eThesis - School of Business electronic theses

PURPOSE OF THE STUDY

In this thesis I examine non-trading and trading period returns and their differences for globally traded equity index, interest rate and commodity futures. I decompose the continuous price pattern of futures into Night and Day returns, thus close-to-open and open-to-close returns. My first objective is to determine the differences between Night and Day returns. Second, I study whether seasonal return anomalies and patterns presented in previous literature hold also during non-trading periods, thus when markets are closed. The anomalies include a day-of-the-week and weekend effect, January effect and holiday effect.

DATA

A data sample includes 17 futures whose underlying assets present three different asset classes: equity indices, interest rates and commodities. Data is gathered from Bloomberg and price source for each future is their primary trading exchange. The sample consists of open and close prices for over 15 years from 1997 to 2012. In total, the time period includes nearly 4 000 trading days and in total 64 428 future day observations.

RESULTS

I find non-trading and trading period price patterns to differ significantly. I report negative Day returns for equity index and commodity futures during the sample period. According to the study the premiums are earned solely during the Night time when markets are closed. Additionally, the volatility of Night returns is lower than volatility of Day returns. The results thus imply that there is no reward of bearing higher risk in daytime volatility. In consistence, Day time returns for interest rate futures are positive and Night returns negative. My findings supports the results of previous literature which argue that non-trading period returns outperform day returns in equity markets in the past decade. In contribution to previous literature I broaden the results across asset classes and internationally.