Trading Versus Non-Trading Returns: Evidence from Russia and the U.K.

Overnight risk is of particular interest for many market participants including traders who provide liquidity to the market, but also to market participants with longer investment horizons who want to determine whether a given risk-return tradeoff can justify possible intermediate portfolio hedging transactions. Overnight risk may in particular play a highly significant role in emerging markets, given that information is incorporated into prices at a slower rate and liquidity may hinder a quick unwinding of portfolio positions. In this study we examine the trading day versus overnight non-trading returns for the Russian and the U.K. stock market. We use data for the FTSE100 index, which represents London as a mature European benchmark market. For Russia we use data from the Russian Trading System (RTS) Stock Exchange in Moscow, an emerging European exchange. Daily returns are converted to Euro during the period January 1, 1999 to December 31, 2000, a period during which the Russian market was subject to substantial political and economic policy uncertainty. Russia has played a central role in the opening of the eastern European economies since the start of the transition process in the late 1980s. Up to the end of the year 1999, Russia was under the political regime of President Jelzin. The subsequent regime under President Putin was established at the beginning of 2000 as a result of political developments that led to early presidential elections. The impacts of liberalization and the opening of the economy during the late Jelzin and the early Putin regime also had large impacts on the Moscow Exchange. Our empirical findings are useful in explaining the effects of transition from emerging to mature stock market status, and they shed light on how emerging markets can potentially improve the investment opportunities for foreign investors. Our empirical results for both markets indicate that expected returns per unit of time are higher during non-trading periods, which further suggests the existence of an overnight risk premium. This premium does not seem to be only attributable to volatility risk, however, because the risk of large price drops for both markets appears to be much higher overnight. The results for Russia in particular are in line with this observation, while being even more pronounced. The Russian emerging market offers relatively high average returns during non-trading hours. Also, the RTS index exhibits time series dependence such that overnight close-to-open returns positively correlate with the subsequent close-to-close trading day returns. Our examination of trading versus non-trading variance per hour ratios in turn shows that relatively more return variability is given during trading hours in the U.K. than in Russia. In other words, volatility stays relatively high overnight in Russia, while it tends to shrink in the mature U.K. market. This chapter adds to previous studies on returns and volatility during trading and non-trading hours. Examples are the seminal paper by French and Roll (1986) and more recent work such as Zdorovtsov (2003). Related market microstructure research on non-trading periods is by Cao, Ghysels, and Hatheway (2000). The remainder of the chapter is structured as follows. Section 2 introduces the dataset, defines the trading and non-trading returns used in the examination and discusses the relevant trading times. Section 3 outlines our empirical analysis and discusses the results. The conclusions are contained in Section 4.