Ebook Liquidity Risk, Credit Risk and the Overnight Interest Rate Spread: A Stochastic Volatility Modelling Approach
The interbank money market is the primary channel for the implementation of monetary policy for a number of central banks, including, for example, the European Central Bank (ECB) and the Bank of England. Steering overnight interest rates is crucial for these central banks as this provides an anchor for the term structure of interest rates. In the case of the euro area, the Euro Overnight Index Average (EONIA) is a weighted average of all overnight lending transactions between the most active credit institutions in the euro area’s money market. Effective steering of the overnight rate by the ECB would therefore imply a low spread between the ECB policy rate and the EONIA rate. Since the intensification of the October 2008, a very large spread became evident, however. This coincided with a range of liquidity easing measures by the ECB, leading to a large liquidity surplus across the Eurosystem. While a number of papers have examined the determinants of the EONIA spread in the pre-crisis period, there are very few (if any) that examine the period of the crisis. The purpose of this paper is to investigate this issue, using a Stochastic Volatility modelling approach. While the primary analysis is on the euro area, we will also carry out a comparative analysis for the UK, which also adopted enhanced liquidity-providing measures to counteract the lack of interbank market activity caused by the crisis.
In non-crisis times, excess volatility is not prevalent in the overnight interest rate as it tracks closely the main central bank policy rate, so that the spread between both is relatively low (i.e. less than five basis points). In crisis times, however, this is not necessarily the case, and in the recent crisis there has been a clear rise in both the level and volatility of the overnight interest rate spread. Clearly, in circumstances when volatility is higher, so too is uncertainty associated with the spread. During the recent crisis of 2007 to 2009, as liquidity dried up, a large policy spread was observed, particularly after the collapse of Lehman Brothers in mid-September 2008. This triggered an intensification of the crisis, and an expansion of central bank balance sheets as liquidity-providing measures were introduced. The result was a large liquidity surplus. In the case of the euro area, the EONIA rate fell below the minimum bid rate (MBR) in the MRO (main refinancing operations), as opposed to non-crisis times when EONIA normally trades above the MBR rate – see Figures A1 and A2 in the Appendix for more details.
Previous work that focuses on explaining the overnight interest rate spread is relatively limited, while to the knowledge of the authors there is existing study aiming to explain the spread in the crisis of 2007 to 2009. Regarding the EONIA spread, Nautz and Offermanns (2008) examined volatility transmission in the European money market over the period 2000 to 2006, specifically assessing the transmission of EONIA volatility to longer term money market rates. In estimating the time-varying dynamic of EONIA volatility, the conditional mean and volatility of the EONIA rate are estimated using an EGARCH model. Their analysis focuses on the period from March 2004 to August 2006. They find that the new framework has reduced the volatility in all money market rates. Interestingly they explain the fluctuation in the EONIA rate as being due to not only to the EONIA spread, but also to the term spread (defined as the spread between the 3-month Euribor rate and the minimum bid rate). These authors find that the latter, as an indicator of interest rate expectations, is an important determinant, even under the new framework. Bartolini and Prati (2005) assess the volatility of overnight interest rates for a range of countries, including the euro area. These authors also use an EGARCH model and focus on the results from the variance equation to identify the effect of monetary policy implementation across countries on interest rate volatility.
Similarly, Würtz (2003) adopts an EGARCH specification to model the volatility of the EONIA spread, finding that expectations on changes in the policy rates and the end of the maintenance period effects are the main drivers of the EONIA spread. The equation used for the policy spread is non-linear to reflect the fact that the EONIA rate is bounded by the standing facilities of the ECB Gaspar et al. (2008) present a model that examines the determinants of equilibrium in the market for daily funds. Using the EONIA panel database over the period 1999 to 2005, the model indicates that there is a rise in both the time series volatility and cross-section dispersion of the lending rates applied by commercial banks towards the end of the reserve maintenance period.
Sarno and Thornton (2003) apply an error-correction framework to US data. Specifically, the overnight rate is assessed by estimating error-correction equations for the Federal funds rate and the 3-month Treasury bill rate. They find that the adjustment of the Federal funds rate to the Treasury bill rate is asymmetric, namely that the effect is more pronounced when the Federal funds rate is below its equilibrium level. Similar effects in, respectively, a Japanese and European context have been found by Kuo and Enders (2004) and Clarida et al. (2006). Nautz and Offermanns (2007) examine how the EONIA rate adjusts to term interest spread, and how the policy rate of the ECB is affected by interest rate expectations and the monetary policy operational framework of the ECB. They find a strong role played by the tender arrangement. Specifically, the introduction of variable rate tenders with a minimum bid rate in June 2000 did not lead to a loss of control over the EONIA, and in fact, the link between EONIA and the policy rate appears to be even stronger when a positive spread is rising.
Posted in :
