Abstract:
This thesis examines the existence of anomalous returns and their risk modeling from
various perspectives in Pakistan Stock Exchange (PSX). Anomalous returns are
empirical results that are unexplained by the theory of “Efficient Market Hypothesis”
(EMH). EMH states that returns are a linear function of risk and it is impossible
to earn superior returns without taking extra risk. However, empirical evidence
shows that there exist numerous anomalies and risk-return models failed to provide
a good description of returns.
To study the risk modeling of returns, first, this study focuses on the role of liq
uidity in explaining the anomalous returns in PSX. In emerging markets such as
Pakistan, liquidity is considered as an important risk factor and returns are assumed
to be associated with liquidly. However, traditional risk-return model(s) such as
“Capital Asset Pricing Model” (CAPM) do not explicitly incorporates the liquidity
factor. To test wether liquidity is priced in PSX, portfolios are constructed as test
assets by using size and volatility related information. It is identified that portfolios
constructed based on size and volatility generate 30% to 50% annualized returns in
PSX during the period from 1994 until 2015. By measuring liquidity as the average
of zero return days in a month and then using the “Liquidity Augmented Capital
Asset Pricing Model” (LCAPM), findings reveal that these enormous returns are
not violation of EMH. Instead, a reward to investors for bearing the market and
illiquidity risk.
Secondly, emerging markets are considered to be weakly integrated with the global
markets. Hence, it is assumed that there exists better investment opportunities for
international investors in emerging markets. Therefore, in this study it is explored
that whether PSX provides risk adjusted opportunities to international investors
after controlling for the global risk factors. Results show that PSX provides risk
adjusted returns to international investors.
Thirdly, momentum anomaly is considered to prevail across the global markets.However, in emerging markets such as Pakistan, on average this anomaly has histori
cally low returns due to the high volatile nature of such markets. Therefore, this
study explores whether volatility is linked with the poor performance of momentum
strategy. To test this, the momentum strategy is adjusted for volatility and then
compares the performance of traditional momentum strategy with volatility/variance
scaled momentum strategies during the period from 1994 until 2016. By using simple
descriptive analyses, results show that the scaled momentum strategies outperform
traditional strategy in terms of higher raw returns and the Sharpe Ratios (SR). In
addition, rationalizing the returns of traditional and scaled momentum strategies in
the framework of standard asset pricing models reveal that the scaled strategies pro
duce larger risk adjusted returns than the traditional momentum. Furthermore, the
probability of negative returns for scaled momentum strategies reduce in comparison
to traditional momentum strategy.
Lastly, this thesis examines the ability of investor sentiment to predict conditional
volatility and excess returns at aggregate market and industry level in PSX by
using daily data from 2001 until 2015. The results show that sentiment induced
investors overreact to information which results in excess demand. As a result,
investor sentiment predicts lower future expected returns. However, this miss pricing
corrects in the next period which brings the sentiment induced prices towards the
equilibrium level. It has also been confirmed that bullish (bearish) sentiment increases
(decreases) volatility which in-turn affect the mean variance relationship. However,
the commonality of the effect of investor sentiment via conditional volatility has not
been uniform across industries.