about how any future banking or credit episode will behave.

Frequently Asked Questions