Liquidity risk and the current crisis
What is liquidity? Why is it at the heart of the crisis? How can we fix it? This column explains it all in terms any trained economist can understand.
Search the site
What is liquidity risk and how can it help us understand the current crisis? How do we solve the crisis - and which measures will only hurt?
Here I provide some answers. The column is based on my 20 October 2008 talk at the International Monetary Fund and the Federal Reserve Board. To see the slides with lots of figures and graphs, click here.What is liquidity risk?
There are two kinds of liquidity: market liquidity, and funding liquidity.
With these notions in mind, the meaning of liquidity risk is clear.
For instance, a levered hedge fund may lose its access to borrowing from its bank and must sell its securities as a result. Or, from the bank's perspective, depositors may withdraw their funds, the bank may lose its ability to borrow from other banks, or raise funds via debt issues.We are experiencing extreme market and funding liquidity risk
Liquidity generally varies over time and across markets, and currently we are experiencing extreme market liquidity risk. The most extreme form of market liquidity risk is that dealers are shutting down (no bids!), which is currently happening in a number of markets such as those for certain asset-backed securities and convertible bonds. We are also experiencing extreme funding liquidity risk since banks are short on capital, so they need to scale back their trading that requires capital, and also scale back the amount of capital they lend to other traders such as hedge funds, that is, hedge funds now face higher margins. In short, if banks cannot fund themselves, they cannot fund their clients.
The two forms of liquidity are linked and can reinforce each other in liquidity spirals where poor funding leads to less trading, this reduces market liquidity, increasing margins and tightening risk management, thus further worsening funding, and so on.
An illiquid security has a higher required return to compensate investors for the transaction costs. Since market liquidity may deteriorate when you need to sell in the future, investors face market liquidity risk as discussed above. Investors naturally want to be compensated for this, so market liquidity risk increases the required return. Indeed, the liquidity-adjusted capital asset pricing model shows how liquidity betas complement the standard market beta. The higher required return in times of higher market liquidity risk leads to a contemporaneous drop in prices, according to this theory, consistent with what we are seeing in the current marketplace. An overview of the liquidity literature is available here.Liquidity risk and the current crisis: downward liquidity spirals
The trigger of the crisis was the bursting of the housing bubble, combined with a large exposure by the levered financial institutions. This led to significant bank losses with associated funding liquidity problems. This started the systemic liquidity spirals. As banks’ balance sheets deteriorated, they had to de-lever. To do this, they:
This put stress on the interbank funding market (as measured e.g. by the TED spread, see slides) as everyone was trying to minimise counterparty exposures.
Banks’ funding liquidity problems quickly spread. Other investors, especially those that rely on leverage such as hedge funds, face funding risk when banks become less willing to lend, they raise margins, and, in the extreme, when the banks fail as Lehman did.
When banks such as Bear Stearns and Lehman started to look vulnerable, their clients risked losing capital or having it frozen during a bankruptcy, and they started to withdraw capital and unwind positions, leading to a bank run.
This funding liquidity crisis naturally lead to market illiquidity with bid-ask spreads widening in several markets, and quoted amounts being reduced by dealers with less available capital. This market illiquidity, and the prospect of further liquidity risk, scared investors and prices dropped, especially for illiquid assets with high margins.
This is the downward liquidity spiral as illustrated in the chart.
Sources: Garleanu and Pedersen (2007) and Brunnermeier and Pedersen (2008)The crisis spreads to other asset classes
The crisis has been spreading across asset classes and markets globally. There are as traders unwind carry trades and lose faith in weak currencies. Further, as an example of the gravity of the liquidity crisis, the “Covered Interest Rate Parity” – the most basic arbitrage condition in the world’s thickest market (foreign exchange) – currently fails to hold even for the major currencies. This must mean that no one can arbitrage because no one can borrow uncollateralised, no one has spare collateral, and no one is willing to lend – arbitrage involves both borrowing and lending.
The increased risk and illiquidity has also lead to a spike in volatility, contributing to the higher margins. Further, correlations across assets have increased as everything started trading on liquidity.The crisis spreads to Main Street
Clearly, the crisis is having a significant effect on the real economy as homeowners see their property value deteriorate, consumers access to borrowing is reduced, main street companies face higher cost of equity and especially debt capital and a lower demand for their products, unemployment goes up, etc.What can - and what cannot - solve a liquidity crisis?
If the problem is a liquidity spiral, we must improve the funding liquidity of the main players in the market, namely the banks. Hence, banks must be recapitalised by raising new capital, diluting old equity, possibly reducing face value of old debt. This can be done with quick resolution bankruptcy for institutions with systemic risk, i.e. those causing liquidity spirals.
Further, we must improve funding markets and trust by broadening bank guarantees, opening the Fed's discount window broadly (giving collateralised funding with reasonable margins), and ensuring the Commercial Paper market function. Further, risk management must acknowledge systemic risk due to liquidity spirals and the regulations must consider the system as a whole, as opposed to each institution in isolation.
If we have learned one thing from the current crisis, it is that trading through organised exchanges with centralised clearing is better than trading over-the-counter derivatives because trading derivatives increases co-dependence, complexity, counterparty risk, and reduces transparency. Said simply, when you buy a stock, your ownership does not depend on who you bought it from. If you buy a “synthetic stock” through a derivative, on the other hand, your ownership does depends on who you bought it from - and that dependence may prevail even after you sell the stock (if you sell through another bank). Hence, when people start losing confidence in the bank with which they trade, they may start to unwind their derivatives positions and this hurts the bank’s funding, and a liquidity spiral unfolds.Banning short selling is a bad idea
In the debate about how to solve the crisis and prevent the next one, it has been suggested that policymakers should ban short selling and impose a transaction tax on stocks. I believe that neither is a good idea. First, short sellers bring new information to the market, increase liquidity, and reduce bubbles (remember the housing bubble started this crisis) so preventing this can be very costly and prohibiting short sales does not solve the general funding problem. While temporarily banning new short sales of financial institutions can be justified if there is risk of predatory trading, this is rarely a good idea since short sellers are often simply scapegoats when bad firms go down fighting. (See here for how shortselling works.)Tobin taxes are a bad idea
Second, a transaction tax on stocks is problematic for several reasons, most importantly because it moves trading away from the official exchanges and into the derivatives world, thus increasing the systemic risk. One of the main arguments in favour of such a transaction tax is that it helps to prevent bubbles, but there is little or no empirical evidence to support this. For instance, in the UK there is a 0.5% tax on trading stocks and a higher tax on trading real estate (up to 4%), but the UK arguably had one of the larger housing bubbles. Further, with a depressed and vulnerable stock market, this does not appear to be the best time to introduce transactions taxes related to potential stock price bubbles in the far future.
To see the problem, consider what happened in the UK due to their transaction tax. The professional investors such as hedge funds found a way around the regulation by executing their trades using derivatives rather than trading stocks directly (while individual investors are unable to avoid the tax). Specifically, in the UK hedge funds typically trade via swaps with counterparties such as investment banks to avoid the transaction tax. There is little doubt that this would also happen in the US if such a tax was introduced here. This would increase counterparty dependencies, systemic risk, and worsen risk management spirals as discussed above.
Another serious problem with the tax is that it lowers liquidity in the marketplace as trading activity may move abroad, move into other markets, or disappear. On top of these distortions to the stability of the financial system, this tax may raise capital costs for Main Street firms because of higher liquidity risk in US financial markets. Indeed, buying US stocks will be less attractive to investors – domestically and internationally – if they must pay a tax to buy and if they anticipate reduced liquidity in the future when they need to sell.
This could make it harder for US corporations to raise capital. And, the importance of being able to raise capital is what this crisis is all about.Conclusion
Market liquidity risk is an important driver of security prices, risk management, and the speed of arbitrage. And the funding liquidity of banks and other intermediaries is an important driver of market liquidity risk. Liquidity crisis are evolve through liquidity spirals in which losses, increasing margins, tightened risk management, and increased volatility feed on each other. As this happens, traditional liquidity providers become demanders of liquidity, new capital arrives only slowly, and prices drop and rebound.References
Viral Acharya and Lasse Heje Pedersen (2005). "Asset Pricing with Liquidity Risk," Journal of Financial Economics, vol. 77, pp. 375-410.
Yakov Amihud and Haim Mendelson (1986). "Asset pricing and the bid-ask spread." Journal of Financial Economics 17, 223–249.
Yakov Amihud and Haim Mendelson, and Lasse Heje Pedersen (2005). "Liquidity and Asset Prices," Foundations and Trends in Finance, vol.1, no. 4, pp. 269-364.
Markus K. Brunnermeier and Lasse Heje Pedersen (2005). "Predatory Trading," The Journal of Finance, vol. 60, no. 4, pp. 1825-1863.
Markus K. Brunnermeier and Lasse Heje Pedersen. “Market Liquidity and Funding Liquidity,” June 2008.
Markus Brunnermeier, Stefan Nagel, and Lasse Heje Pedersen (2008). "Carry Trades and Currency Crashes," NBER Macroeconomics Annual, forthcoming.
Douglas Diamond and Philip Dybvig, 1983, "Bank Runs, Deposit Insurance, and Liquidity," Journal of Political Economy, 91(3), 401-419.
Darrell Duffie, Nicolae Garleanu, and Lasse Heje Pedersen, (2002). “Securities lending, shorting, and pricing,” Journal of Financial Economics 66 307–339.
Darrell Duffie, Nicolae Garleanu, and Lasse Heje Pedersen, (2005). “Over-the-counter markets,” Econometrica, Vol. 73, No. 6 1815–1847.
Darrell Duffie, Nicolae Garleanu, and Lasse Heje Pedersen (2007). "Valuation in Over-the-Counter Markets," The Review of Financial Studies, vol. 20, no. 5, pp 1865-1900.
Nicolae Garleanu, and Lasse Heje Pedersen (2007). "Liquidity and Risk Management," The American Economic Review, P&P, vol. 97, no. 2, pp. 193-197.
Mark Mitchell, Lasse Heje Pedersen, and Todd Pulvino (2007). "Slow Moving Capital," The American Economic Review, P&P, vol. 97, no. 2, pp. 215-220.
Owen A. Lamont, 2003. "Go down fighting: Short sellers vs. firms," working paper, University of Chicago.