Basket Trading Using A Directed Acyclic Graph by Dave Cline
A Tight Clique
When trading a basket of securities, one or two nonconforming stocks will throw your system off track. Here’s one way to ensure that the components of your portfolio travel together.
To successfully trade a mean-reverting basket of securities means the participants must behave and revert to the mean. If you get a couple of nonconformist stocks in your group, your entire system can get thrown off its wheels. Confirming that you have a collection of securities that constantly move together is a must in building such a basket. In this article, I’ll demonstrate how to use a directed acyclic graph when selecting the components in your basket of securities to help ensure that your portfolio of instruments, in general, travel as one.
DIRECTED ACYCLIC GRAPH
A directed acyclic graph, or DAG, is like a network graph on a social site. If friend A and friend B are both friends with C, then chances are A and B might also be friends. A DAG represents the relationship map between nodes. You could think of it as the six degrees of separation of the equity world. In this case, the nodes of the graph are equities and the relationship metric I’ll use is correlation. My objective is to build a network of equities that are all as correlated as a group as you can get them. See Figure 1 for an example.