Equity Curve Management Crucial to Long-Term Success
Nerdy Little Factoid: The Equity Curve is actually the 1st derivative of a trading strategy, and multiple derivatives may be taken from its derivation. This implies some interesting mathematical properties. The first derivative is often smoother than the strategy itself and is hence easier to trade with a simple trend following approach.
Relevant background links:
This is a great article by Michael Stokes of MarketSci that covers some of the issues involved in trading the equity curve. At the end of the article in the comment section there is a reponse that I present that is highly relevant to this discussion.
Note: Rob Hanna shows the deterioration in a historically successful strategy–an important example of why it is important to keep an eye on strategy performance. Even good adaptive strategies must build in these filters to avoid drawdowns or poor trading performance. Rob has been a major inspiration to my work and many others and most of the stuff he publishes is long on common sense and yet put together with rigorous analysis.
Note: Jeff Pietsch is known most for his street-smart market commentary, but every once in a while he publishes a great article that is quite thought-provoking. This is an older article but still a classic–and even though it does mention the DV2 (a little self-dealing!) the most important aspect is the notion of being flexible to rotate to different strategies.
Note: Here Woodshedder shows a unique way of tracking a trading system and its health via control charts. He also mentions the possibility of “buying on the dips.” Again another great way to keep an eye on systems and inspire a new rotational strategy method.
I will note that mean-reversion is present in many strategies and this is largely a function of the behavior of people. A good piece on the reasons underlying this mean-reversion effect was written by Jared Woodward of Condor Options here http://www.condoroptions.com/index.php/meta/some-perspective-on-performance/ . Thus there are actually two sides to this coin–1) you can ride a strategy equity curve as a trend follower or 2) hop on to a successful long-term strategy when it underperforms as a contrarian strategy. There are costs and benefits to both, and I might add that both are NOT mutually exclusive. The most important consideration to determine which approach is best is how many strategies you can potentially track at one time and your preference for return vs risk. If you can track dozens of strategies and prefer the highest returns, than a trend following approach is better simply because you can always rotate to what is working best at the time. Think of this as like buying momentum stocks with a trailing stop and continuously rotating to the better performers. The best strategies are even better than good momentum stocks because they tend to move up more smoothly and for longer periods of time–sometimes for years at time without pulling back very much. The danger here is that when strategies become overused they can quickly become victims of their own success . The lowest risk and most desirable strategies can absolutely collaspe if there is too much money chasing them that must simulaneously exit all at one time (ie portfolio insurance and the crash of 1987). Contrarian strategies ironically are more ideal if you are risk averse, and cannot track many strategies. In this case buying value with a built in stop loss –ie you will exit a strategy below its long term or intermediate term regression line– is the best risk-adjusted method from my own testing. Of course the variations are endless, and as indicated effective combination methods are well within reach.