Designing an adaptive ETF Model Portfolio in three systematic steps.

Step 1 - ETF Selection - How Twenty20 selects which ETFs to use from those 1,328 listed on the LSE

Cluster Analysis

At Twenty20 we use cluster analysis to break down the large set of ETFs at our disposal into smaller sub-groups, many of which turn out to correspond to well-known sub-asset classes.

For example, one of the clusters comprises many government bond ETFs as they possess very similar risk/return characteristics. To learn more about this subject refer to Cluster Analysis for Researchers.

Asset Diversification

If one takes the view that asset diversification is the only free lunch left on the table, then the hunt for uncorrelated assets is key to building an investment portfolio.

Machine Learning

In many other fields of application, such as machine learning, image analysis, and pattern recognition, various techniques from the field of cluster analysis are used to understand the relationships between large sets of data. With over 1,328 ETFs listed on the LSE, the financial markets provide another field of application.


Next: Step 2 of the Investment Process

< style > .mfp - iframe - holder.mfp - content{ line - height:0; width: 100 %; max - width:570px; min - height:725px; background - color:transparent; }