Vladimir Patras
Master of Business Administration Dissertation
Nottingham Business School of Nottingham Trent University
May 2008
Abstract
Efficient Markets Hypothesis has been recently more often challenged on base of the empirical evidence which was suggested not be consistent with the theory such as excess volatility, market seasonalities, autocorrelation and predictability of equity returns. There is still ongoing discussion on relevancy and interpretation of these phenomenons and its consequences in investment strategy. This work link empirical data gathered from equity markets to equity markets theory and investment strategy framework. Importance of the work lies in providing guidance to investors on capital allocation on equity markets.
Utility of three investment strategies was evaluated in this work which were buy&hold (index), contrarian and momentum. Research is focused on research of mutual funds and custom portfolios in regard to their returns and investment strategy. In mutual funds research the funds were categorized on base of their investment strategy and performance parameters were evaluated to generalize which strategies provided the best results. In research of stock returns three portfolios were constructed and returns analyzed in search for mean reversion and autocorrelation patterrns. Using automation of some calculations provided by VBA programming language enabled processing of large data sets for more than 70 stock with daily trading data mostly back to 1990.
Research output suggests superiority of buy&hold strategy which is linked to Effcient Market Hypothesis. Buy&hold was shown to produce best risk-adjusted returns. This apply to both mutual funds and stock portfolios returns. Some less common patterns were observed in stock portfolios and possible explanation suggested within Efficient Markets Hypothesis framework.
Introduction
Since establishment of the first modern stock exchange in Amsterdam in 1602 investors and speculators have formulated various investment strategies aimed to maximize investment returns. While many of them had not been proven to be workable failures and successes led to increase of understanding of how equity markets operate. Effort of traders had later been joined by academics and led to establishment of new disciplines within financial economics such as portfolio, risk management. Since then stock markets has emerged as leading institutions mediating capital allocation and risen to the most prominent place in advanced economies.
In financial economics there is ongoing discussion on efficiency of equity markets. This issue has principle importance in economic science as it relates to rational expectations theory which is one of backbones of modern economy science. Nonetheless are practical implications as market efficiency form has direct consequences in acceptance of a investment strategy.
Today still mainstream view explaining equity markets behavior is Efficient Markets Hypothesis (EMH) contributed by many authors and rising to prominence in 1960s. EMH states that equity markets operate with high degree of efficiency. At each point of time all securities of the same risk are priced to offer the same expected rate of return. This imply that under high market efficiency it is not possible to earn risk unadjusted returns. Maximal possible returns are defined by risk and this relationship constitutes efficient frontier suggested by Markowitz in his Portfolio Theory. EMH is backed by sophisticated mathematical models and great amount of evidence from equity markets trading.
However, some empirical evidence is suggested not be consistent with efficient markets. On the other hand also most of EMH authors admit that absolutely efficient markets can exist only in academic theory. On this bases EMH has been challenged by theories that deal with EMH weaknesses. One of most prominent recently emerging financial school is Behavioral Finance that compiles evidence of both economics and human psychology to explain some phenomenons that it suggests cannot be explained by EMH.
These theories are closely related to investment strategies as they deal with risk-return relationship and predictability of returns what are primary points of focus in investing. Most common investment strategies applied in nowadays portfolio management – buy&hold, momentum and contrarian have justification in the mentioned theories. Buy&hold (and its index modification) is relatively low risk approach most consistent with EMH and often claimed to provide the best risk adjusted results. Momentum strategy is approach the most often practiced by current portfolio managers. There are some indications that momentum provides higher returns but at expense of much higher risk. Contrarian strategy is especially popular among some Behavioral Finance proponents although some authors suggested that it may be consistent with EMH as well.
This theoretical framework and the three investment investment strategies are used as theme of this work. Its aims are practically oriented with outcomes applicable in portfolio management. Using more analyses and research cases utility of the investment strategies are evaluated on basis of their risk adjusted returns. The work focus on comparison of index investing and mean reversion research as mean reversion is unifying idea of momentum and contrarian strategies. Empirical data are analyzed and plotted against trends expected on base equity market theories or corresponding investment strategies. Specific target or research question of this work is to answer which investment strategy is best fitting given market circumstances. This work does not have ambition to evaluate overall efficiency of equity markets although empirical evidence will always play important role in formulating theories on equity markets functioning.
Analyses which were searching evidence of utility of the investment strategies were evaluation of mutual funds performance, analysis of mutual funds on base of their investment strategy, construction and evaluation of portfolios expected to produce mean reversion patterns in their returns. Returns analysis is coupled with risk analysis where appropriate.
In analyses of mutual funds performance sample consisting of 63 mutual funds with substantial assets was chosen. Their results were categorized and statistically evaluated. Valuation and size were bases for categorization using Morningstar methodology. Funds of the sample which do not fully fit given categories were further evaluated on base of their value and portfolio measures. Data tracking up to 10 years was used where available. In line with EMH returns should follow primarily by investment risk. There is also focus on index funds and their assumed ability to over perform actively managed funds in long term in the analysis. More of observed measures are correlated one against other in search for returns determinants.
The second scope is evidence for investment strategies utility from stock portfolios. Tests are conducted on three portfolios They are derived from current composition of Dow Jones Industrial Average, Nasdaq 100 index and the third portfolio is not bounded to any index. Most calculations were performed using VBA (Visual Basic for Applications) programming language. Patterns that are assumed to point to mean reversion if present were analyzed and are summarized in presented tables. One year and two year periods were used for the calculations. Using more than one time frame is justified by claims of some authors that mean reversion presents differently depending on time series observed. Portfolios used track data back to 1990 were available. Time end point for the analyses is January 2008. Presence of mean reversion would favor momentum and/or contrarian strategies over buy&hold. Otherwise returns distribution corresponding to expected volatility is pointing to market efficiency and superiority of index investing.
Answers to research questions as they are possible to be generalized from the analyses outputs point to buy&hold as most efficient investment strategy. In this connection research limitations should be recognized. Data credibility depends on its sources – mostly Yahoo! Finance and Morningstar. However, these providers are accepted as trustworthy by financial professionals. Data availability may be considered as restriction as incomplete portfolios from older time series (mostly in Nasdaq index) complicates determination of deciles structure of the portfolio. This is reason why sections where complete data are not present are not plotted one against another for comparison. Reason why data for some companies in Nasdaq 100 index are not available is their IPO after 1990. Also scope of the research is less complex compared to published works. Larger funds sample, more portfolios included would provide results with higher statistical power.
Restriction that is not within scope of the work methodology is based on assumption that overall market efficiency may change over time. It is generally accepted that stock markets operate now more efficiently then few decades ago. Market understanding, technologies facilitating market research and higher number of analysts are accounted for the change. As mean reversion is in some contexts related to market inefficiencies its effect may diminish in future as suggested in case of some other market anomalies. Therefore future utility of findings of this research are dependent also on future market circumstances.
No comments:
Post a Comment