In principle, unsecured creditors of banks and other financial inst.i.tutions could impose market discipline; their funds, after all, could be jeopardized if the inst.i.tutions took too much risk. But in the recent crisis, even these unsecured creditors did not impose market discipline. The reasons were various: the unsecured claims were too small to make a difference; the unsecured creditors were treated mostly like secured creditors (insured depositors) and did not experience losses as they were bailed out; the lender-of-last-resort support of central banks prevented the working of market discipline.

Not all financial inst.i.tutions are covered by deposit insurance, but if there's one lesson of financial crises that does get remembered by everyone, it's that when the going gets rough, a lender of last resort will appear to save the day. Ever since the Great Depression, central banks have consistently stepped into the breach and acted as the lender of last resort. It happened in the LTCM crisis in 1998, when the New York Fed orchestrated a private bailout, and it happened again in the midst of the recent crisis, when the Federal Reserve made unprecedented levels of liquidity available to inst.i.tutions like investment banks and others that fell outside the aegis of deposit insurance.

Knowing that a probable lender of last resort existed reduced financial inst.i.tutions' incentive to hold a large amount of liquid a.s.sets as a buffer against a bank run. It also helped remove any remaining incentive that the de facto depositors had to monitor these inst.i.tutions' performance: should a crisis strike, they knew, central banks around the world could be counted on to save the day. And in this respect, the calculations of all the financial system's players proved remarkably correct: in both the United States and abroad, central banks fell all over themselves to provide lifelines to ailing firms. There was one dramatic attempt to address the problem of moral hazard-by letting Lehman Brothers fail-followed by frantic efforts to bail out huge swaths of the financial system.

If there was ever an argument for tightly regulating banks and other financial firms, this is it. Banks have to be forced to hold enough liquidity, and shareholders must have enough skin in the game and an incentive to monitor the firms they supposedly supervise. And these requirements mean that government must play a major, if controversial, role. Unfortunately, in the years leading up to the crisis, government was nowhere to be found. In fact, government helped foment this crisis, not merely through its absence, but through its not-so-subtle interventions as well.

Government and Its Discontents.

The Federal Reserve is arguably the most powerful instrument of government control over the economy. Its power can be used for good or for ill, as the career of Alan Greenspan amply suggests. That Greenspan presided over the Federal Reserve is ironic. After all, as a young man he became smitten with the power of the free market. In the 1950s he even became an acolyte of Ayn Rand, whose hard-core libertarian beliefs he admired. Yet Greenspan's growing conviction that government should stay out of the economy did not prevent him from serving in government when the opportunity arose.

Greenspan's first major appointment came in 1974 as chairman of President Gerald Ford's Council of Economic Advisers. But this service paled in importance compared to his 1987 appointment as chairman of the Federal Reserve. His ambivalence about government's role in regulating the free market was evident from the beginning. Four months after his appointment, the stock market crashed, and Greenspan immediately rode to the rescue. Out the window went any principled opposition to government intervention. As he memorably put it, "In a crisis environment . . . we shouldn't really focus on longer-term policy questions until we get beyond this immediate period of chaos." If Greenspan could acknowledge that the central bank had a role to play in mitigating the effects of a financial crisis, he declined to do anything to prevent such crises from developing. He seems to have had little interest in a long-standing central banking philosophy that these powerful inst.i.tutions should prevent bubbles from forming in the first place. Nicely summarizing that belief, former Federal Reserve chairman William McChesney Martin, Jr., once said that the job of the central banker was to "take away the punch bowl just as the party gets going."

Greenspan revealed himself to be unwilling to take it away. In 1996, as the stock market spiraled into a giddy bubble focused on tech and Internet stocks, he warned of "irrational exuberance," then did next to nothing to stop the bubble from inflating, aside from inst.i.tuting a token increase of 25 basis points in the Federal funds rate. When the dot-com bubble finally popped in 2000, Greenspan poured plenty more alcohol into the proverbial punch bowl. In the wake of the attacks on September 11, he kept cutting the funds rate, even after signs of a recovery started to appear. When he finally resumed raising rates in 2004, he did so in tiny and slow and predictable (a policy of "measured pace" tightening) increments of 25 basis points every six weeks, when the Federal Open Market Committee met. This policy kept rates too low for too long and normalized them too late and too slowly.

The result was the housing and mortgage bubble. By pumping vast quant.i.ties of easy money into the economy and keeping it there for too long, Greenspan muted the effects of one bubble's collapse by inflating an entirely new one. This policy was the inevitable consequence of the contradiction at the heart of his approach to central banking: helplessly watching bubbles on the way up, and moving frantically to arrest the downward slide. Unfortunately, it created a Greenspan put. By the end of Greenspan's final term as chairman of the Federal Reserve, the Greenspan put was an article of faith among traders: the markets believed that the Fed would always ride to the rescue of reckless traders ruined after a bubble collapsed. It created moral hazard on a grand scale, and Greenspan deserves blame for it.

Greenspan also deserves blame for refusing to use the power of the Federal Reserve to regulate markets. For example, in 1994 Congress pa.s.sed the Home Ownership and Equity Protection Act in order to crack down on predatory lending practices. Under its terms, Greenspan could have regulated subprime lending, but he refused to do so. He continued to refuse even after Edward Gramlich, one of the members of the Federal Reserve Board, beseeched him to do so. Greenspan later defended his refusal to monitor subprime lenders: "For us to go in and audit how they act on their mortgage applications would have been a huge effort, and it's not clear to me we would have found anything that would have been worthwhile without undermining the desired availability of subprime credits."

Revealing words, these. Greenspan considered the advent of subprime lending to be entirely a good thing, the inevitable consequence of letting markets run free. Until very recently he continued to praise the role that financial "innovation" played in making credit available to growing numbers of Americans. At one public event in 2005 he hailed the way financial innovation had "led to rapid growth in subprime mortgage lending, . . . fostering constructive innovation that is both responsive to market demand and beneficial to consumers."

In all fairness, Greenspan had plenty of company in the relentless drive toward deregulation. For the previous three decades, freeing financial markets from "onerous" regulations had been an article of faith among conservatives. It also became public policy. From the 1980s onward, tight regulations of the financial system inst.i.tuted during the Great Depression were phased out or eliminated.

The most notable casualty was the Gla.s.s-Steagall Act of 1933. Part of that landmark legislation had created a firewall between commercial banks (which took deposits and made loans) and investment banks (which underwrote, bought, and sold securities). Those provisions suffered death by a thousand cuts. Beginning in the late 1980s, the Federal Reserve Board permitted commercial banks to buy and sell a range of securities. At first commercial banks could derive only 10 percent of their profits from securities operations, but in 1996 the Federal Reserve Board raised that threshold to 25 percent. The following year Bankers Trust became the first commercial bank to purchase a securities firm; other banks soon followed suit.

The catalyst for the final repeal of Gla.s.s-Steagall was the proposed merger of Travelers with Citicorp. This combination, which brought commercial banking, insurance underwriting, and securities underwriting under the same roof, forced the issue: the new financial behemoth was illegal under existing laws. Late in 1999, after intense lobbying, Congress repealed the remnants of Gla.s.s-Steagall via the Financial Services Modernization Act, paving the way for additional mergers between investment banks, commercial banks, and insurers.

One of the key players in the repeal of Gla.s.s-Steagall was Republican economist-turned-senator Phil Gramm. Gramm continued to lead the crusade against financial regulation, most famously in 2000, when he attached the Commodity Futures Modernization Act to a budget bill. This act, which was never debated in the Senate or the House, effectively declared huge swaths of the derivatives market off-limits to regulation. Among the instruments thus removed from regulation were credit default swaps, which permitted a purchaser to buy "insurance" to protect against defaults on bonds both very simple (such as those issued by an automaker) and extremely complex (collateralized debt obligations backed by pools of mortgage-backed securities). Credit default swaps, which mushroomed to reach a notional value of over $60 trillion by 2008, became one of the most important sources of "systemic risk"-perils that threaten the entire financial system. (For more on credit default swaps, see chapter 8.) The push for deregulation also took place outside Congress. In 2004 the five biggest investment banks lobbied the Securities and Exchange Commission (SEC), hoping to persuade it to loosen rules that restricted the amount of debt their brokerage units could a.s.sume. Obtaining an exemption would allow the firms to tap billions of dollars. .h.i.therto held in capital reserve should they sustain major losses on their investments. It would allow cuts in the cushion these firms had maintained, even as it magnified their potential for profits. In a unanimous decision, the SEC complied with the banks' request, though not without some recognition that the move might be risky. "We've said these are the big guys," observed one commissioner at the spa.r.s.ely attended hearing, "but that means if anything goes wrong, it's going to be an awfully big mess." Investment banks reacted to this deregulation by ma.s.sively increasing their leverage (ratio of a.s.sets to capital) to ratios of 20, 25, or even more, well above the ratio of 12.5 imposed on the more regulated commercial banks.

Not everyone thinks that deregulation alone is to blame for the crisis. Some conservative commentators have claimed that it was the product of too much government, not too little. The key claim here is that the Community Reinvestment Act of 1977 helped inflate the bubble. That piece of legislation, which prevented banks from discriminating against low-income neighborhoods when they made loans, made it easier for the poor and minorities to obtain mortgages. In the conservative interpretation, the original and the amended legislation-with the a.s.sistance of Fannie Mae and Freddie Mac-helped spur the subprime market and caused the eventual meltdown.

It's an interesting argument but misplaced. The huge growth in the sub-prime market was primarily underwritten not by Fannie Mae and Freddie Mac but by private mortgage lenders like Countrywide. Moreover, the Community Reinvestment Act long predates the rise of the housing bubble. True, legislation pa.s.sed in the 1990s compelled Fannie Mae and Freddie Mac to purchase mortgages that effectively included subprime loans. In 1997, for example, some 42 percent of the loans they purchased had to come from borrowers whose income was below the average for their neighborhood. Some of these loans were subprime, though the precise number is not known with certainty. Regardless, overblown claims that Fannie Mae and Freddie Mac single-handedly caused the subprime crisis are just plain wrong.

What is true is that the federal government has long sponsored and subsidized home ownership, making it a far less expensive and burdensome proposition than it would otherwise be. Its subsidies include allowing homeowners to deduct property taxes and mortgage interest payments on their federal income tax returns. Similarly, it does not tax a certain portion of capital gains from the sale of a primary home. Most important of all, several government-sponsored enterprises-not only Fannie Mae, Freddie Mac, and FHA but the Federal Home Loan Banks, among others-support and subsidize the housing and mortgage market. These subsidies may not have caused the housing bubble, but they certainly created conditions that encouraged and sustained its growth.

The Shadow Banks.

If government policies helped inflate the bubble, and deregulation helped remove existing constraints on financial firms, the failure of government to keep pace with financial innovation also played a role. This failure goes far beyond merely neglecting to regulate exotic derivatives, or leaving the bonus system favored by the financial services industry untouched. It goes to the heart of the dramatic if unheralded rise, over the past thirty-plus years, of what Pacific Investment Management Company's Paul McCulley dubbed the "shadow banking system."

The shadow banking system consists of financial inst.i.tutions that look like banks, act like banks, and borrow and lend and invest like banks, but-and here's the important part-are not regulated like banks. Think for a moment about what const.i.tutes a bank. In simplest terms, a bank borrows money on a short-term basis, usually in the form of deposits "lent" to it by depositors. These deposits const.i.tute most of the bank's liabilities: at any moment the depositors can demand their money, and the bank has no choice but to give it back.

But banks don't just sit on deposits; they lend them out in the form of mortgages and other long-term investments, such as a ten-year loan to a corporation. In other words, they borrow the deposits, make loans, and thereby make a profit for themselves via the interest they charge. However, there's a catch: while the bank's liabilities are liquid (they're in the form of deposits), its a.s.sets are illiquid (they're tied up in land, new equipment on a factory floor, and other things that can't immediately be turned into cash).

Normally, this disparity isn't a problem; it's highly unlikely that all the depositors will rush to the bank at once, demanding their money back. But occasionally they do precisely that, and the Great Depression is the example of what happens when panicked depositors flood a bank. The perils of this dynamic were beautifully dramatized by Frank Capra's It's a Wonderful Life, which profiles the ups and downs in the life of small-town banker George Bailey.

One day, as Bailey is besieged by anxious depositors demanding their money back, he gives an impromptu lesson on banking. "You're thinking of this place all wrong," he tells the depositors, who cling to the idea that their money is simply lying idle in the vaults. "As if I had the money back in a safe," he remonstrates. "The money's not here. Your money's in Joe's house. . . . And in the Kennedy house, and Mrs. Backlin's house, and a hundred others." The liquid deposits, in other words, have been transformed into illiquid investments that are not readily converted back to cash. As Bailey explains to the depositors, "you're lending them the money to build, and then, they're going to pay it back to you as best they can."

Bailey's predicament was typical of banks in the darkest hours of the Great Depression. He was grappling with the "maturity mismatch" between liabilities that are short-term "demand deposits," and a.s.sets held for the long term that can rarely be turned into cash on the spur of the moment. As a consequence, it's next to impossible to use the one to pay off the other without incurring tremendous costs. A bank caught in a run might sell off its a.s.sets, such as mortgages and other loans it has made. Unfortunately, if a general panic has seized the banking system, every bank will be trying to do the same thing, and these sales will fetch only a fraction of what they would command in normal times.

So in practice, banks that fall victim to liquidity runs can swiftly go from being illiquid to being insolvent. Sometimes banks deserve that fate, as when their a.s.sets are insufficient to accommodate the demands of depositors, regardless of the price at which they are sold. But in plenty of other cases, a bank is solvent but has simply made illiquid investments. As a consequence, its short-term liabilities far outweigh its liquid a.s.sets. During the Great Depression, banks failed for both reasons. Some could never have made good on the obligations to their depositors, whether there was a bank panic or not. Others could have made good on their obligations if they'd had help.

That help could have come in two forms: lender-of-last-resort support and deposit insurance. The first was available during the Great Depression, but the Federal Reserve failed to use it effectively; the second came into being when New Deal banking legislation created the Federal Deposit Insurance Corporation (FDIC). These two antidotes to bank runs are slightly different. Lender-of-last-resort support stops a bank run by giving banks ready access to cash so they can pay off their depositors, thus sparing them having to liquidate a.s.sets at fire-sale prices. Deposit insurance, by contrast, prevents bank runs from occurring in the first place: it rea.s.sures depositors that they will get their money back if the bank becomes illiquid or even insolvent.

In the postwar era, both lender-of-last-resort support and deposit insurance became the norm, not only in the United States but in most capitalist nations. These protections came at a cost for the partic.i.p.ating banks: they had to give up some of their autonomy in order to avoid the problem of moral hazard. They thus submitted to regulation and supervision in the form of controls on their liquidity, leverage, and capital, which necessarily limited how much money they could make. As a result, banking became a rather humdrum if dependable business. A running joke had it that banking operated according to the 3-6-3 rule: bankers paid their depositors 3 percent interest, lent it out at 6 percent interest, and lined up to tee off at the golf course by three P.M. A slight exaggeration, perhaps, but the joke had more than a grain of truth.

As if that weren't enough to tame banking, international regulations imposed further restrictions. In 1974 central bank governors from the countries that made up the G-10 established the Basel Committee on Banking Supervision, named after the Swiss city that is home to the Bank for International Settlements, a linchpin of the global financial system. In 1988 the committee introduced a capital adequacy system that laid down methods for determining the relative risks of different kinds of a.s.sets held by banks around the world. This system, called the Basel Capital Accord, spelled out in no uncertain terms how much capital banks had to hold, relative to the risk of the a.s.sets in their custody. The core of the agreement held that banks had to maintain a minimum capital standard of 8 percent, that is, hold reserves equivalent to or exceeding 8 percent of the total value of their "risk-adjusted a.s.sets" (which meant that riskier a.s.sets would incur higher capital charges). Though the committee had no legal authority over member nations, its recommendations were adopted in most countries throughout the world.

The committee did not rest on its laurels: in succeeding years it issued additional recommendations. The stakes of those revisions were clear. As one 1997 report of the committee noted, "Weaknesses in the banking system of a country, whether developing or developed, can threaten financial stability both within that country and internationally." That spirit informed a revision to the Basel Capital Accord in 2006, known as Basel II. Unlike the first accord, not all the recommendations of Basel II have been implemented. (For more on the Basel accords, see chapter 8.) Why? Simply put, not everyone in banking was looking for stability and security. A growing number of people who joined the financial services industry from the 1980s onward realized that they could make plenty of money, so long as they were willing to walk the banking tightrope without a safety net underneath. There were ways to conduct banking free of regulations, but also free of the protections afforded ordinary banks. So began a game of "regulatory arbitrage," the purposeful evasion of regulations in pursuit of higher profits. This quest gave rise to the shadow banks.

The shadow banks didn't have tellers; they didn't stand on street corners in neighborhoods across the country. They had funny acronyms, or what Paul McCulley aptly called a "whole alphabet soup of levered up non-bank investment conduits, vehicles, and structures," many of which lurked off the balance sheets of conventional banks. The shadow banks came in all shapes and sizes: nonbank mortgage lenders; structured investment vehicles (SIVs) and conduits, which financed themselves with complex short-term loans known as a.s.set-backed commercial paper; investment banks and broker dealers, which financed themselves with overnight "repos," or repurchase agreements; money market funds, which relied on short-term funds from investors; hedge funds and private equity funds; and even state- and local-government-sponsored pools of auction-rate securities and tender option bonds, both of which had to be rolled over at a variable rate in weekly auctions. Most of these shadow banks had one thing in common: a profound maturity mismatch. They mostly borrowed in short-term, liquid markets, then invested in long-term, illiquid a.s.sets. They looked quite different from Bailey Bros. Building & Loan, but they suffered from the same vulnerability to a bank run.

This would not have been a problem had the shadow banks made the same bargain as regular banks, submitting to increased regulation in exchange for access to lender-of-last-resort support and the equivalent of deposit insurance. But they did not. Even worse, these inst.i.tutions grew to rival the conventional banking system, lending comparable amounts of money. It's little wonder that the shadow banking system was at the heart of what would become the mother of all bank runs.

A World Awash in Cash.

All of these factors-financial innovation, failures of corporate governance, easy monetary policy, failures of government, and the shadow banking system-contributed to the onset of the crisis. In many aspects, the United States and other countries in the English-speaking world took the lead. But the rest of the world helped set the stage for the crisis, even if it was hardly their intent to do so.

Alan Greenspan was one of the first to recognize the problem. He rightly noted that when he raised the federal funds rate from 1 percent to 5.25 percent between 2004 and 2006, long-term interest rates and fixed-rate mortgage rates barely moved. Greenspan's belated policy of monetary tightening had had no effect. This was not what the textbooks would have predicted. In theory, long-term interest rates and mortgage rates should have crept upward in tune with the rate hikes.

Greenspan called it the "bond market conundrum," but it turned out to have an explanation. In an integrated world economy, the rates at which the United States could borrow money were increasingly determined in global markets. And in global markets there was a surplus of savings from j.a.pan, Germany, China, and a range of emerging economies. All that savings had to be invested somewhere, and in the end, it went into purchasing debt generated by the United States. But the low rates of return on the federal government's short-term and long-term debt understandably caused investors to prefer debt paying a higher rate of return. So they purchased the debt of Fannie Mae and Freddie Mac, along with the mortgage-backed securities guaranteed by those inst.i.tutions. All were implicitly guaranteed by the U.S. Treasury.

But overseas investors did not stop there. Private creditors of the United States-particularly investors and financial inst.i.tutions in Europe-became major purchasers of securitized products. Estimates vary, but between 40 and 50 percent of the securities generated by American financial inst.i.tutions ended up in the portfolios of foreign investors. In other words, the income stream from credit card debt, home equity loans, auto loans, student loans, and mortgages ended up in the portfolios of foreign investors via the process of securitization. By making those purchases, foreign creditors helped finance the borrowing binge that drove the bubble.

Just how much foreign investors underwrote the boom remains an open question. Much rides on the answer: some commentators have used the "global savings glut" hypothesis to blame the crisis on China and the other creditors of the United States. That misplaced a.n.a.lysis wrongly shifts the blame from problems in the United States. But what is indisputable is that this pool of savings in search of investments ended up in the United States. In the process, it inadvertently helped the United States live far beyond its means for far too long. Indeed, had the United States been an emerging economy instead of the world's sole superpower, its creditors would have pulled the plug long ago.

But they didn't. Instead, easy money poured into the United States, and this powerful global trend sustained the boom. Combined with lax monetary policy, reckless financial innovation, the problems of moral hazard and poor corporate governance, and the shadow banking system, easy foreign money helped brew a disaster of epic proportions. Still, none of these developments could alone cause a crisis. An essential, additional factor made a catastrophe all but inevitable: the fact that almost everyone connected to the financial system was increasingly reliant on debt or leverage.

The Lure of Leverage.

Let's recall Minsky's taxonomy of borrowers (see chapter 2). The most conservative are the hedge borrowers, whose short-term income flow can cover payments not only of interest but of princ.i.p.al too. More risky are the speculative borrowers, whose income can cover only payments on interest; they must roll over the princ.i.p.al each time it comes due. Most dangerous are the Ponzi borrowers, who can service neither the princ.i.p.al nor the interest and must take on new debt just to stay afloat.

Minsky grasped an essential truth: that an economy would become vulnerable to collapse should its various players resort to debt to finance their activities. He believed that the greater the reliance on debt and leverage, the more fragile the financial system.

Leverage has been on the increase for years. From 1960 to 1974 the leverage ratios of banks in the United States increased by some 50 percent. This process only accelerated from the 1980s onward. Look, for example, at the statistics on debt for households, financial inst.i.tutions, and the other corporations that const.i.tute the private sector. In 1981 the debt of the U.S. private sector was equal to 123 percent of the gross domestic product (GDP); by the end of 2008 it had soared to 290 percent.

Debt soared in every part of the private sector. The corporate sector was the most prudent: its total debt increased from 53 percent to 76 percent of GDP. Households showed less restraint. In 1981 household debt in the United States was 48 percent of GDP, but by 2007 it had risen to 100 percent. The household-debt-to-disposable-income ratio went from 65 percent in 1981 to a staggering 135 percent by 2008. Much of this debt came in the form of leverage in the housing sector, as home buyers purchased increasingly expensive homes with less and less of their own equity. Indeed, at the height of the housing boom, homes could be purchased with no money down, thanks to "innovations" like piggyback loans and other contrivances.

But if debt increased among households and corporations during this period, the financial sector came to rely on debt in a big way: between 1981 and 2008 its debt went from 22 percent of GDP to 117 percent, more than a fivefold increase. The use of debt to supplement investment is known as leverage. For example, an investment bank that finances the purchase of $20 million worth of mortgage-backed securities by putting up $1 million of its own capital and borrowing the other $19 million is leveraged at the rate of twenty to one.

Leverage comes in many flavors. Plain-vanilla leverage is the sort just described, but embedded leverage offers the potential for gains (and losses) that are many times the value of the underlying a.s.set. For example, CDOs, as we have seen, come in different tranches, the riskier ones taking the brunt of the losses in the event things go south. That means in practice that for holders of certain slices of the CDO, losses are magnified to astonishing levels; a tiny loss on the underlying portfolio can hit certain investors particularly hard. Much of this kind of leverage is invisible to the market at large; there's no way to measure it, but when it unravels, the consequences can be dramatic.

Then there's systemic or compound leverage, in which one opening bit of leverage becomes the tip of a vast inverted pyramid of debt. Suppose, for example, that a wealthy individual borrows $3 million from a bank, adds $1 million of his own equity, and invests it into a "fund of funds" that invests in other hedge funds. At this point he has leverage of four to one. Then suppose this fund of funds takes that $4 million and borrows another $12 million from another bank, and sinks it into yet another hedge fund. Again, the leverage is still only four to one, but the initial stake of $4 million has grown to $16 million. Now imagine that this hedge fund borrows another $48 million-again, a leverage of four to one-to invest a total of $64 million in some high-risk tranches of a CDO. In an ill.u.s.tration of the power of exponential math, a tiny initial stake of $1 million has become the basis of a $64 million bet.

That's all fine if the value of those securities holds steady or, better yet, increases. But it's altogether different if the value of the a.s.sets declines. Consider the simple example of leverage given above, where the investment bank is levered twenty to one. Recall that the investment bank's equity totals a million dollars. But let's say the value of the a.s.set falls from $20 million to $19 million, a drop of 5 percent. Should that happen, the investment bank's equity has been wiped out; the effective return on its investment is a rather unpleasant negative 100 percent. The same logic applies no matter what the level of leverage. If the leverage ratio was one hundred to one (a dollar of equity for every ninety-nine dollars of debt) even a minuscule drop of 1 percent in the value of the a.s.set wipes out the underlying equity.

To make matters worse, lenders often expect leverage ratios to remain constant, even when the a.s.set purchased with the loan loses some of its value. Say, for example, that a hedge fund borrows $95 million from an investment bank, kicks in $5 million of its own money, and purchases a CDO worth $100 million. Then the market price of the a.s.set falls to $95 million. The equity has been wiped out. That's not necessarily a problem: the price may rebound to $100 million in due course. But the investment bank may get worried and make a margin call, demanding that the original leverage ratio be restored. That means that the hedge fund must come up with $4.75 million in new equity ($4.75 million is one twentieth of $95 million). If the hedge fund can raise the money, things may work out. If not, the hedge fund has to sell the a.s.set at $95 million and see its equity wiped out.

That alone is hardly a tragedy: margin calls happen, and people (or hedge funds) lose their shirts all the time. The bigger problem is the risk that the hedge fund scrambling to raise equity is not alone. What if many such hedge funds and other financial inst.i.tutions suddenly have to answer a margin call? It could happen if the a.s.set that everyone is using leverage to obtain is the object of a speculative bubble, and prices have risen to unsustainable levels. In the recent crisis, that a.s.set was real estate, not only land and buildings but the exotic securities that derived their value from timely mortgage payments.

When the value of an a.s.set plateaus, then falls-say, because some sub-prime loans go sour, and the revenue stream of a CDO slows to a trickle-the effect ramifies throughout the financial system. Suddenly countless investors are seeing their $100 million CDO fall in value to $95 million. Suddenly all of them get margin calls, demanding that they put up more equity. Perhaps some of them can raise it, but many more will be forced to sell their CDO at whatever the market will bear. And if too many of them do this at once, the CDO may no longer fetch $95 million; it may drop to $90 million or $85 million.

When that happens, borrowers must sell off even more of their a.s.sets to meet new margin calls. This creates a cascade of fire sales, as too many sellers chase too few buyers. Even worse, lenders nervous about the solvency of borrowers may start requiring an even greater equity margin and lower leverage ratios as a condition of rolling over the debt. This adds fuel to the fire, and selling pressures intensify. Of course, borrowers may sell other a.s.sets to make margin calls: Treasury bonds or plain-vanilla equities. Unfortunately, if everyone pursues that strategy at the same time, the same dynamic that played out with CDOs will affect these other a.s.sets too: there will be too many sellers and too few buyers, and prices will fall in a range of a.s.set cla.s.ses.

In this way, what began as a problem in, say, housing, can suddenly spread to other markets. And what began as a subprime problem can suddenly become everyone's problem. Sound familiar?

Chapter 4.

Things Fall Apart.

Walter Bagehot was a giant in the nineteenth-century British financial world. Aside from editing The Economist for many years, he wrote extensively about financial crises, most famously in Lombard Street, published in 1873. Writing about the great banks of his day, Bagehot complained that they "are imprudent in so carefully concealing the details of their government, and in secluding those details from the risk of discussion." That veil of secrecy was all well and good in prosperous times, he observed, but in a downturn it could become a terrible liability. Suppose, he wrote, that one of the "greater London joint stock banks failed." The result "would be an instant suspicion of the whole system. One terra incognita being seen to be faulty, every other terra incognita would be suspected." In short, he concluded, "the ruin of one of these great banks would greatly impair the credit of all."

If Bagehot had been alive in 2007, he would have recognized a familiar but deeply unsettling scene: Citigroup, a financial inst.i.tution with impeccable credentials but an impenetrable balance sheet, was ailing because of mysterious dealings with shadowy SIVs and conduits and a baffling a.s.sortment of structured financial products. A big bank was in trouble, and the extent of its problem was not apparent. Other financial inst.i.tutions came under suspicion; uncertainty and unease roiled the markets.

What happened next was precisely as Bagehot antic.i.p.ated. The first failures of 2007 set the stage for a collapse of confidence and an evaporation of trust, not merely in the shadow banks but in conventional banks as well. In no time at all, the ordinary bank-to-bank lending that supports global finance collapsed. The reason was simple: the entire financial system was one great terra incognita. As one market economist at the doomed firm Lehman Brothers observed late in the summer of 2007, "We are in a minefield. No one knows where the mines are planted." The result was the paralysis of the entire financial system.

That paralysis was a function of not knowing which banks were merely illiquid and which banks were truly insolvent. To have trouble rolling over some debt because of a seizure in the markets was one thing; to be bankrupt was altogether another. In a panic, it's difficult to tell which is which, and absent any clarification, panic can only grow. When that happens, inst.i.tutions can swiftly slide from illiquid to insolvent, as a.s.set values drop amid countless fire sales.

The only thing that can reliably arrest the descent into fear and terror is a lender of last resort. Bagehot is generally credited with coming up with the idea. He believed that a bank of banks-something like the Bank of England or the Federal Reserve-must step up to the plate and lend to those caught in the crunch. The holders of what he called the "cash reserve" must "advance it most freely for the liabilities of others. They must lend to merchants, to minor bankers, to 'this man and that man,' whenever the security is good." After all, he observed, "in wild periods of alarm, one failure makes many, and the best way to prevent the derivative failures is to arrest the primary failure which causes them." Yet Bagehot was against indiscriminate bailouts: only solvent inst.i.tutions should be able to gain access to loans, which would be made at penalty rates so as to discourage all but the most desperate. His philosophy has often been distilled to its essence: "Lend freely at a high rate, on good collateral."

Over the course of 2007-8, Bagehot's perceptive diagnosis, along with a deeply flawed version of his prescription, played out dramatically. Panic struck the markets, uncertainty spread, liquidity evaporated, and central banks around the world threw lifelines to banks large and small and to financial inst.i.tutions of every stripe. It was a rescue effort on a scale that Bagehot never foresaw. For this crisis, although a textbook case, was bigger, swifter, and more brutal than anything seen before. It was a nineteenth-century panic moving at twenty-first-century speed.

The Minsky Moment.

By the spring of 2006, the financial system, with its extraordinary reliance on leverage-and its blind faith that a.s.set prices would only continue to rise-was primed for a breakdown of monumental proportions. Financing increasingly depended on the sort of speculative and Ponzi borrowing that Minsky predicted. Euphoria that began in the housing sector and percolated upward throughout the entire financial system only encouraged further risk taking, and the few skeptics who raised the alarm were not heard. As Minsky himself said of these euphoric moments, "Ca.s.sandra-like warnings that nothing basic has changed, that there is a financial breaking point that will lead to a deep depression, are naturally ignored in these circ.u.mstances."

And so it was with this boom. Throughout 2006 and into 2007, one of the authors-Nouriel Roubini-warned of the coming collapse, as did a handful of other prescient commentators. In general, their warnings fell on deaf ears, much as Minsky antic.i.p.ated. Naysayers at the height of a bubble, Minsky observed, "do not have fashionable printouts to prove the validity of their views," and those in the establishment inevitably "ignore arguments drawn from unconventional theory, history, and inst.i.tutional a.n.a.lysis."

Indeed, by the time a bubble peaks, its partic.i.p.ants do more than scorn the skeptics; they proclaim that a new age of prosperity has arrived. The particulars vary from era to era, but the language is the same. On October 15, 1929, the otherwise accomplished economist Irving Fisher announced that having dropped downward from their remarkable highs, "stock prices have reached what looks like a permanently high plateau." Likewise, in December 2005 the somewhat less accomplished (and more subjective) spokesman for the National a.s.sociation of Realtors, David Lereah, looked at a similar precrash drop and uttered this sage p.r.o.nouncement: "Home sales are coming down from the mountain peak, but they will level out at a high plateau, a plateau that is higher than previous peaks in the housing cycle."

It seems quaint in retrospect, but what inaugurates a financial crisis is rarely something dramatic or out of the ordinary, merely a leveling off, a movement sideways, and a few unsettling signs. Those arrived in the spring of 2006, as housing starts flattened out, and home prices-which had doubled in real terms over the previous decade-stopped rising. The reason was simple enough: the supply of new homes began to outstrip the demand, and a rise in interest rates made variable-rate mortgages more expensive. Prices leveled off.

At the same time, as in every financial crisis, a "canary in the coal mine" signaled that all was not well: subprime mortgages issued in 2005 and 2006 began to exhibit unusually high rates of delinquent payments. These same mortgages came with features-superlow teaser rates, option ARMs, negative amortization-that depended on refinancing at low rates. But the option of refinancing-particularly for those mortgages that had no down payment and no equity-was available only if home prices kept rising. As a consequence, delinquencies and defaults started to crop up; cracks appeared in the facade.

Still, there was little indication that this was the beginning of a colossal banking crisis. But beginning in late 2006, the shadow banking system became the focus of a slow-motion run that George Bailey himself would have recognized. The hundreds of unregulated nonbank mortgage lenders who had been at the forefront of originating subprime mortgages relied heavily on short-term financing from larger banks. Once subprime mortgages were going into default at accelerating rates, the larger banks refused to renew these lenders' lines of credit. Unable to tap a lender of last resort, the nonbank lenders began to fail, victims of a twenty-first-century bank run.

The first lender to go under was the hilariously misnamed Merit Financial, which had allegedly spent all of fifteen minutes training its loan officers before setting them loose to originate loans with little doc.u.mention, liar loans, and no-income, no-job NINJA loans. But Merit Financial was not alone. Other nonbank lenders may have kept up professional appearances, but their lending practices were no less suspect. By the end of 2006, ten inst.i.tutions had gone bust, and the flow of mortgages through the securitization pipeline began to slow. By end of March 2007, the number of nonbank lenders that had collapsed soared to fifty or more. On April 2 the nation's second-largest subprime lender-New Century Financial-went bankrupt after its funding dried up. At the same time, others who had battened on the business of originating mortgages-thousands of small-time mortgage brokers-went out of business.

Most market commentators claimed that the problem was restricted to one small sector of the financial system. This too often happens as financial crises gather steam: the problem is widely seen as "contained"-in this case, to a handful of reckless mortgage lenders and the loans they made. Federal Reserve chairman Ben Bernanke fell into this trap when he appeared before Congress in May 2007. While he conceded that the subprime market had plenty of problems, he portrayed these troubles as an isolated disease outbreak rather than the beginnings of a pandemic.

Then a London-based company called Markit Group introduced something called the ABX Index, which measured stress in the market for subprime securities. It did so by measuring the prices of a basket of credit default swaps, used to transfer the risk of default on securities derived from subprime home loans. The goal, a company spokesman said, was "visibility and transparency." Using the ABX, one could measure the cost of buying insurance-in the form of credit default swaps-against defaults of tranches of mortgage-backed securities and CDOs rated from an abysmal BBB to a supposedly high-grade AAA. Over the course of 2007, the ABX Index went into a free fall, as bottom-of-the-barrel tranches lost upwards of 80 percent of their value. Even the safest AAA tranches lost 10 percent by July 2007.

The fall in the ABX Index revealed that something was going horribly awry. Worse, the ABX figures made all the shadow banks look at their a.s.sets and recalculate the value of the securities they held. Collateralized debt obligations that had been worth one hundred cents on the dollar sustained enormous losses, leaving financial inst.i.tutions with fewer a.s.sets relative to their outstanding liabilities. Faced with dwindling reserves, both the traditional and the shadow banks began to h.o.a.rd cash, refusing to lend on the basis of collateral that looked more dubious by the day.

A sudden aversion to risk, a sudden desire to dismantle the pyramids of leverage on which profits have until so recently depended, is the key turning point in a financial crisis. In earlier times, it was called "discredit" or "revulsion"; more recently it has been called a "Minsky moment." By late spring of 2007, that moment had definitely arrived.

The Unraveling.

Hedge funds may not look like banks, but they operate much as banks do, getting short-term investments from individual and inst.i.tutional investors as well as short-term repurchase agreements, or repos, from investment banks. Like conventional banks, hedge funds invest their short-term borrowings for the long term. For example, two hedge funds run by Bear Stearns sank billions of short-term loans into highly illiquid subprime CDO tranches.

The collapse of those two funds in the summer of 2007 portended the fate not only of hundreds of other hedge funds but of the shadow banking system as a whole. Like many players in this system, these two funds were virtually unregulated but highly leveraged; the riskiest had a debt-to-equity ratio of twenty to one. As the ABX Index revealed the market's growing belief that subprime CDOs might lose much if not most of their value, these two hedge funds started to suffer major losses.

At that point the banks that had lent billions to the two funds made margin calls and threatened to sell the collateral-some AAA CDO tranches-that the two funds had pledged to secure financing. This step was fateful: up until now CDOs and other forms of structured finance had rarely been traded. The ABX Index was merely a proxy for prices, not an actual reflection of the going market price. The hedge fund managers knew that these securities would never fetch their original price; trying to sell them into a panicked market would have revealed that the entire CDO enterprise was, like the fabled emperor, without clothes. Instead, Bear Stearns injected money into the funds. But to no avail: by the summer of 2007, one of the funds had seen 90 percent of the capital put up by the investors of wiped out, while the equity of the more leveraged fund disappeared altogether. Both funds filed for bankruptcy at the end of July. They were not alone: another hedge fund created by UBS perished under similar circ.u.mstances.

These early failures showed how hedge funds could fall victim to the equivalent of a bank run on their a.s.sets. Inst.i.tutional creditors could suddenly refuse to roll over the repo loans, leaving them high and dry. Alternatively, those who'd invested equity-wealthy individuals and the like-could demand their money back, just as depositors used to demand money back from old-fashioned banks like Bailey Bros. Building & Loan. Either way, the result was the same: the short-term financing of the hedge funds could readily disappear, forcing them to shut their doors.

The failure of the first three hedge funds conformed to the cla.s.sic narrative of a financial crisis. Most crises see a few initial high-profile failures, then a period of unsettling uncertainty, as people try to determine whether the troubles that have befallen once-healthy inst.i.tutions are part of a larger problem. More often than not a larger problem is emerging, and this crisis was no different: in the two years following the failure of the Bear Stearns and UBS funds, some five hundred hedge funds perished, the victims of a slow-motion bank run. The reason was simple: the creditors of the hedge funds couldn't and didn't know how much exposure individual hedge funds had to the toxic a.s.sets. Faced with so much uncertainty, they curtailed credit to all of them.