Reviewed by Sep 30, 2020| Updated on
It is a method used to group sets of objects that share similar properties. Investors use cluster analysis to establish an approach around cluster trading that will help them build a diversified portfolio. Stocks that show high return correlations fall into one bucket, those that are less correlated will fall in another, and so on until every stock is placed into some group.
The different clusters will show minimal influence from each other if done correctly. Investors, thus, reap all the benefits of diversification: decreased downside risks, capital preservation, and the ability to make more risky trades without adding to the overall risk.
A dominant drawback to cluster analysis is the degree of overlap amongst clusters. Clusters that are closer in the distance, implying a high return connection, also share similar risk factors. Hence, a down day in one cluster could translate into an equally weak result in another cluster. For this reason, investors should consider stocks with a large distance between them and cluster them.
The methodology may also reveal other stock groups, such as cyclical stocks and growth stocks. Such particular approaches come under the umbrella of smart beta or investment factor. They try to capture better risk-adjusted returns from particular risk premiums such as low uncertainty, growth, and momentum.
Smart beta or factor investment in some way embodies the grouping and categorization principles taught by cluster analysis. The clustering theory on a single common activity mirrors the underlying approach behind factor investment, which recognizes stocks that are vulnerable to similar systemic risks and share similar characteristics.
It is not always the case that assets reside in the same industry within a cluster. Clusters also contain inventories from multiple industries, such as technology and finance.