To understand the crisis, it is first necessary to understand the reasons for the creation and rapid growth of structured products in the credit markets. Beneath the new jargon and greater mathematical complexity of today’s financial markets, investment behaviour is little changed. Banks, as they have done for centuries, have borrowed short to lend long. The difference in this cycle has been that the financial innovation of securitisation has allowed a wider range of investors to ride the yield curve and permitted banks to shift traditional areas of business off balance sheet.
There is a good reason for banks to move loans off their books: securitisation permits the transfer of risk to those who can best bear it. But less valid reasons have also been influential in the growth of the “originate and distribute” banking model. Under Basel regulations, direct loans must be backed by regulatory capital. Shifting loans off balance sheet to Structured Investment Vehicles (SIVs) permits regulatory arbitrage. In the most extreme case, where the sponsoring bank does not extend a formal credit line to the SIV, but merely an implicit obligation to help in the event of financing difficulties – described by one member of the panel as “reputational liquidity enhancements” – then no regulatory capital is necessary.
The originate and distribute model led banks to focus only on their “pipeline risk” – the stock of loans awaiting sale. The distance between lenders and borrowers increased and business became less transparent. Until this July, opaqueness was a term more usually applied to describe hedge funds. Now it applies also to banks, despite the fact that they are regulated. With the banks retaining less risk, lending standards deteriorated. In the mortgage market, loans were made to NINJAs, individuals with no income, no job or assets, helping fuel the US housing market bubble. In the corporate bond market, the frenzy to finance leveraged buyouts promoted the growth of PIK and Covenant-lite bonds.
Demand for structured products, meanwhile, was underpinned by the search for yield in a low volatility market environment supported by benign macroeconomic conditions. Stable growth, low inflation and the reflationary policies of central banks after the collapse of the tech bubble produced an unprecedented increase in liquidity. The shift towards liability driven investment and the recycling of ballooning foreign exchange reserves in countries with currencies pegged to the dollar reinforced a shift into long-dated fixed income securities. As the trend became established volatility declined. This encouraged strategies selling volatility, which in turn reduced realised volatility further, reinforcing the cycle and prompting investors to increase leverage.
In such an environment structured products provided a creative way to use leverage to enhance portfolio returns and track records. The lack of liquidity in many structured credit products did not detract from their appeal. Reliance on “mark-to-model” valuations helped smooth the volatility of returns and improve Sharpe ratios, in a process described by Warren Buffett as “marking to myth.”
Holdings of Level 3 assets – securities that are so illiquid, or trade so infrequently, that they have no reliable price – increased. These leveraged products at the bottom of the capital structure can be neatly encapsulated in the phrase “return free risk securities.”
The search for yield produced a frantic hunt for new people to lend to. It is unsurprising, therefore, that problems arose first in the US sub-prime market. Loans were made that never should have been made and securitised products were built around those bad credits. Through securitisation, which was a gargantuan fee-generating machine for the banks, the market had become an enormous ratings-sensitive jigsaw puzzle based on backward-looking credit ratings. Ultimately there was a disconnection between the backward-looking ratings and the actual credit performance of sub-prime loans.
More complicated, however, was the transformation of a credit problem in one part of the US mortgage market into a far wider liquidity problem. US sub-prime is a $1.5 trillion segment of the $10 trillion mortgage market. Even if the extreme assumption were made that 50 per cent of sub-prime loans default and only 50 per cent were recovered from those defaulted loans, then total losses would amount to $375 billion. Such a loss is big, but not large enough in comparison with the value of losses witnessed in sharp equity market corrections to explain the seriousness of this summer’s crisis.
To explain how the sub-prime crisis was transmitted, first to the Asset Backed Commercial Paper market and then the interbank lending market requires an understanding of the amplification mechanisms at work in the financial system.
Recent academic work by Professor Brunnermeier and Lasse Heje Pedersen of New York University links an asset’s market liquidity ie. the ease with which it can be traded, with traders’ funding liquidity, ie. the ease with which they can obtain funding1.
Different market participants face slightly different versions of funding liquidity risk. Hedge funds that have funded trades through a prime broker face the risk of margin calls and are also vulnerable to investor redemptions. Banks, too, face a redemption risk in the form of withdrawals by depositors. SIVs must contend with the risk of not being able to roll over short-term commercial paper requirements. In each case, the risk is that short-term borrowing that has been used to financing long-term lending is withdrawn.
If funding liquidity is readily available, margins will be low and leverage correspondingly high. During a market crisis, however, increased volatility and permanent price shocks (ones that are not expected to be quickly reversed) raise financiers’ expectations of future volatility. In turn this raises banks’ Value-at-Risk at the same time as collateral values are declining. The end result is that funding liquidity dries up as margins and haircuts are increased. Between January and May 2007, for example, haircuts on investment grade bonds were between zero-3 per cent, but increased to 3-7 per cent in July-August. Similarly, between the same periods, haircuts on second lien leveraged loans increased to 20-30 per cent from 15-20 per cent.
When funding liquidity is tight, traders become reluctant to take on positions, especially in capital-intensive high-margin securities. This reduces market liquidity and increases volatility. In these circumstances margins are a destabilising mechanism because they set in train two self-reinforcing liquidity spirals – which Professor Brunnermeier describes as a margin spiral and a loss spiral. Initial losses create funding problems for traders forcing them to reduce their positions, which drives prices lower. As prices and collateral values fall, traders are forced to put up more margin to finance their positions. This prompts further asset sales to raise capital, which drives prices down even further, adding to losses on existing positions.
The margin and loss spirals also offer a plausible explanation for why market turmoil spread to quantitative hedge funds during the first week of August. Faced with the need to reduce market exposure, funds chose to close out relatively liquid positions first. For a multi-strategy fund, this shifted liquidation into popular strategies, even though they were unrelated to credit and the rationale for holding them remained valid.
Additional reinforcing mechanisms were also evident this summer. Problems with the value of collateral were exacerbated by ‘asymmetric information’ regarding the quality of that collateral. This is a version of the ‘lemons’ problem, outlined by George Akerlof in 1970. If buyers cannot tell the difference between the good and the bad (lemons) in a collection of assets, they will only offer an average price for the pool of collateral. Faced with this, sellers will withdraw the better collateral from the pool, prompting buyers to offer even lower prices and so on, until the market ultimately breaks down. The problems SIVs faced this summer in rolling over Asset Backed Commercial Paper programmes fit into such a pattern.
Uncertainty has not been confined to the quality of collateral. Banks were faced with three additional areas of uncertainty. First was an increase in credit counterparty risk. Second, the prospect that SIVs might be forced to draw on credit lines and third, that other banks with whom they conducted business in the interbank lending market might have SIV exposure. The net result was ‘precautionary hoarding’ to build up a funding cushion – behaviour which explains the persistently high premium between Libor market interest rates and official policy rates.
A more extreme version of this mechanism might also have been at work. When it is difficult to assign probabilities to different outcomes, such as the value of CDOs, investors face ‘Knightian uncertainty’ – risks that cannot be measured. In these circumstances investors and banks focus on the worst-case scenario.
The same uncertainty that led banks to hoard cash also helps explain the spillover from sub-prime into corporate credit markets, despite the fact that, as yet, there has been no uptick in corporate default rates. There had been, however, a serious erosion of credit underwriting standards – leverage had increased markedly, both in first lien loans and through the explosive growth of second lien loans and new products such as ‘PIK Toggle’ notes. Securitisation drove demand for these products, with over half of all new leveraged loans in the run-up to the crisis being bought by CLOs. When the dislocation in the sub-prime market destroyed confidence in the ratings agencies, distrust was fostered in other ratings-based securitised products.
The close correlation in August between movements in the ABX index, which represents a basket of credit default swaps on high-risk mortgages and the CDX high yield index, which tracks the costs of debt insurance for sub-investment grade companies in the US, clearly illustrates this spillover effect (Fig.1).
Brunnermeier also identifies an amplification mechanism based on market timing, in which self-interest prompts investors to cut and run before others do. The most extreme example is a withdrawal of deposits from a bank, as seen with the run on Northern Rock. But the refusal to roll over SIVs’ ABCP programmes is part of the same phenomenon. Indeed, the liquidation policy of many SIVs makes a run more likely, given that early withdrawals are paid in cash, while those remaining may have to wait for forced assets sales to get their money out. Hedge funds face a similar dynamic – both through increased margin calls by prime brokers and investor redemptions.
There is also a risk of ‘gridlock’, because of the interlocking nature of financial obligations. With many financial institutions fulfilling the role of lender and borrower at the same time, there is the risk that an institution will not be able to meet its obligations until the debts it is owed have been paid. The opaqueness of the system makes matters worse. Without full knowledge of where the losses lie, it is less likely that a third party buyer will step in to help out.
The potential for a problem in one part of the credit markets to be amplified throughout the entire system makes it difficult to conclude that the crisis is over. The next chapter, which is when sub-prime defaults lead to an increase in corporate defaults in consumer facing industries, has yet to be seen. Moreover, because of the weakness of covenants and the higher debt multiples applied to leveraged loans, by the time a company is forced into default it is likely to be in worse shape than in previous credit cycles. Recovery levels will be lower and some junior debt will be wiped out completely.
Policymakers have responded to the risk of further market turmoil by injecting liquidity, albeit reluctantly in the case of the Bank of England, which only overcame its fears of moral hazard after the run on Northern Rock. The Federal Reserve, meanwhile, has cut the Fed Funds rate by 75 basis points in an attempt to limit the impact of credit market dislocation on the wider economy.
Those spillover effects could follow one of two paths. In the more optimistic scenario the US economy muddles through, but the risk remains of a more acute slowdown in which consumption declines, stock markets fall, the Fed cuts rates further and the dollar continues to depreciate. A falling dollar in turn reduces the attractiveness of long-term US fixed income assets for managers of foreign exchange reserves. Without the depressing effect on long-term yields of this overseas demand, there could be an adverse impact on mortgage rates, providing a negative feedback loop into housing and consumer markets.
The panel believed that a period in which the search for yield had led to insensitivity to risk had come to an end. A return to more normal conditions in long-dated fixed income markets is now in prospect, as the distorting effect of the buying of long-dated US assets by reserve managers diminishes. The dysfunctional behaviour of buying the most illiquid debt securities and leveraging them as much as possible was the most highly rewarded in 2006 and the first half of 2007.
Hedge funds have been among the casualties of the crisis, but there are plenty of opportunities for non-traditional investors in current market conditions. Hedge funds’ ability to redeploy capital, for example by moving into the market for bank debt, can help reduce market volatility. Multi strategy hedge funds are not constrained by mandate in the same way as traditional institutions and can therefore fill the liquidity void created when those traditional buyers withdraw from the market. The share of loan syndications bought by non-traditional investors, for example, has increased sharply during the credit dislocation (see Fig.2 & 3).
The speed and flexibility of opportunistic hedge funds provides them with a role in shifting the backlog of bridge loans on banks’ balance sheets as well as helping finance large new deals where investment bank risk appetite has diminished. As corporate default rates increase there will also be opportunities for distressed debt funds.
There is indeed more to come from the 2007 liquidity crunch, although the panel believed it would not be of a catastrophic systemic nature. Hedge funds, meanwhile, are expected to play an important role in restoring equilibrium to capital markets.
1. ‘Deciphering the 2007 Liquidity Crunch’ by Professor Merkus Brunnermeier, Princeton University
Commentary
Issue 33
The Liquidity Crunch
Is it all over or is there still more to come?
VINCE HEANEY
Originally published in the December 2007/January 2008 issue