A solution would be to mandate that all inst.i.tutional investors pay into a common pool that would be administered by regulators. For every new issuance of debt, this pool would be used to purchase ratings from a group of sanctioned agencies. This solution would require that all players in the financial system-even less-regulated ent.i.ties like hedge funds-contribute to the pool.

This requirement would upend the economy of the ratings agencies, but that's precisely the point: the idea that the issuers of debt pay for their ratings is bizarre. To return to our earlier a.n.a.logy, it's comparable to having students pay professors for their grade. Imagine that students have a choice of professors (just as debt issuers have some choice of ratings agencies). Those professors who hand out lots of Fs will soon find that their more easygoing colleagues-who hand out nothing but As-are attracting more students and more revenue. But all those As will be just as spurious as the AAA ratings handed out at the height of the housing bubble.

Make no mistake: reforming the ratings agencies will be no easy task; they occupy a unique place within the financial firmament. But unless some of the foregoing reforms are put in place, the conflicts of interest will almost certainly continue.

But let's a.s.sume for a moment that those conflicts can be made to disappear, and that ratings agencies henceforth bestow accurate ratings on things like mortgage-backed securities. Unfortunately, not even the ratings agencies will touch the opaque, mysterious, and often baffling instruments known as derivatives.

Dealing with Derivatives.

In 2002 Warren Buffett penned a now-legendary annual report to investors in Berkshire Hathaway. He decried the growing use of derivatives, which he ominously described as "time bombs, both for the parties that deal in them and for the economic system." Buffett didn't pull any punches, characterizing derivatives as "financial weapons of ma.s.s destruction." They carry dangers, he warned, "that, while now latent, are potentially lethal." Most presciently of all, he warned that "the derivatives genie is now well out of the bottle, and these instruments will almost certainly multiply in variety and number until some event makes their toxicity clear."

Buffett was right, but the story is a bit more complex. Derivatives have been around for several centuries; not until recently did they a.s.sume forms that posed a significant danger to the global financial system. After all, a derivative is simply a bet on the outcome of some future event: a movement in interest rates, oil prices, corn prices, currency values, or any number of other variables. They go by various names-swaps, options, futures-and they've worked just fine for decades, enabling people to "hedge" against risk. In their original iteration, farmers could hedge against fluctuations in the prices of their crops in advance of a harvest, giving them peace of mind they would otherwise lack.

But in recent years derivatives have grown into something altogether different, thanks to the rise of several new varieties, such as the credit default swap (CDS). This instrument has been compared to an insurance contract but was in fact very different. It superficially did resemble insurance in that it allowed a buyer to purchase protection in the event that a debtor defaulted on his obligations. If that happened, the seller of the "insurance" would be on the hook to help the buyer recoup his losses. However, unlike the purchaser of an insurance contract, the purchaser of a CDS didn't have to actually own a chunk of the a.s.set that was the subject of the bet. Worse, anyone who had placed a bet that someone would default had every incentive to make this happen. In these cases, purchasing a CDS was akin to buying homeowners' insurance on a house that you didn't actually own-and then trying to set fire to it.

The CDS market grew from next to nothing to astonishingly large. By the time the crisis broke in 2008, the notional value of CDSs (the amounts of money being insured) topped out at $60 trillion. Much of the credit-or blame-for their meteoric rise lies with free-market zealot Senator Phil Gramm, who from 1995 to 2000 presided over the Senate Banking Committee. In his final year he managed to insert a provision in the Commodity Futures Modernization Act that exempted CDSs and other "over-the-counter" derivatives from regulation by the Commodity Futures Trading Commission (CFTC).

The key phrase here is "over the counter" (OTC). It might seem to be the opposite of "under the table," but in fact "under the table" is as good a definition as any of an OTC transaction. OTC transactions are ones in which the derivative contract is signed by two private parties-typically a "bilateral contract," to which no one else is privy. The lack of transparency is complete: no one knows the extent of anyone else's exposure, much less where it's concentrated. The financial firms that created many of these instruments were all too happy to keep the details secret; after all, their trading strategies were proprietary information and their trading fees very high. But during the crisis this secrecy proved corrosive to investor confidence.

No less troubling was "counterparty risk": the chance that the inst.i.tutions that had sold this "insurance" would be unable to make good on their promises, particularly during a systemic financial crisis. That's precisely what happened as the recent crisis gathered steam: major financial inst.i.tutions, confident that they would never have to pay, didn't set aside the necessary reserves. This posed a risk to the entire financial system, particularly in the case of AIG, which had insured-via CDS transactions-over a half trillion dollars' worth of toxic CDO tranches. AIG was in no position to cover the losses on these tranches, and because its failure would have bankrupted the firms whose a.s.sets it insured, the U.S. government stepped in and bailed it out. In effect, counterparty risk created a financial system that was not only too big to fail, but too interconnected to fail.

Derivatives have been a.s.sociated with a number of other infamous episodes in financial stress and crisis. For example, a kind of derivative known as portfolio insurance was implicated in the 1987 stock market crash, and in 1994 losses on derivative positions brought Orange County, California, to the brink of bankruptcy. Derivatives played a major role in the LTCM fiasco in 1998, and they aggravated the boom and bust in oil prices during 2008 and 2009. Derivatives can wreak havoc in other ways as well, hiding liabilities, avoiding taxes, frustrating attempts to restructure debt, and even serving as a means of purposely triggering defaults of banks, firms, and nations.

Given this rap sheet, banning derivatives may seem like a good idea. But it's not: most derivatives operate without ill effect. What we need to do is control the excesses that certain derivatives can cause. As with everything else in this mess, that's easier said than done; there is no panacea. But some sensible steps should be taken immediately.

First we must correct the problem of transparency. True, some derivatives have long been traded over the counter without problems-like plain-vanilla interest-rate swaps and currency swaps-and could reasonably remain that way. But CDSs are another story altogether. These must be brought into the light of day and subjected to rigorous regulation by the SEC and the CFTC. The Obama administration has already taken steps along these lines, and several proposals to make this regulation a reality are already on the table.

One school of thought advocates forcing credit derivatives onto the kind of central exchange similar to those where simpler derivatives are bought and sold. This idea makes perfect sense, as exchanges guarantee that derivatives will be cleared and settled in a straightforward, transparent fashion. Such a new inst.i.tution could also ensure that the parties to derivatives have the necessary collateral to make good on their promises.

Unfortunately, while some of the existing credit derivatives could be standardized and traded on such a central exchange, not all of them can be: many of the OTC derivatives are next to impossible to standardize and are not traded in sufficient volumes; their price can't be consistently quantified like a stock or bond (or a common derivative).

These more esoteric credit derivatives should be registered in a central clearinghouse. Such inst.i.tutions already exist for other, simpler kinds of derivatives: the Options Clearing Corporation, for example, handles a host of derivatives related to equities and commodities. Though a private organization, it has the imprimatur of both the SEC and the CFTC. Part of its job is to make certain that the parties to a derivatives contract have sufficient collateral to make good on their promises. (In other words, no financial firm could offer to "insure" against defaults without posting sufficient collateral.) In return, the clearinghouse would a.s.sume the burden of the contract in the event the counterparty failed. All this helps reduce the problem of counterparty risk.

While a clearinghouse is a good idea, a few caveats are in order. First, if the markets undergo a systemic meltdown, the clearinghouse may be unable to cover all the contracts, forcing it to default too. To some extent, that risk can be minimized by raising margin requirements. Nonetheless, as recent events have shown, it's easy to underestimate systemic risk, so such a clearinghouse would necessarily operate under the scrutiny of regulators, who would be charged with making sure that it had the necessary reserves to ride out a storm.

More problematic is the risk of regulatory arbitrage: if the clearinghouse handles only straightforward, standardized credit derivatives, financial engineers are likely to deliberately create exotic derivatives that the clearinghouse cannot accommodate, simply to evade regulation. Far better to have the clearinghouse handle all such derivatives.

This solution could go hand in hand with other reforms to increase transparency. For example, CDS transactions could be registered in a central database that would be accessible to the public. As our colleagues at New York University have suggested, that sort of data collection could be modeled on the Trade Reporting and Compliance Engine (TRACE), the database administered by the Financial Industry Regulatory Authority. Making these markets less opaque will have the added benefit of making pricing more compet.i.tive and cutting into the ability of firms to game the system, quoting higher prices than market conditions warrant.

In some cases, it's worth banning certain derivatives altogether, or severely restricting their use. For example, regulators should consider forbidding CDS contracts entirely. One of the cardinal rules of insurance holds that the party purchasing the policy must have an "insurable interest": a direct stake in the outcome. Most CDS contracts dispense with this custom. To return to an earlier a.n.a.logy: CDS contracts effectively gave Wall Street "huge incentives to burn down your house," as one reporter for the Financial Times pithily put it.

If they cannot be banned outright, insurance companies should be barred from selling these guarantees. The only ones trafficking in these instruments should be hedge funds and other high-risk players in the financial markets. Moreover, they should be subject to rigorous margin and collateral requirements via a clearinghouse. If a hedge fund is going to get into the insurance business, it should be required to demonstrate beyond a shadow of a doubt that it will be able to make good on its obligations.

One final measure that would go a long way toward streamlining the regulation of derivatives would entail changing the relative responsibilities of the SEC and CFTC. These agencies regulate different slices of the derivatives market, effectively dividing regulatory authority. Consolidating the responsibility for overseeing all derivatives within a single agency would permit a more systematic approach to regulating and supervising derivatives and, more important, reducing the potential threat they pose to the stability of the international financial system.

These sorts of reforms would address several problems: counterparty risk, a lack of price transparency-even the outsize fees that are a feature of the mysterious market in over-the-counter derivatives and that enable insiders to reap enormous fees and fleece investors.

Nonetheless, these suggestions are not a cure-all. Derivatives are among the trickiest things to regulate and monitor, and their explosive growth over the last decade has only made that job harder. They have gone from being a means of hedging risk to a purely speculative instrument that permits often naive investors-pension fund managers, for example-to a.s.sume ma.s.sive amounts of leverage and risk. Increasingly exotic, opaque, and impenetrable to nonspecialists, they pose a very serious danger to the financial system that the foregoing reforms alone will not cure.

For this reason, the new generation of derivatives should be the subject of far more systematic and ruthless scrutiny by regulators. Put differently, this is not a situation where regulators need to fear that cracking down on these instruments will somehow imperil economic growth. Far from it: their continued existence poses a far greater danger to global economic stability, and the sooner legislators, policy makers, and regulators understand that, the better.

Unfortunately, that stability has to be sh.o.r.ed up worldwide. That means reexamining some of the global guidelines that shape how banks do business too.

Basel and Beyond.

The quaint Swiss city of Basel has many claims to fame: the oldest university in Switzerland, the country's first zoo, and, more recently, its tallest building. Basel's schools have been home to intellectual giants like Friedrich Nietzsche, and its storied chemical and pharmaceutical companies have given the world everything from Valium to LSD. And its banking community has given the world something a little less exciting but no less important: the Basel Committee on Banking Supervision.

Born in 1974, this little-understood inst.i.tution draws its members from central banks of the advanced economies known as the G-10. Its mandate is to come up with better ways to regulate and supervise banks and other financial inst.i.tutions. While its recommendations are nonbinding, it nonetheless carries a great deal of weight. Much of the financial system as it existed on the eve of the crisis was a creature of the Basel Committee's guidelines.

Those guidelines, or accords, have evolved over the years. The first accord, the Basel Capital Accord, known as Basel I, asked banks to differentiate between the various cla.s.ses of a.s.sets they held in order to better a.s.sess the relative risk posed by holding them. This risk a.s.sessment would affect how much capital a bank had to hold.

Consider two hypothetical banks, each of which borrows a billion dollars from other sources and invests it. One invests in low-risk, supersafe U.S. Treasuries, the other in high-risk corporate junk bonds. Under the Basel I guidelines, the two banks would a.s.sign a different risk factor (a percentage) to these different a.s.sets. This risk factor would guide how much capital they had to hold relative to these risks. In practice the bank with the supersafe government debt didn't need to hold as much capital as the bank with its money in junk bonds.

Basel I had a few other stipulations. Banks that operated in multiple countries had to hold capital equivalent to 8 percent of their risk-weighted a.s.sets. In an additional wrinkle, the guidelines spelled out the form that this capital or equity could take: common shares, preferred shares, and other high-quality capital (called Tier 1 capital), and everything else (Tier 2 capital).

The first Basel accord went into effect in the 1980s, and by 1992 most of the G-10 had adopted its recommendations. Many emerging-market economies voluntarily adopted these guidelines too, as a demonstration of financial stability and prudence. Unfortunately, standards that made sense for advanced industrial economies proved more difficult for emerging economies to maintain, particularly in times of crisis, and proved their undoing.

No less troubling, bankers had found ways to hide risk that Basel I did not antic.i.p.ate-for example, by securitizing a.s.sets. These sleights of hand gave bank balance sheets the appearance but not the reality of stability. Bankers had obeyed the letter but not the spirit of the Basel I guidelines.

These failures led to Basel II. While its predecessor had filled a mere thirty-seven pages, the new accord was almost ten times as long. It gave much more precise technical guidelines on weighing the relative risk of various a.s.sets; suggested methods for making these calculations; expanded the definition of risk to encompa.s.s other perils, such as the likelihood that a.s.sets might fall in value on the open market; sought to close various loopholes by which banks had hidden risk; urged regulators to move more aggressively to monitor compliance with capital reserve requirements; and spelled out terms by which banks would make their financial condition public. Though many European nations wanted Basel II to apply to all banks, the United States, Canada, and the United Kingdom successfully argued that it should apply only to large international banks.

The members of the G-10 hammered out a final version of Basel II in 2006. It then went to the individual nations for implementation, a process that was under way when the crisis. .h.i.t. It immediately became apparent that for all its specificity, Basel II had serious flaws. Although many of the revisions were a response to the crises of the 1990s, Basel II did not protect large banks from the kind of disruptions that attend a major financial crisis. Simply stated, Basel II a.s.sumed that the world's financial system was more stable than it actually was. This was a serious mistake.

The crisis underscored several realities. One, banks needed higher-quality capital and more of it. Two, the "capital buffer" that many banks had established was nowhere near large enough to shelter them from the kind of shock delivered by the housing bust and the credit crisis. Three, the quality of the capital as defined by Tier 1 and Tier 2 could deteriorate significantly in a time of crisis.

Reforming Basel II will take years, but a few things stand out. For starters, the way Basel II defines and ranks capital should be changed. Rather than relying on the Tier I definition to calculate bank capital, it might use a narrower measure known as Tangible Common Equity (TCE). TCE counts only common shares in its calculation of capital; by contrast, Basel's Tier I capital definition includes both common and preferred shares. TCE is thus a more conservative estimate of the capital a bank has on hand. As such, it may be a more realistic way to a.s.sess a bank's health in the face of a crisis.

There's a deeper problem with the way that Basel II is structured. The methods it used to calculate capital had the perverse effect of exaggerating the amount of capital that banks had on hand during boom times while doing the opposite when the crisis. .h.i.t, causing them to curtail their exposure to risk in excessive and disruptive ways. This happened because during boom times the price of a.s.sets held by the banks rose, simultaneously reducing their need for capital and encouraging them to take on more risk. In the crisis, the process went into reverse: a.s.set prices fell, and suddenly the banks needed more capital just when it was most difficult to obtain.

In economics, this phenomenon is known as "procyclicality." The meaning of this unwieldy term is simple: something that's procyclical amplifies fluctuations in the economy at large-for example, a boom-and-bust cycle. When it comes to capital, that's obviously a problem; if anything, you want the guidelines not to amplify economic fluctuations but to shield banks from them. To avoid procyclicality, one could adopt a different way of calculating capital called "dynamic provisioning." Instead of forcing banks to hold a static amount of capital at all times-like the 8 percent imposed by the Basel accords-a dynamic system would allow it to vary over time. In boom years, capital requirements would go up. When things go sour, the capital requirements would decrease. A version of dynamic provisioning has already been used by Spanish banks, and while not a cure-all, its widespread adoption under a new Basel regime may be worth trying.

Another possible solution to the problem of procyclicality is the use of "contingent capital." This idea, which is gaining adherents, deserves a close look, though it's not without problems. Here's how it works: in good times, banks issue a special kind of debt known as "contingent convertible bonds." It differs from ordinary debt in that if a bank's balance sheet declines past a particular "trigger point," the debt will "convert" into shares or equity in the bank.

Contingent capital potentially provides several benefits. First, the bank gets more capital when it really needs it, enabling it to stay afloat. The ailing bank survives, but at a price: its former bondholders lose their money but end up with a significant ownership stake in it. This sudden influx of outside shareholders diminishes the power of the existing shareholders.

That's not in the interest of the original bondholders or shareholders. So both groups have an incentive to keep a closer eye on what the bank is doing so that it never finds itself at this unfortunate juncture. In theory, if the bondholders think the bank is headed in the wrong direction, they will impose market discipline in the form of higher borrowing costs. Likewise, the shareholders will rein in risky activity that might land the bank in trouble.

This idea may work better on paper than in practice. For it to work, after all, banks would have to issue enough contingent convertible bonds. Otherwise, the sort of self-discipline it's meant to encourage-never mind the capital it's supposed to provide-will be insufficient. Still, it's one of the better proposals on the table at the moment, and it may address one of the major failures of Basel II. A central conceit of that agreement was that if a bank's regulators, shareholders, and managers couldn't restrain risky behavior, its creditors might be encouraged to do so in their stead; purchasers of the bank's debt would impose "market discipline," jacking up borrowing costs on banks that took too many risks.

Unfortunately, when the recent crisis. .h.i.t, there wasn't enough bank debt out there to impose meaningful market discipline. But even if there had been, governments around the world threw market discipline out the window when they decided to guarantee all that debt and make it easy for banks to borrow more. Contingent capital may be one way around that problem too, ensuring that debt gets converted into equity rather than guaranteed, backstopped, and bailed out. It might-perhaps-hold the line on moral hazard.

Another problem for the sages of Basel to consider is the peril a.s.sociated with a lack of liquidity. As the recent crisis amply demonstrated, financial inst.i.tutions-both traditional banks and members of the shadow banking system-are highly susceptible to a liquidity crunch. In the past, the Basel Committee paid little attention to this problem. That needs to change. As we've noted in previous chapters, the recent disaster was a banking crisis writ large, with financial firms of all stripes borrowing money on a short-term, liquid basis, then investing it in long-term, illiquid a.s.sets. When panic struck, banks and shadow banks couldn't renew their short-term loans. They became illiquid.

Therefore an essential element of future regulation of all financial inst.i.tutions-banks and nonbanks-should be a greater emphasis on the management of liquidity risk. One way to manage it would be to require firms to avoid short-term borrowing-in other words, to significantly lengthen the maturity and duration of their liabilities. A bank that has to renew its borrowing once a year is less likely to suffer a liquidity crunch than one that renews loans every day. Again, this technique for managing liquidity risk won't solve everything: as the recent crisis demonstrated, even inst.i.tutions that traffic almost exclusively in liquid investments-money market funds, for instance-can still get themselves in trouble by putting just a tiny bit of their money into less liquid a.s.sets.

Another flaw with the existing Basel regime is that it gives too much discretion to financial inst.i.tutions to use their own internal models for a.s.sessing risk. The Value at Risk model (VaR) is one, a mathematical formula that purports to calculate the likelihood that a firm will suffer a loss on its a.s.sets; another is the Gaussian copula, a statistical tool that was used to price esoteric a.s.sets like CDOs. Both models disregarded the likelihood of crises and other disruptive events, leaving banks vulnerable when the crisis. .h.i.t. Sadly, the solution is not simply to devise a better set of equations. Models are only as good as the people using them, and if individuals wish to jack up short-term profits by ignoring risk, they will be able to fiddle with any formula in order to get the answer they want.

So we must think about other ways to minimize risk. For starters, it's worth rethinking how financial inst.i.tutions themselves manage risk. At present, the "silo" approach is dominant: each division or business line considers risk in isolation, never contemplating how its ventures might interact with those of other divisions. The cla.s.sic example is AIG: a small branch of the company based in London with only 375 employees managed to insure enough toxic CDOs to bring down countless other divisions that employed more than a hundred thousand people.

No less troubling, many risk managers are marginalized; after all, they stand in the way of bigger profits. That makes it all the more difficult for them to be effective in controlling the traders with whom they work. For this reason, risk management in the future should dispense with the silo model and instead strive for a bird's-eye view of risk throughout the organization. That means appointing a more powerful chief risk officer who reports directly to the CEO and the board of directors, along with a staff capable of monitoring what's going on in multiple divisions of the bank and, more important, cracking down on risky behavior.

The Coming Crisis.

Financial crises have a funny way of making radical reforms seem reasonable. Much of what we've described in this chapter would have seemed extreme and unnecessary in advance of the crisis that hit in 2008. That's no longer the case.

The reforms we've outlined would go a long way toward increasing accountability and transparency in the financial system by reforming compensation, regulating securitization, bringing derivatives under public scrutiny, and putting the putative guardians of the system-the ratings agencies-on a very short leash. These reforms would also ensure that banks and other financial firms have enough capital and, no less important, enough liquidity to weather a major financial crisis.

These reforms would also put an end to the potential abuses a.s.sociated with having a small oligopoly of banks and investment banks controlling so much of the financial system. If derivatives and other financial instruments were more transparent, it would be much harder for these firms to charge their clients outsized fees and bid-ask spreads. These firms are understandably reluctant to see these instruments' trading moved from dealers' markets onto exchanges: doing so would deprive them of the ability to extract the kind of profits that come with having access to information their clients lack. Under the current system, firms can very easily charge high bid-ask spreads because there is no price transparency; they can effectively "front run" their clients, meaning that they can use information about their clients' investments to make money in other departments, such as in proprietary trading or even in the actual market-making and market-dealing activities. Greater transparency will frustrate those sorts of behaviors.

Skeptics might reasonably point out that if investors want to pay through the nose for the privilege of alpha-or schmalpha-returns, that's their business. But the rise of a small coterie of incredibly powerful, opaque financial firms has generated a far more unsettling problem. These firms now house a staggering a.s.sortment of financial functions: everything from securities origination and underwriting, market making and dealing, prop trading, private equity, hedge funds, and a.s.set management to bread-and-b.u.t.ter banking. The interconnections among those various functions, never mind the interconnections among the handful of surviving firms, have created a system that is extraordinarily vulnerable to systemic risk.

Moving most trading from OTC markets that rely on market makers/ dealers to exchanges would reduce these rent-extracting distortions but, more important, would radically reduce the counterparty risk that makes the financial inst.i.tutions too interconnected to fail. Indeed, the more that transactions occur on exchanges, the less the system becomes interconnected as counterparty risk is significantly reduced. Not only do we need to reduce the TBTF problems by making each inst.i.tution smaller, we also need to unbundle financial services within financial inst.i.tutions to reduce the too-interconnected-to-fail problem: with exchanges, broker dealers would be involved only in the efficient execution of trades for clients, not in market making/dealing, which is rife with conflicts of interest, lack of price transparency, and large and systemic counterparty risk. So we need to go back to Gla.s.s-Steagall, and even beyond it, to a financial system in which both inst.i.tutions and their activities are unbundled to make them less too big to fail and less too interconnected to fail.

In short, the concentration of financial power has created a system that is too interconnected to fail. The proposals outlined in this chapter represent the first steps toward curing these problems. But far more radical reforms must be implemented if the financial system is to achieve any semblance of stability in the coming years.

Chapter 9.

Radical Remedies.

In early January 2010, Ben Bernanke defended the Fed's handling of the recent financial crisis. The lesson he drew was straightforward and simple: better regulation could have prevented it. As he put it, "The lesson I take from this experience is not that financial regulation and supervision are ineffective for controlling emerging risks, but that their execution must be better and smarter."

This is correct. But how, precisely, might regulators do a "better and smarter" job of executing their responsibilities? In this chapter we consider some of the significant obstacles that stand in the way of this ambition. These obstacles include the intentional evasion of regulatory oversight, or what's known as regulatory arbitrage; the too-many-cooks-in-the-kitchen problem, in which an overabundance of regulators and a lack of coordination frustrate effective supervision of the financial system; and last but not least, the problem that regulations are only as good as the regulators who enforce them. The recent history of regulation shows that these problems have a way of derailing even the most rigorous and well-conceived rules and regulations. Any successful reform of the regulation and supervision of the financial system will have to confront them.

That's a start. So too are the commonsense, middle-of-the-road recommendations laid out in the previous chapter. But sometimes it's not enough to impose new regulations on the status quo; sometimes a bit of regulatory "creative destruction" is in order. In this chapter, we take up some of the more radical changes that can and should be imposed in the coming years, including breaking up big banks and imposing a number of new firewalls in the financial system.

Finally, we weigh an entirely different but equally radical idea: the use of monetary policy to prevent speculative bubbles. Though it is relatively simple and straightforward, most economists and policy makers consider this idea heretical and dangerous. In fact, Bernanke explicitly ruled it out when he delivered his postmortem, arguing that better regulation and supervision offered a more "surgical" approach to addressing the problem.

We beg to differ. When used appropriately, monetary policy is one of the most effective and powerful ways to deal with a.s.set bubbles and the crises they cause. Its effects are not surgical, but that's precisely the point: monetary policy can have a broader, systemic influence on the speculative climate that creates a bubble in the first place. It's therefore worth adding to the toolkit available to policy makers.

What follows is a glimpse of the possible future of finance-if and only if policy makers and politicians recognize that confronting crises requires some big-picture thinking and some radical reforms. Anything less would be tantamount to rearranging the deck chairs on the t.i.tanic.

Avoiding Arbitrage.

When people think of new financial regulation, they generally think of actual rules, guidelines, and laws that promise to end bad behavior, encourage stability, and otherwise prevent crises. That's all well and good, but bankers and traders have a funny way of dodging even the most carefully constructed regulations. Such regulatory arbitrage is one of the issues that policy makers must confront if reforms are going to have any effect.

In the years before the crisis, ordinary banks were reasonably well regulated, and in exchange, they had access to a government safety net: deposit insurance and explicit lender-of-last-resort support. This bargain was stifling for bankers who wanted more freedom to take risks. So they increasingly shifted banking activities to the shadow banks: inst.i.tutions that looked and acted like banks but weren't regulated like them. This was regulatory arbitrage: the purposeful movement of financial activity from more regulated to less regulated venues.

In the wake of the crisis, the consensus holds that these nonbank financial inst.i.tutions-the shadow banks-have to be regulated like ordinary banks. Most of them have received unprecedented levels of government support; now it's time for them to return the favor and submit to greater regulation. In addition, the recent crisis demonstrated that many of these firms are systemically important: their failure can send shock waves through the entire financial system-all the more reason that they should be regulated.

This all sounds good, right? Unfortunately, many policy makers are drawing the wrong lessons from history. They look at the monster shadow banks-inst.i.tutions like AIG-and conclude that only the big fish deserve to be regulated. That's the philosophy embodied in some of the proposals now on the table. For example, the Obama administration's proposals for regulatory reform specifically target "systemically significant financial firms."

A selective application of regulations would be a profound mistake: it would simply open the door to more regulatory arbitrage. Next time around, financial intermediation would move from the bigger, newly regulated inst.i.tutions to their less significant brethren. However small, these less regulated inst.i.tutions would become increasingly important to the larger system in the aggregate, and their collective failure could be equally problematic. For example, during the savings and loan crisis about fourteen hundred such thrifts went bust; no single one of them was systemically important, but their collective poor loan underwriting and losses had systemic effects.

For this reason, any new regulations must be applied across the board to all inst.i.tutions, not just to those deemed to pose a systemic risk. The same regulations governing everything from capital to liquidity ratios to compliance and disclosure standards should be imposed without exception, even if systemically important inst.i.tutions should be more tightly regulated than smaller ones given their potential systemic effects. Financial engineers should have no place to run and hide; regulation should not be applied only here and there. Otherwise there will be more arbitrage-and more crises.

In all fairness, regulatory arbitrage was too easy a game to play in recent years, thanks to what was euphemistically called "self-regulation." This was the idea that regulators could establish general "principles" for financial inst.i.tutions to follow, then request that the firms find some way to conform to them. Almost inevitably, this approach (also known as "soft-touch" regulation) bred more arbitrage. Charged with managing risk on their own terms, firms selected models that almost always underestimated the amount of capital needed to weather a crisis.

In dealing with the problem of self-regulation, regulators should be cautious about immediately issuing vast numbers of very specific rules governing each and every kind of structured financial product. This would be a fool's errand: such granular regulation will only generate another burst of regulatory arbitrage, as financial engineers figure out how to tweak products so as to evade the law.

Superspecific regulations would also be pointless on another level: the degree of financial innovation that has already taken place is staggering. One measure can be seen in the relentless expansion of a standard industry guide to derivatives. Originally a mere seven hundred pages long in 1989 (when it was called Swap Financing), the latest edition, published in 2006 (ret.i.tled The Das Swaps & Financial Derivatives Library), fills almost five thousand pages. It's astonishingly difficult to keep pace with financial innovation.

But this doesn't mean that regulators should throw up their hands, return to the old system of setting out "principles," and trust that the good traders and bankers at financial firms will abide by them. Rather, they should establish a robust set of simple rules governing key features of the financial system. For example, they should set clear caps on leverage-not on "risk-adjusted leverage" but on absolute leverage. The same specificity should apply to capital requirements and liquidity buffers. These can and should be absolute and should be applied across the board, to firms both large and small. Moreover, any discretion in interpreting them should lie not with the bankers, as has been the practice, but with the regulators.

That raises a somewhat vexing problem that transcends any particular set of reforms. How should regulations be administered? What official inst.i.tutions should take the lead in administering them? And how should those inst.i.tutions coordinate their efforts?

Enforcement and Coordination.

In the United States, a bewildering array of regulatory bodies, at both the state and the federal levels, share responsibility for the financial system. These bodies have evolved haphazardly over the course of more than a century.

Let's begin on the state level. Over the course of the nineteenth and twentieth centuries, each of the fifty states created its own separate banking and insurance commissions. These bodies operate in different ways and have varying levels of experience monitoring the firms in their jurisdictions, which include both state-chartered banks and national banks with federal charters. Some states also have securities regulators, akin to the SEC, as well as credit union regulators. All of these regulatory authorities vary greatly in their funding, their sophistication, and the rules they are responsible for enforcing.

Atop this decentralized structure sits the federal regulatory edifice, sharing enforcement in some cases, superseding it in others. The number of federal regulatory bodies is staggering, ranging from the Office of the Comptroller of the Currency to the Federal Reserve to the Commodity Futures Trading Commission, along with many, many others. Many of these agencies patrol the same territory, or overlap in substantial ways. Finally, as if that weren't confusing enough, there are several government-sanctioned, private supervisory bodies that advise the government, set rules for financial firms, and even police their members-inst.i.tutions like the Financial Accounting Standards Board (FASB).