When is the best time to begin trading a portfolio of algorithmic strategies?

  1. After a big winning period or run-up.
  2. After a big losing period or drawdown.
  3. After a small winning or losing period.
  4. Anytime.

There is a case to be made for each of these scenarios. Let’s look at two perspectives: emotions and long-term performance to help develop a game plan for beginning to trade a strategy or portfolio of strategies.

From an emotional perspective, which would you regret more:

  1. Going into a big drawdown after starting to trade.
  2. Missing a good profitable period because you postponed trading.

Many people look for strategies or portfolios to trade by searching for the best recent performance. It feels good to start trading after a big winning period, maybe after the best compared to other strategies or portfolios.

However, based on experience, it is usually better to wait for a more typical month or quarter of performance—even to wait for a larger drawdown—rather than entering right after a big winning month or quarter.

This can be more difficult emotionally, getting in after 1-3 months of losses, but it tends to minimize the regret of going into a large or long drawdown soon after starting to trade.

From a long-term performance perspective, the practice of rebalancing a portfolio regularly tends to smooth out quarterly performance, which also tends to improve compounded annual returns.

Applying this principle to algorithmic trading, it is a good idea to cutback the position size, number of strategies, or leverage after a big month or quarter. Then, after a losing month or quarter, begin adding back position size, strategies or leverage.

An example

Here is an example for you to think about, based on how I currently consider whether to add a new automatically traded strategy or portfolio. Maybe this will give you some ideas to help in your own trading!

For me, I like to make these decisions monthly. I regret it more when I go into a long drawdown after beginning to trade; I don’t mind missing a good profitable period because I know another one is always around the corner.

On a monthly basis, I look at my portfolio as a whole, and see if it has had more than an average month’s loss.

  • If more than an average month’s loss in the portfolio, and
    • If the portfolio or strategy I’m considering has lost more than a typical losing month, then I wait for the next month’s results to decide.
  • If the portfolio has a monthly profit down to an average month’s loss, and
    • If the portfolio or strategy I’m considering has lost more than a typical losing month, then I start trading it at the beginning of the next month.
  • Either way,
    • If the algorithmic trading portfolio or strategy I am interested in trading has a typical winning or losing month, then I start trading it at the beginning of the next month.
    • If the portfolio or strategy I’m interested in has made two or more times what a typical winning month makes, then I wait for the next month’s results to decide.
    • I can always make a more conservative decision for a good reason that I document.

I review these rules once a year to see if they still fit my goals and objectives

– Jim