In my trading practice I am always looking for better ways to robustly manage risk, but complicated strategies for this do not get me excited. The more complicated, the more likely it will breakdown in the future. It is the sophisticatedly simple, elegant and easily repeatable trading and risk management practices that tend to attract me. I have come to believe trading your own equity curve as a level of risk management fits this description.
To determine if my hunch on this had merit, it had to be tested. Because I have run a particular short-term trading system as a little side test for several years (now being implemented as a new trading operation), I have a decent amount of data to use for further testing and tinkering. Every trade over the period of the multi-year experiment has been recorded down to the number of times stops were systematically moved by their algorithm. Therefore, I was able to go back, find one of the most challenging 12 month periods in the test, and see what would happen if I traded the equity curve for that period. By retesting this data with no changes other than the risk profile per trade, the behavior we are trying to analyze is isolated.
Lets define terms. By equity curve, I am talking about the equity of the account as it moves through time. Very similar to prices of stock and commodities on a time and price based chart. The equity curve is either going down, sideways, or up. There are no other options.
By trading the equity curve, I am talking about taking cues from what just happened to the equity curve in order to make a future decision on how much risk to use on the next trade. The opposite of trading the equity curve would be to keep risk at a static proportion to equity no matter what has recently happened in the trading operation.
I already had a static proportional risk profile as a baseline. The prior test had been run risking 5% of equity on every single trade signal. Now all that had to be done was to rerun the trades with an adaptive proportional risk profile.
Here is what I did: The first losing trade in a series resulted in proportional risk being reduced from 5% of current equity to 2% of current equity. The first winning trade in a series resulted in proportional risk being increased from 2% of current equity to 5% of current equity. Like I said this is simple. I don't like overly complex.
The premise was that a system based on technical price trending rules or heuristics will undergo periods that are better suited for the system to be profitable as well as periods that are not so well suited for profit. When a losing trade occurs, it is a signal that a drawdown period may be at hand. When a wining trade occurs, it is a signal that a profitable period might be afoot. These signals in the equity curve's trend tell us when it is best to increase or decrease risk exposure.
Again, this data is from a short term system. The average month in this twelve month period saw sixteen round trip trades. Therefore, there are ample realized gains and losses in the data. The equity curves you are about to see only represent realized gains and loses (closed trades). Although not really a direct factor in what we are measuring here, it should be noted that commissions were accounted for in each trade.
Here is a chart of the two trade by trade equity curves -adaptive and static:
The adaptive proportional risk profile was 69% more profitable in the end than was the static proportional risk profile.
It is clear that equity curve trading worked in this example. I speculate that it is robust and will work over other long periods of time as well. In fact, I will probably try to find time to continue testing this on various periods of data just to see. In fact, it makes me wonder if the gap between the static risk model and the adaptive risk model would continue to widen exponentially in the future. Perhaps it is time for me to learn EasyLanguage and try to find more automated ways to test these things.
It is my current plan to implement equity curve trading into the new short term operations I have recently embarked on.
Does anyone else out there have any experience with this? If so, I would love to hear from you.




