Equity Curve Trading
What is Equity Curve Trading?
First, let’s start off by defining what an equity curve is. It can be simply described as the growth of capital over time. Below is an equity curve that shows how an account would have performed had it invested only one contract in the fourth Free Friday strategy.
(given away on twitter @dburgh).
A major fear in system trading is that often times your profit and loss or “equity curve” is your only feedback loop. That is, you don’t know when to stop trading a “broken” system until it has already cost you significant money.
One industry idea is to monitor the equity curve in comparison to a moving average of the equity curve. That is, if our equity curve falls below its rolling average of the equity curve then maybe it is a red flag or early warning sign the system is broken. On the other hand, if you are confident in your system then maybe this temporary sign of system weakness is actually an opportunity to trade larger!
Some systems might miss significant drawdowns by simply turning off once the equity curve has weakened below its moving average and turning back on once the equity curve reclaims the moving average of the equity curve. Additionally, some other systems may go through significant periods of run up after their equity curve has fallen below its moving average. The idea being a period of temporary weakness will soon be followed by some strategy out performance.
Either way system traders need to know how their system performs when the equity curve weakens. If not, we are only to blame for our losses especially if we can identify some loss saving techniques ahead of taking a system live.
Build Alpha offers simple views of all the possible scenarios regarding Equity Curve trading.
In the picture below you can see the thicker blue line represents our original backtest’s equity curve. This is how the strategy has performed ignoring all equity curve fluctuations. The nasty drawdown that concluded around trade number 740 is not a drawdown most could endure.
The baby blue line shows how the strategy performed only taking trades when the equity curve fell below its 5 period simple moving average. On the other hand, the yellow equity curve shows if you only took trades when the strategy’s equity curve was above its 5 period equity curve moving average.
You can see the baby blue returns are much smoother and avoided the most recent drawdown experienced around trade 740. In contrast, the yellow equity curve – only trading when the strategy was above its equity curve – caught all of the drawdown. In fact, the yellow equity curve had a worse drawdown than the original backtest.
Here is another example. Same idea, however. The blue and red lines show nice smooth returns whenever this strategy is below its 5 or 10 period equity curve moving average, respectively. On the flip side, the green and purple lines show how choppy and risky equity returns have been when only trading this strategy when above its 5 or 10 period equity curve moving average, respectively.
In conclusion, this can be valuable information that temporary weakness in this strategy has historically been followed by large drawdowns or by periods of out performance/strategy recovery.
This can give insights to the trader on when to increase position size/leverage or to just simply have confidence in periods of strategy weakness, for example.
The main takeaway being… often times smoother returns or a better risk-adjusted return can be achieved by taking into account the strategy’s “health” vis-a-vis the equity curve in relation to a moving average of the equity curve.
These type of ideas can be considered Meta Strategies. A Meta strategy is a trading system that trades trading systems. The idea is to develop a strategy that will improve an existing trading system. Ideally, these layered decisions will increase overall performance as opposed to just following the original strategy.
The simplest Meta Strategy is to only take a trade if the underlying or original trading system’s last trade was a loser. Obviously we can get more complex and say only trade if the original trading system’s last three trades were winners or, more probable, do NOT trade if the original trading system’s last three traders were winners.
To run all these tests checking for autocorrelation of returns would be cumbersome, but Build Alpha displays all the 2 and 3 trade Meta scenarios for each and every strategy. BA also shows equity curve trading results for every strategy.
A key feature of Build Alpha is the ability to import systems built outside Build Alpha and run all the tests on them. For example, you could easily import your system and check all these Meta system ideas across your system.
Thanks for reading,