Part 8 (2/2)

A variant of bonus pooling has been proposed by Raghuram Rajan. In his scheme, traders would be compensated for their high returns, but their bonuses would be held in escrow for several years. Should a trader incur a loss in subsequent years, it would be subtracted from the existing bonus account. In this ”bonus-malus” system, bonuses can be clawed back and nullified according to the ups and downs of a trader's long-term performance. The longer the bonuses are held in escrow, the more likely the traders will be to think more carefully about a.s.suming risk at the expense of long-term revenue.

The bonus-malus system works best if applied on an individual level. Unfortunately, bonuses are often calculated on an inst.i.tutional level, so that when bets pay off, everyone shares in the proceeds. Traders and bankers do not directly suffer the consequences of their bad decisions, which are borne by the pool at large. Still, collective clawbacks-the repossession of bonuses across the board-may nonetheless cast a pall of prudence over all those trading desks.

The problem of compensation has a more diabolical solution: to compensate traders and bankers not with money or with stock but with the very same esoteric securities that they're cooking up in their laboratories. Traders and bankers would get bonuses, but in a very specific form: a little slice of, say, that CDO that they had a hand in making. If traders cook up toxic securities, they get paid with the same. The thinking here is that if traders know that the proverbial chickens will come home to roost in their bonus package, they may be a little bit more careful about the eggs they lay.

A version of this plan is already operational. At the close of 2008, Credit Suisse announced that it was s.h.i.+fting some $5 billion worth of toxic a.s.sets off its own balance sheet and into a special fund. It then paid bonuses to employees out of this fund, replacing the usual form of compensation (shares of stock in the company) with shares of this fund. This raised some howls of protest; after all, many of those compensated had had nothing to do with the bad bets. Still, however imperfect, this is a good start.

Yet another kind of compensation scheme could draw from all of these proposals. For example, rather than retroactively saddling employees with the consequences of their bad bets (as the Credit Suisse scheme does), make it clear from the beginning that bonuses of bankers and traders will be paid in the securities they have a hand in creating. Better yet, put those securities-c.u.m-bonuses in escrow for several years, letting enough time pa.s.s to determine whether they are toxic or not. Finally, forbid employees to hedge against any potential losses on these future bonuses. (They are traders, after all, and if there's one thing they're good at, it's making money regardless of where the market is moving.) Whatever change in compensation is ultimately adopted should be implemented across the board. If one major financial firm adopts some version of the bonus-malus system but no one else does, employees from the more prudent firm will likely flock to high-rolling firms, where they'll be better compensated.

That means government must be involved. In the United States, only the federal government has the power to reform the compensation system in comprehensive fas.h.i.+on. It has plenty of justification for doing so: the government-or more to the point, the taxpaying public-has effectively bailed out and backstopped the entire financial system and has a cogent interest in making sure it doesn't have to do it again. Moreover, given the tangled web of princ.i.p.al-agent problems, shareholders cannot possibly be expected to reform compensation. But government could set across-the-board changes along the lines above.

Let's be clear: we're not suggesting that government cap compensation, though it would certainly be well within its rights to do so, particularly with banks still on government life support. What we're proposing is in a way more radical: that compensation be completely overhauled to reduce risky behavior, and by extension, the likelihood of another systemic collapse in the global financial system.

Those caveats aside, removing traders' incentives for taking on short-term risk (and creating disincentives, in the form of clawbacks) will probably cause compensation to decline. Is this a bad thing? No. In recent years, the financial services industry-and compensation within it-has undergone exorbitant and utterly unwarranted growth, driven by financial liberalization, financial innovation, elimination of capital controls, and the globalization of finance.

In the process, finance's ”contribution”-if that's the word-to the U.S. gross domestic product has soared from 2.5 percent in 1947 to 4.4 percent in 1977 to 7.7 percent in 2005. By that time financial firms accounted for upwards of 40 percent of the earnings of the companies listed in the S&P 500, and these firms' share of the total S&P 500 market capitalization doubled to approximately 25 percent. Even more startling, the combined income of the nation's top twenty-five hedge fund managers exceeded the compensation of the combined income of the CEOs of all companies listed in the S&P 500. In 2008 no less than one in every thirteen dollars in compensation in the United States went to people working in finance. By contrast, after World War II a mere one in forty dollars in compensation went to finance workers.

This outsize and excessive growth of the financial system did little to create any ”added value” for investors. While many hedge funds, investment banks, private equity funds, and other a.s.set managers claimed that they could provide investors with superior ”alpha” returns (in other words, bigger returns than those provided by more traditional a.s.set managers), ”schmalpha,” not ”alpha,” became the norm. These high-flying a.s.set managers often got higher returns, but investors saw little of it, because the managers charged higher fees for their allegedly superior services.

The various players in the financial system parted investors from their money in other ways too. Take securitization: at every step of the process, someone-a mortgage broker, an originating bank, a home appraiser, a broker dealer, a bond insurer, a ratings agency-charged high fees for its ”services” and transferred the credit risk down the chain. But it was an oligopoly of investment banks that profited the most from this arrangement, exploiting the lack of transparency about these operations to extract profits from credulous investors, most of which ended up in the pockets of these firms' employees rather than those of the shareholders of the firms.

The cancerous growth of finance has arguably had significant social costs too, as innovation and creativity have fled from manufacturing and other old-fas.h.i.+oned industries in favor of Wall Street. Indeed, since the 1970s, as our colleague Thomas Philippon has revealed, finance has attracted an ever-growing number of intelligent, highly educated workers. As compensation soared, graduates of elite schools increasingly went to Wall Street. In fact, among Harvard seniors surveyed in 2007, a whopping 58 percent of the men joining the workforce were bound for jobs in finance or consulting. In a curious paradox, the United States now has too many financial engineers and not enough mechanical or computer engineers.

Not coincidentally, the last time the United States saw comparable growth in the financial sector was in the years leading up to . . . 1929. In the 1930s, compensation in the financial sector plummeted, a victim of regulatory crackdowns that made banking a boring, if more respectable, profession. Reforming today's warped compensation structure is a necessary first step toward making banking boring once more.

Making Better Sausage.

Compensation is hardly the only problem that cries out for reform; the elaborate system of securitization that helped cause the recent crisis must be fixed as well. In the originate-and-distribute model of securitization (see chapter 3), a potentially risky a.s.set-a subprime mortgage, for example-was pooled with similar a.s.sets and turned into securities that would be sold to investors better able and willing to tolerate the risk.

One obvious flaw with this system was that it reduced incentives for anyone to actually monitor the creditworthiness of the borrower. Instead, the various players in the securitization process pocketed a fee while transferring most, if not all, of the risk to someone else. Everyone was complicit in this chain: the mortgage broker who handled the initial loan; the home appraiser, who had every incentive to give inflated values; the bank that originated the mortgage and used it to make mortgage-backed securities; the investment bank that repackaged these securities into CDOs and far more esoteric investments; and the ratings agencies that bestowed coveted AAA ratings along the way; and the monolines that insured those toxic tranches.

Any solution to the problem of securitization must somehow force these different players to more carefully consider the risks involved. In other words, each player must somehow be encouraged to pay more attention to the quality of the underlying loans. One way to do so is to force intermediaries-the originating bank and the investment banks-to hold on to some of the MBSs or CDOs in question. Forcing them to retain some risk, the thinking goes, will induce them to do a better job of monitoring the creditworthiness of the original borrowers (and leaning on the mortgage brokers and others who serve as the first link in the chain).

A number of proposals in circulation push this idea. Some came out of international bodies, including working groups within the G-20; others are homegrown, such as the Credit Risk Retention Act, which pa.s.sed in the House of Representatives in December 2009. This legislation proposes that banks involved in creating a.s.set-backed securities (not only mortgages, but any number of loans) be forced to retain 5 percent of the securities they create; a separate proposal in the Senate would increase that figure to 10 percent. Both proposals wisely forbid the banks to hedge or transfer any risk that they incur by retaining these securities.

Unfortunately, these low amounts of retained risk may be insufficient to change behavior. In the recent crisis, many banks and other financial inst.i.tutions maintained a significant exposure to the various securities that they had a hand in creating. Most of the AAA supersenior tranches of CDOs, for example, were retained rather than sold to investors. At the time of the crisis, in fact, approximately 34 percent of all the a.s.sets of major banks in the United States were real-estate-related; the figure for smaller banks was even higher, at roughly 44 percent. The originate-and-distribute model transferred some risk, but it certainly didn't transfer all of it; most financial inst.i.tutions had plenty of skin in the game. Otherwise, they would not have sustained the losses they did.

Firms retained that risk because traders made money by doing so. For this reason, relying on retained risk or ”skin in the game” as the princ.i.p.al method of reforming securitization is questionable. While it's a useful complement-and certainly will focus attention down the line on the risk incurred by holding such a.s.sets-it's unlikely to be a cure-all. Traders may gladly comply with requirements to retain risk, particularly if they can find a way of doing so that yields a bigger bonus. But a bigger bonus, as we've emphasized already, is no guarantee of stability.

Forcing firms to retain risk won't do much to resolve an even more pressing problem: the fact that securitization has, despite government subsidies, all but ceased. The reason it remains comatose is that even now, it's not really clear what went into the alphabet soup of securities that fed the boom. Indeed, securitization in the go-go years was a bit like sausage making before the creation of the Food and Drug Administration: no one knew what went into the sausage, much less the quality of the meat. And so it remains today: financial inst.i.tutions can still churn out the sausage, but given what we know might (or might not) go into these things, is it any surprise that investors have lost their appet.i.tes?

Some people believe that securitization should be abolished. That's shortsighted: properly reformed, securitization can be a valuable tool that reduces rather than exacerbates systemic risk. But in order for it to work, it must operate in a far more transparent and standardized fas.h.i.+on than it does now. Absent this s.h.i.+ft, accurately pricing these securities, much less reviving the market for securitization, is next to impossible. What we need are reforms that deliver the peace of mind that the FDA did when it was created.

Let's begin with standardization. At the present time, there is little standardization in the way a.s.set-backed securities are put together. The ”deal structures” (the fine print) can vary greatly from offering to offering. Monthly reports on deals (”monthly service performance reports”) also vary greatly in level of detail provided. This information should be standardized and pooled in one place. It could be done through private channels or, better, under the auspices of the federal government. For example, the SEC could require anyone issuing a.s.set-backed securities to disclose a range of standard information on everything from the a.s.sets or original loans to the amounts paid to the individuals or inst.i.tutions that originated the security.

Precisely how this information is standardized doesn't matter, so long as it is done: we must have some way to compare these different kinds of securities so that they can be accurately priced. At the present time, we are stymied by a serious apples-and-oranges problem: the absence of standardization makes comparing them with any accuracy impossible. Put differently, the current system gives us no way to quantify risk; there's far too much uncertainty.

Standardization, once achieved, would inevitably create more liquid and transparent markets for these securities. That's well and good, but a few caveats also come to mind. First, bringing some transparency to plain-vanilla a.s.set-backed securities is relatively easy; it's more difficult to do so with preposterously complicated securities like CDOs, much less chimerical creations like the CDO2 and the CDO3.

Think, for a moment, what goes into a typical CDO2. Start with a thousand different individual loans, be they commercial mortgages, residential mortgages, auto loans, credit card receivables, small business loans, student loans, or corporate loans. Package them together into an a.s.set-backed security (ABS). Take that ABS and combine it with ninety-nine other ABSs so that you have a hundred of them. That's your CDO. Now take that CDO and combine it with another ninety-nine different CDOs, each of which has its own unique mix of ABSs and underlying a.s.sets. Do the math: in theory, the purchaser of this CDO2 is supposed to somehow get a handle on the health of ten million underlying loans. Is that going to happen? Of course not.

For that reason, securities like CDOs-which now go by the nickname of Chern.o.byl Death Obligations-must be heavily regulated if not banned. In their present incarnation, they are too estranged from the a.s.sets that give them value and are next to impossible to standardize. Thanks in large part to their individual complexity, they don't transfer risk so much as mask it under the cover of esoteric and ultimately misleading risk-management strategies.

In fact, the curious career of CDOs and other toxic securities brings to mind another, less celebrated acronym: GIGO, or ”garbage in, garbage out.” Or to return to the sausage-making metaphor: if you put rat meat and trichinosis-laced pig parts into your sausage, then combine it with lots of other kinds of sausage (each filled with equally nasty stuff), you haven't solved the problem; you still have some pretty sickening sausage.

The most important angle of securitization reform, then, is the quality of the ingredients. In the end, the problem with securitization is less that the ingredients were sliced and diced beyond recognition than that much of what went into these securities was never very good in the first place. Put differently, the problem with originate-and-distribute lies less with the distribution than with the origination. What matters most is the creditworthiness of the loans issued in the first place.

That's why reform should focus on the circ.u.mstances in which loans are originated. It's not as if the regulatory apparatus for doing so isn't in place. In the United States, the Federal Reserve, the FDIC, the Office of Thrift Supervision, the Office of the Comptroller of the Currency, and the National Credit Union Administration all have jurisdiction to oversee and regulate the sorts of loans that end up in various kinds of a.s.set-backed securities. The existing regulations and guidelines must be beefed up and given real teeth to ensure that what ends up in the securitization pipeline isn't toxic.

The Federal Reserve has already taken steps in that direction, proposing significant changes to Regulation Z (also known as Truth in Lending). These changes would make it significantly easier for potential borrowers to recognize the true costs of the mortgages they're a.s.suming. It would also place restrictions on those originating the loans. Compensation of mortgage brokers and loan officers would no longer be linked to the interest rate of the loan, much less any of the other terms. Likewise, mortgage brokers and loan officers would be expressly forbidden to steer consumers to bigger or more expensive loans, simply to increase their compensation.

The changes would be sensible, but cleaning up securitization requires that policy makers consider another important aspect of sausage making: the meat inspectors who grade these products. Their financial equivalent would be the ratings agencies, and like their counterparts in the USDA, they haven't always lived up to their responsibilities.

Reforming Ratings.

In the United States, three major private ratings agencies-Standard & Poor's, Moody's Investors Service, and Fitch Ratings-wield remarkable power, slapping grades on everything from mortgages to corporate bonds to the sovereign debt of entire nations. These grades reflect the likelihood that the borrower or borrowers will default on their debt, and they are central to how financial markets ascertain risk. In effect, ratings are a way to outsource due diligence: if Moody's says a particular CDO tranche is supersafe and gets an AAA rating, then it spares anyone else from having to pop the hood of the security and look at the underlying a.s.sets.

The ratings agencies' rise to power began in the 1930s. Their forerunners issued ratings that federal regulators used to a.s.sess the quality of bonds held by banks. This stamp of government approval helped cement their influence, and while their power waned in the immediate postwar era, it rose again in the 1970s, a time when bond defaults rose, and ratings became increasingly important to evaluating risk.

In 1975 the Securities and Exchange Commission created a category known as Nationally Recognized Statistical Rating Organization (NRSRO). Fitch, Standard & Poor's, and Moody's were among those granted this coveted designation. In effect, anyone selling debt had to get a rating from one of these specially designated agencies. While the SEC eventually recognized seven of these agencies, mergers reduced their ranks to the three familiar firms, though the SEC has recently given this designation to a handful of lesser-known companies.

Over the course of their careers, the ratings agencies have changed dramatically. In the early years, they made their money from investors, who paid them to evaluate potential investments. Over time this revenue model s.h.i.+fted, partly because some investors obtained ratings by photocopying their friends' rating manuals instead of paying for them. To get around that problem, the ratings agencies adopted a new business model: they would sell their services to issuers of debt rather than to investors. At the same time the SEC's reforms effectively put the onus of obtaining a rating on anyone who wanted to sell debt. By the 1980s the transition was complete: issuers of debt now paid for ratings.

This arrangement, however, created a ma.s.sive conflict of interest. Banks looking to float some securities could shop around among the agencies to find the best rating. A ratings agency that looked at a proposed offering and slapped a subprime rating on it risked losing business. Increasingly, the ratings agencies had an interest in giving the customers what they wanted-and if a customer wanted a AAA rating for an MBS made up of subprime mortgages, there's a good chance that's what it got.

As if that weren't bad enough, the ratings agencies started to generate revenue from other, equally problematic sources. A bank putting together a structured financial product would go to one of the ratings agencies and pay for advice on how to engineer that product to attract the best possible rating from the very agency the bank would ultimately pay to rate its securities. This service was described as ”consulting” or ”modeling.” Perhaps. In fact, it was a bit like a professor's accepting a fee in exchange for telling students how to get an A on an exam. That's not kosher.

How, then, might the ratings agencies be reformed? At a bare minimum, the agencies must be forbidden to offer any consulting or modeling services. They should exist for one purpose: to a.s.sign a rating to a debt instrument. That's it; anything more introduces a possible conflict of interest. While the SEC has issued rules that forbid the ratings agencies to consult with the companies they rate, this ban is extraordinarily difficult to enforce. Instead, the SEC should forbid the ratings agencies to consult or model for anyone.

It also makes sense to open up compet.i.tion in this privileged realm. While this proposal might have been difficult to justify ten years ago, when the collective reputation of the big three was still intact, it's an easier sell now. Unfortunately, the SEC makes it very difficult for new companies to obtain that coveted NRSRO rating: newcomers must have been in business for several years and have many major clients. But it's hard to get major clients without first being inducted into the sacred circle. To address this problem, the SEC needs to lower the barriers to entry, so that more compet.i.tion-free market compet.i.tion, if you will-enters into this vitally important industry.

More radical still would be to take away the semiofficial role that the ratings agencies now enjoy. Everything from SEC regulations to Basel II capital requirements formally recognize the NRSROs as the only place from which ratings can be obtained. That recognition invests them with disproportionate, if not excessive, power. Taking that power away would be another way of opening up compet.i.tion.

An even more comprehensive reform would be to force the ratings agencies to return to their original business model, in which investors in debt-not the issuers of it-pay for the ratings. Unfortunately, this is easier said than done. One reason is the ”free rider problem”: once a set of investors pays for a rating and makes a decision based on it, other investors can figure out the rating and make their own decision free of charge.

<script>