Renaud Beaupain

Inferring trading dynamics for an OTC market: The case of the euro area overnight money market

The overnight money market is in essence the segment of the money market in which financial institutions fill their short-term liquidity needs. In the euro area, the overnight segment nevertheless operates mostly over-the-counter (OTC), meaning that effective transactions are not recorded and therefore investigations of its trading dynamics usually rely on market proxies. Although the euro overnight index average (EONIA), disseminated daily by the European Banking Federation (EBF), conveys valuable information in this respect, policymakers and credit institutions monitoring this segment traditionally examine alternative sources of information providing a more detailed insight into its dynamics. In this respect, two competing sources of data are commonly encountered. A number of studies (see, e.g., Moschitz 2009, Alonso and Blanco 2005 or Idier and Nardelli 2008) makes use of the information conveyed by quotes posted on wire services (e.g., Reuters) by market participants when revealing their intention to trade. Other authors (Angelini 2000 and Barucci et al. 2004, among others), however, rely on records of the effective transactions that are executed through the e-MID electronic platform as a more accurate approximation of the real flow of exchanges of funds in the overnight money market. The common objective behind these studies is to construct a market proxy that accurately replicates the dynamics observed in the actual EONIA, which remains the main reference of the unsecured overnight segment of the money market in the euro area. Beyond the issue of using a superior proxy for the transaction prices in the overnight market, the accuracy of more detailed data sets is also of particular interest for policymakers and market participants. To monitor the liquidity conditions in the money market and to ensure a smooth evolution of the overnight rate, policymakers indeed constantly pay particular attention to the overnight market. By the same token, the accuracy of high-frequency data sets conveys valuable information for credit institutions to infer the daily realized overnight rate or the evolution of liquidity conditions in the market. This paper therefore examines whether, and to what extent, e-MID and Reuters data are representative of the

Liquidity Co-Movements, Market Capitalization, and Volatility

Liquidity co-movements are studied within three different market capitalization indices, each made up of 100 NYSE stocks. To condition the analysis of liquidity comovements upon index volatility, three regimes of volatility are defined using the Markov-switching methodology. Our results shows that volatility bears on liquidity co-movements in a subtle way. First, volatility affects liquidity co-movements among small caps differently than liquidity co-movements among mid and large caps. Second, spread-based liquidity co-movements react to volatility differently than quote-based liquidity co-movements.