The Beta intervalling effect during a deep economic crisis : evidence from Greece

Purpose - The intervalling effect bias of beta refers to the sensitivity of beta estimation with respect to the reference time interval on which returns are measured and its manifestation may indicate the degree of market inefficiencies. The purpose of this paper is to study the intervalling effect bias within an environment and during a sample period that embraces the evolution of a deep economic crisis and show in particular that its intensity is profoundly magnified. Design/methodology/approach - The Athens Stock Exchange is studied via the market model during the sample period 2007-2012 that embraces the Greek debt restructuring. Two portfolios are formed to distinguish between large and small market capitalizations, three reference intervals are considered for measurement of returns (daily, weekly, monthly) and the respective betas are calculated via OLS simple regression. The results are compared to similar studies. The results are further confirmed by using a second proxy for the market portfolio. Findings - The intensity of the intervalling effect bias was very pronounced during this sample period with regard to all aspects of the phenomenon that similar studies have reported and to which the results of this paper are compared. Originality/value - This is the first time that the intervalling effect is examined in conjunction to a deep economic crisis environment. The intensity of the intervalling effect reflects the depth of the inefficiencies of a market for some period. As a consequence, some function measuring this intensity may be devised to serve as a measure of market inefficiencies.