A Pragmatic Approach
Fundamental research is heavily focused on gathering and analysing first-hand information with the ultimate aim of determining whether a company’s earnings growth is under or over-estimated by the market. Analysts often spend a significant amount of time interviewing company management and performing industry research. As a result, opinions can depend on the working relationship between the analyst and management and can often be distorted by management optimism on company prospects. Quantitative analysis, on the other hand, tends to be more objective and scalable. When constructing an investment portfolio, it would be prudent for investors to use both forms of research which are complementary to each other.
||Disciplined; Informed Decisions
Below are the benefits and limitations of quantitative investing
Opportunity Set | Computer based models and screening processes can efficiently reduce a large universe of stocks into a shortlist that can be scrutinised and evaluated to identify the best reward/risk opportunities.
Behavioural Bias | Investors tend to act irrationally (through greed and fear) and an emotional bias can consistently hinder optimal returns. An objective quantitative strategy can exploit these behavioural inefficiencies.
Time Testing | Unlike traditional fundamental analysis, quantitative strategies can be empirically tested over long time periods and through different market cycles. With traditional bottom up analysis which relies on subjective forecasts and qualitative judgements, it is difficult to ascertain whether the investment performance is the result of manager skill, market cycles or simply luck.
Risk Control | Effective risk management is critical in the success of a long term portfolio. Hence it is important to understand the sources of portfolio risk and the manner in which it is compensated. For example, investors should expect to be rewarded for exposure to systematic, sector and style risks, whereas idiosyncratic risks can be eliminated through diversification.
Cost | Portfolios can be managed more efficiently through the use of quantitative driven models by reducing the need for expensive and time consuming qualitative research.
Data Biases | During the model development and back-testing process, survivorship bias can lead to overly optimistic beliefs because company failures have been excluded from the empirical analysis. Quantitative models may also be prone to "look ahead bias" which is created by the use of information or data that would not have been known or available during the historical period that is analysed. Another form of bias is "in-sample bias" which is common when one focuses only on a particular time period not fully reflective of all market conditions.
Data Mining | The vast quantity of information in the stock market allows a person, if they look hard enough, to find a relationship which appears to generate spectacular returns even though in practice the relationship may not work.
Model Complexity | Overfitting occurs when a model becomes too complicated e.g. having too many parameters relative to the data set. Usually these types of models have poor predictive powers because they can overreact to minor fluctuations in the inputs.
Data Reliability | Quant models rely on the accuracy of inputs but some of these such as analyst earnings estimates, are subject to high forecast error. In addition, some inputs may not be readily available and need to be estimated using interpolation techniques.
Note: This article is intended to provide general advice only, and has been prepared without taking account of your objectives, financial situation or needs, and therefore before acting on advice contained in this document you should consider its appropriateness having regard to your objectives, financial situation and needs. If any advice in this document relates to the acquisition or possible acquisition of a particular financial product, you should obtain a copy of and consider the Product Disclosure Statement for that product before making any decision.