Connectedness of money market instruments: A time-varying vector autoregression approach

The heightened volatility of LIBOR rates relative to other money market funding rates following the 2012 manipulation scandal and the 2007/2008 global financial crisis led to financial regulators’ recommendation to transit to more robust rates. During this period, uncertainty and heightened credit risks led to the drying up of liquidity in underlying money markets, especially for the longer-dated money market instruments. The need to shift to alternative rates was reinforced during the covid-19 crisis in March 2020 when the LIBOR rates’ vulnerability to short-term liquidity, and therefore volatility was amplified. This paradigm shift has economic and financial consequences. While connectedness studies exist for various financial markets and/or instruments, studies on money markets are limited. This is despite the uniqueness of money markets. This study fills the gap by investigating the volatility connectedness of overnight index swaps, LIBOR rates, and foreign exchange swaps using the time-varying vector autoregressive model. Specifically, the study measures the extent and dynamic connectedness of three major currencies (EUR, GBP, and JPY) in three maturity categories (1-month, 3-month and 6-month), for the period 2007-2020. The findings show that the connectedness of instruments is time-varying, event dependent for these currencies, with a high integration during crisis periods. However, the integration reduces when markets are calm. Notably, the bi-directional volatility connectedness of instruments varies across currencies. This is not surprising considering the domestic institutional and monetary policy specificities affecting these currencies.