A generation has prospered from the wholesale transfer of risk through securitization. Now it is paying the price“THE medium-term outlook for the company is very positive,” declared Northern Rock’s chief executive, Adam Applegarth, unveiling its first-half results in July. He spoke of a credit book that was “robust”. Who would have guessed that Britain’s fifth-largest mortgage lender would be fighting for its life, its branches besieged by customers demanding their savings back?
Disclaimer: This article was written originally in September 2007, so some data points may be outdated.
The run on Northern Rock is the most dramatic symptom of the contagion gripping the financial markets. Here was a bank that had grown rich from the innovations of recent years, using abundantly stocked wholesale markets to fund its lively growth, using those same markets to offload bits of its loan book as and when they became unattractive.
But the very innovations on which Northern Rock thrived have savaged its business. The company does not lend to speak of overseas. Nevertheless, its fate was determined by the distant turmoil in America’s mortgage market. When that spilled over into the securities markets, the money markets that Northern Rock had depended on for years dried up in a single day at the start of August.
The brave new world that enabled banks like Northern Rock to grow so fast is founded on “securitization”—the process that transforms mortgages, credit-card receivables, and other financial assets into marketable securities—and the innovation it spawned in “structured” products. This was a revolution that brought huge gains. But across the financial world investors and regulators are asking themselves whether it also brought costs that are only now becoming clear.
In America, the President’s Working Group on Financial Markets, which includes the Federal Reserve and the Treasury, has launched a probe into securitization and the rating agencies, which monitor it. IOSCO, a global regulators’ body, has begun an investigation of its own. And the two biggest securitization clubs on either side of the Atlantic held an emergency summit this week to look for ways to coax investors back into moribund asset-backed markets.
It is hard to overstate the extent of this reversal in fortunes, if only because it is hard to overstate the effect that securitization has had on financial markets. Until the early 1980s, finance hewed to an “originate and hold” model. When that spilled over into the securities markets, the money markets that Northern Rock had depended on for years dried up in a single day at the start of August. This began to give way to an “originate and distribute” model after America’s government-sponsored mortgage giants issued the first bonds with payments tied to the cash flows from large pools of loans.
Wall Street built on this innovation, and securitization took off soon after, then paused before exploding in the 1990s (see chart 1). It was given a lift by America’s savings-and-loan crisis, which encouraged mortgage lenders to jettison their riskier loans, and by new technologies, such as credit scoring, that facilitated loan pooling. Around 56% of America’s outstanding residential mortgages were packaged in this way, including more than two-thirds of the subprime loans issued in 2006. Thanks largely to securitization, global private-debt securities are now far bigger than stock markets.
Banks have come to see securitization as an indispensable tool (see chart 2). Global lenders use it to manage their balance sheets since selling loans frees up capital for new business or for return to shareholders. Small regional banks benefit too. Gone are the days when they had no choice but to place concentrated bets on local housing markets or industries. Now they can offload credits to far-away investors such as insurers and hedge funds, which have an appetite for them.
Michael Milken, of junk-bond fame, called securitization the “democratization of capital”. Studies suggest that the explosion of this “secondary” market for bank debt has helped to push down borrowing costs for consumers and companies alike. There are other “systemic” gains, too. Subjecting bank loans to valuation by capital markets encourages the efficient use of capital. And the broad distribution of credit risk reduces the risk of any one holder going bust.
Even in the midst of turmoil, it is hard to find a banker, regulator, or academic who wants to see the clock turned back. But the crisis has exposed cracks in the new model that were hidden or ignored during the credit bubble. The three most glaring are complexity and confusion, a fragmentation of responsibility, and the gaming of the regulatory system. Take each in turn.
Too clever by 50 basis points
The past few weeks have shown that financiers did not fully understand what they were trading. The boom in derivatives was one of those moments when financial engineering raced ahead of back offices and risk-management departments, leaving them struggling to value or account for their holdings. Pierre Pourquery, of Boston Consulting Group, says it is not uncommon for investors to break their exotic purchases into smaller pieces in order to feed them into their risk-management systems. This brings new risks, particularly that the parts will behave differently from the whole under stress.
Steven Schwarcz, a professor at Duke University and writer on securitization, has come across contracts that are so convoluted that it would be impractical for investors to try to understand them: they would have to spend more money hiring experts to deconstruct them than they could ever hope to earn in extra returns.
In a recent paper* on credit derivatives, David Skeel and Frank Partnoy concluded that collateralized debt obligations (CDOs), one of the most common derivatives, are too clever by half. The transaction costs are high, and the benefits are questionable. They conclude that CDOs are being used to transform existing debt instruments that are accurately priced into new ones that are overvalued.
Complexity confuses investors about the risks they are taking on. The more eclectic CDOs bind together the fate of assets that have few real economic links. Imagine a lowly-rated energy bond and top-notch bank paper in the same structure. Separately, they would not normally move in tandem. Put them in a CDO, however, and in a credit squeeze they fall together, by virtue of being in the same murky structure, as investors rush for the exit or seek to hedge their risks.
The lack of transparency plagues the bundling of loans into securities, too. These days, for instance, lenders are less likely to foreclose on defaulting borrowers: in America, less than a quarter of loans 90 days late or more are in foreclosure, compared with three-quarters in the late 1990s, points out Charles Calomiris, of Columbia University. When a late payer gets back on track his loan is once again labeled “current”, and his chequered history does not have to be fully disclosed when the loan goes into a securitized pool. So even the most diligent buyer would struggle to spot that some of the “prime” collateral of mortgage-backed bonds was, in fact, of questionable quality.
Investors seeking redress have encountered unforeseen problems. Securitizations are generally structured as “true sales”: the seller wipes its hands of the risk. In practice, buyers have some protection. Many contracts allow them to hand back loan pools that sour surprisingly quickly. Some have done this with the most rotten subprime mortgages, requesting better-quality loans into the pool. But there were so many bad loans that originators could not oblige. “What we thought was an effective secondary-market punishment mechanism turns out to be faulty when the problem grows beyond a certain size,” says Anthony Sanders, a subprime expert.
The second lesson of the past few weeks is that securitization has warped financiers’ incentives. It is sometimes portrayed as bank “disintermediation”, but in fact, it replaces one middleman with several. In mortgage securitization, for instance, the lender is supplanted by the broker, the loan originator, the servicer (who collects payments), the investor, and the arranger, not to mention the rating agencies and mortgage-bond insurers.
This creates what economists call a principal-agent problem. The loan originator has little incentive to vet borrowers carefully because it knows the risk will soon be off its books. The ultimate holder of the risk, the investor, has more reason to care but owns a complex product and is too far down the chain for monitoring to work. For all its flaws, the old bank model resolved the incentives in a simple way. Because loans were kept in-house, banks had every reason to underwrite cautiously and keep tabs on the borrower after the money left the vault.
Investors in loan-backed securities could have pushed for tougher monitoring. But most were too taken with the alluring yields on offer—an addiction Alan Greenspan, the Fed’s former chairman, has likened to cocaine abuse. Debt investors are usually sober types, but as the bubble grew, it was increasingly their urges, and not the creditworthiness of homeowners, that determined loan-underwriting standards.
Wall Street took full advantage of this appetite. It was well known that investors such as Germany’s IKB, a lender to small companies which was bailed out last month, had a weakness for exotic products. The securities firms peddling mortgage-backed bonds did little to disabuse them of the notion that a CDO with a high rating must be as safe as houses—after all, the buyers were sophisticated institutions, not widows or orphans.
Moreover Wall Street has every reason to shovel securitised debt out as fast as it can. The loan-origination platform has high fixed costs, so it is a scale business. This can lead to trouble when there are not enough creditworthy new borrowers, as in subprime lending. Banks may be tempted to keep feeding the machine at the expense of laxer lending standards. “Once you get into it, it’s a bit like heroin,” says Joseph Mason, an academic who has written on securitization. When AAA paper is repeatedly compared to Class A drugs, you know something is wrong.
Rating the raters
Complexity and warped incentives foster the third cost of securitization: gaming the regulations. Politicians are scrutinizing the role of rating agencies, as they did with auditors after the dot-com bubble burst. Regulatory dependence on ratings has grown across the board. Banks can reduce the amount of capital they have to set aside if they hold highly rated paper, for instance, and some investors, such as money-market funds, must stick to AAA-rated securities. But not all top-rated paper is the same.
The agencies appear to have been too free in giving out prized AAA badges to structured products, especially CDOs. This was partly because their models were faulty, failing to pick up correlations between different markets, and partly because of a conflict of interest: theirs is one of few businesses where the appraiser is paid by the seller, not the buyer.
This made it easier for the banks securitizing and further repackage debt to create the greatest possible number of securities with the lowest regulatory cost (that is, the highest rating). Investors restricted to investment-grade paper assumed (or at least hoped) that the rating was a guarantee of strength. It might have helped if the agencies had properly monitored their ratings after issuing them. But with low fees per security, there is little incentive to stay on the case.
From dodgy American mortgages to the British high street
Avinash Persaud, of Intelligence Capital, a consultancy, argues that securitization has let banks (as regulated “holders” of credit risk, with the capacity to keep it through bad times) pass it on to unregulated “traders” of risk with smaller balance sheets, such as hedge funds, which sell when trouble strikes. As a result, he says, although the risk of bank runs has fallen, the risk of market runs has increased.
In fact, some hedge funds are patient investors in illiquid assets. But Mr. Persaud is right that risk changes, chameleon-like, depending on the holder. A bank with a plump capital cushion can use its balance sheet to hold on to out-of-favour credits until markets regain their balance. An investment fund that is several steps removed from the borrower, vulnerable to margin calls, and constrained by daily risk models will have less room for maneuver. The rise of such investors has led, paradoxically, to more of a herd mentality because of “convergence in the way the risk has been diversified”, says Mr. Schwarcz.
Fans of securitization argue that it has made the system safer because risk ends up with those who want to shoulder it. It is true that few banks have failed in recent years, in spite of the Asian crisis, the failure of Long-Term Capital Management, the dotcom bust, and so forth. But this could have more to do with economic stability than securitization.
Indeed, André Cappon, a consultant, warns of the “circularity of risk”. The hedge funds that buy mortgage-linked debt also borrow heavily from the prime-brokerage arms of banks that originated many of the underlying loans. So a bank can push risk out of the front door, only to find it sneaking through the back.
Shine a light
What should banks and regulators do about all this? In the short run, the focus will be on transparency. Investors need to know who is holding what and how should it be valued. One idea is to force investment banks to reveal more about the performance and price of privately traded asset-backed instruments. Another is for the Securities and Exchange Commission, Wall Street’s primary regulator, to ensure the consistent valuation of such assets across firms. More information on the vehicles that issue asset-backed commercial paper would also help.
There will also be calls for greater standardization of structured products. This could undermine the “over the counter” (off-exchange) markets. Exchange groups like the Chicago Mercantile Exchange and Deutsche Börse have been trying to take business away from the banks, offering centrally cleared foreign-exchange trading and credit products. Regulators like these because they are more transparent than privately traded deals. The crisis could tip the balance in the exchanges’ favor, says Benn Steil, a market-watcher at the Council on Foreign Relations.
If investors continue to shun the most complex products, Wall Street will have to offer simpler fare. Tom Zimmerman, head of asset-backed research at UBS, sees parallels between the CDO bust and the blow-up in collateralized mortgage obligations (CMOs) in the mid-1990s. CMOs, which pool prepayment risk, became so convoluted that investors could no longer see where the dangers lay. When they stopped buying, investment banks made the product more straightforward and it took off again.
As for the rating agencies, they have probably grown more powerful than anyone intended. Regulators will want to see their interests aligned more closely with investors and to ensure that they are quicker and more thorough in reviewing past ratings. Moody’s is thinking about adding new letter codes to cover liquidity and volatility risk for complex products.
Fixing the problem of fragmented responsibility will be a balancing act. Mr Sanders thinks subprime default rates would not have spiked if loan originators had been forced to set aside capital to cover, say, 10% of each securitized pool. But framing the terms of this sort of co-insurance would be tricky. Would 10% be enough? Or too much? Should a reputable bank have to pay as much as a small specialist lender?
With investors, too, there is a risk of heavy-handedness. Calls for federal legislation on “assignee liability”, which would hold secondary-market investors liable for loan losses, could do more harm than good. The soul-searching has led some analysts to speculate that mortgage lending could, for a time, go back to being a bank-led lend-and-hold market.
But do not expect a rush back to the ways of the 1960s. Securitization has become far too important for that. Indeed, it has not yet fulfilled its promise. Wall Street eggheads may be licking their wounds at present, but they will soon be coming up with even more products. And, given time, there will no doubt be another wave of buying. More importantly, the transformation of sticky debt into something more tradable, for all its imperfections, has forged hugely beneficial links between individual borrowers and vast capital markets that were previously out of reach. As it comes under scrutiny, the debate should be about how this system can be improved, not dismantled.