Behavior of Stock Return, Calendar Effects and Adaptive Market Hypothesis (AMH): Evidence from Pakistan by Using Historic Data with Special Focus on Gregorian and Islamic Calendar

Abstract

Beyond the traditional verdict of efficiency/inefficiency of market, the current thesis aims to provide additional insights to portray the clear picture of behavior of stock market. Although the previous studies have broadly examined the classical EMH, but there exists no consensus among academicians and researchers whether markets are efficient or not. Earlier studies favor EMH and claim that it holds. However, the recent studies cast a serious doubt by showing a strong evidence against the validity of the EMH. In addition, numerous research studies used a predetermined sample-period to investigate the EMH while ignored the fact that the levels of market efficiency may change/evolve over time. Thus, testing efficiency in previous studies is confined to investigate whether the returns generating process of an asset is stochastic (market efficiency) or deterministic (market inefficiency). In this regard Lo (2004) tries to reconcile both the varying degree of market efficiency and EMH through Adaptive Market Hypothesis (AMH). This theory facilitates the efficiency of capital markets to evolve over time and allows market inefficiencies to co-exist along with market efficiency (EMH). The thesis aims to examine the behavior of stock returns and calendar effects through AMH to identify whether AMH delivers a better depiction of behavior stock return and calendar anomalies than traditional EMH. For the purpose 107 listed companies from Pakistan Stock Exchange (PSX) are selected over the period from 1996 to 2015 to contribute to limited literature available on emerging markets. Similar to the studies of (Urquhart, 2013; Urquhart & Hudson, 2013), five different types of behavior are studied. The data set is divided into 4 equal 5-yearly-subsamples to detect the evolution of market efficiency (through linear & non-linear tests) and calendar effects (through GARCH model & Kruskal-Wallis test). The simultaneous utilization of linear and non-linear methods would probably improve the prediction of market participants. The results of linear & nonlinear tests reveal that firms exhibit significant predictability of returns during certain sub-samples while insignificant in others indicating that returns go under periods of market efficiency and inefficiency thus, consistent with AMH. Similarly, the results of GARCH (1,1) and Kruskal-Wallis test show that behavior of calendar effects fluctuates and goes under periods of predictability and no predictability . Overall, results of study suggest that AMH well elucidates the behavior of stock return and calendar effects than conventional EMH. The findings of the study may helpful to investors to assess the time varying volatility to manage their portfolios, investment decisions and trading strategies at (PSX).