The sub-prime crisis emerged in the United States in mid-2007 and spilled over to Europe and other economies. From mid-2007 to mid-2008, the spillovers were relatively modest. The situation began to change in mid-2008. Then, following the bankruptcy of Lehman Brothers in mid September 2008, developments took a dramatic turn leading to a global financial crisis. During the crisis, the US dollar has seen some remarkable swings against major currencies. For example, from September 2007 to March 2008, it depreciated about 16% against the euro and yen, while between March and September 2008, it gained sharply (22%) against the euro. On the other hand, the dollar depreciated against the yen about 21% from August to December 2008, in particular after the Lehman’s default (see Figure 2 below). During this crisis, banks reportedly faced severe liquidity problems. US dollar funding shortages put intense pressure on the balance sheet capacity of the banking sector due to financial sector deleveraging. In response central banks around the world took unprecedented policy measures to supply funds to the banks (see McGuire and von Peter, 2009).
The purpose of this paper is to investigate any link between the market wide liquidity risk caused by the deleveraging process and exchange rate movements during the crisis. Adrian and Shin (2008) document that aggregate liquidity can be understood as the rate of growth of the aggregate financial sector balance sheet. When asset prices increase, financial intermediaries’ balance sheets generally become stronger, and, without adjusting asset holdings, their leverage declines. The financial intermediaries then hold surplus capital which they use to expand their balance sheets. On the liability side, they take on more short term debt. On the asset side, they search for potential borrowers. Aggregate liquidity is intimately tied to how hard the financial intermediaries search for borrowers, including through the interbank market. Conversely, when asset prices decline during a financial crisis, the financial intermediaries’ balance sheets contract and are thus reluctant to lend. Aggregate liquidity then declines.
Spreads of interbank interest rate over overnight index swap (OIS) in the US as well as Europe, the UK and Japan widened substantially in August 2007, and then persisted at high levels during the financial crisis in 2007 - 2009. The rise in spreads could represent heightened perceived default risk or greater compensation demanded by risk averse investors against the risk of default. Alternatively, it could represent a risk premium demanded by investors to induce them to hold comparatively illiquid assets. Schwarz (2009) constructs new microstructure measures of credit and market liquidity and find that liquidity effects explain more than two-thirds of the widening of one and three-month euro LIBOR-OIS spreads. Taylor and Williams (2009) find that while counterparty risk is a key factor in the movements in the term lending spreads including LIBOR-OIS spreads, they do not rule out that liquidity has been reduced by the increase in counterparty risk since the crisis began. The argument is that banks are reluctant to lend funds in the interbank market because of uncertainty about their own future need for funds, perhaps because of concerns about risk in their own balance sheet.
According to the theory in Adrian and Shin (2008), when the asset prices declined during the crisis, banks were reluctant to lend in the interbank market. This in turn would reduce market liquidity and require a higher risk premium (i.e. higher aggregate price of risk) for lending with longer maturity (which is more illiquid). Their reluctance to lend to each other in money markets at longer maturity should also contribute to the rise in spreads between the term and overnight interbank lending. The LIBOR-OIS spread is therefore an appropriate measure of the market wide liquidity risk. Figure 1 shows the negative relationship between the leverage of US banks and the spread of three-month US dollar LIBOR over OIS during 2007 - 2008.
The abrupt escalation of the crisis during 2008 marked an important turning point for the exchange rates of many currencies. Traders, bankers, and economists often attribute these exchange rate movements to a decline in risk appetite (i.e. an increase in market wide risk premium of holding risky assets), in particular due to the unwinding of carry trades (i.e. decline in “carry-trade incentives”). As the market wide liquidity problem due to US dollar funding shortages also occurred during this period, the risk spreads of interbank interest rate over OIS should also explain these exchange rate movements, if the theory in Adrian and Shin (2008) is correct. Against this background, this paper investigates the contribution of the market wide liquidity risk on the one hand, and “carry-trade incentives” on the other, to the value of the US dollar against several currencies during the crisis.
A related paper by Hattori and Shin (2009) studies the conjunction of deteriorating credit conditions in the US and the weakness of the dollar against the yen in the early stages of the credit crisis of 2007 - 2008. They argue that the carry trade should be viewed in the broader context of global credit conditions. Both can be thus seen as consequences of financial sector deleveraging in the US. However, our paper makes a distinction between exchange-rate movements that are caused by market wide liquidity risk (itself a consequence of the deleveraging process) and those that were due to carry trades. We view the latter as being the result of changes in risk appetite of participants in the foreign exchange market. We include the exchange rate movements of the euro, British pound, Swiss Franc, Japanese yen, Australian dollar and New Zealand dollar against the US dollar in the study.
The remainder of this paper is organised as follows. The next section describes the measures of market wide liquidity risk and carry trade incentives. Section III discusses the data used and the model specification. Section IV presents the estimation results and discusses the related issues.
Section V concludes.
