How to Manage Your Trades Part 7: Managing Systematic Breakouts
“Trend followers trade at new highs and at new lows [if going short]” – Ed Seykota, Market Wizard
In this installment of our Trade Management series, we are going to analyze Systematic Breakouts, taking inspiration from Andrew Abraham’s book “The Trend Following Bible”. In reality this morphed into a broader investigation of systematic breakouts and relative performance of different lookback windows.
Once again, simple entry strategies continue to work in the markets, if complemented by solid trade management practices.
Abraham Breakout Rules
In his book, Andrew Abraham was quite vague about particular lookback periods he uses. What he did say, is that shorter-term lookbacks will offer more trades with more false-breakouts and more volatile equity curves but (if managed well) the expectancy will be higher due to the number of opportunities taken during a typical year. That’s the trade-off: higher ending equity curve with larger drawdowns.
- Buy the X day high and cover on the Y day low (for example, X = 20, Y = 10)
- MACD (not the histogram) must be above zero and increasing.
- Short the X day low and cover the Y day high.
- MACD must be below zero and decreasing.
with those, which improved the SQN. Note also, that the strategy was still positive without any entry filter.
SQN. The principle that trade management has a strong influence on the system performance is clear, again.
Investigation of Lookback Periods
This investigation should be influential for any traders that are still looking for “the best lookback period” or “the perfect fit”. Market wizards have said time & time again that there is no perfect fit! We’re going to demonstrate that with cold hard data.
The break-out and initial stop-loss look-backs were tested with the SMA(60) Slope filter and SuperTrend(2,7) trailing stop.
Break-out Look-back Stop-loss Look-back
10 days 5 days
20 days 10 days (the Turtle shorter values)
40 days 20 days
55 days 20 days (the Turtle longer values)
80 days 40 days
The longer look-back pair used by the Turtles gave the best results – the total return was as good as the others but the equity curve was smoother (better SQN). With look-backs shorter than this, the number of trades increased, the average R-multiple return reduced and the standard deviation increased. Generally, longer look-backs gave smaller standard deviation.
Up to now, the systematic models we have programmed suffer from the same issues:
- the entry/trigger alone does not produce acceptable/tradable results;
- the market type (filtering a trend) is the first ingredient that significantly changes the nature of the results;
- trade management can turn a winning system into a losing system if it does not compliment the principle of the model itself.
While we do have one or two mean-reversion strategies to test in the coming weeks, we seem to be finding a pattern. Diversification really is the only free lunch. There is no one perfect system and even very good systems will suffer from periods of poor performance. The only way to reduce the overall volatility of the portfolio is to have multiple systems working in sync, and our resident programmer Tony is doing an excellent job at rediscovering the principles that Robert Carver (and other systematic traders) has outlined.
One final consideration is this: what edge can a human trader have over a machine? Why should a retail trader trade manually?
The answer depends on the experience of the trader and the resources available. For the most part, retail traders are most likely better off to use low-frequency systems, and trade like Austin for example. By studying and working with one trading strategy, a human trader can learn to stack the odds in his favour and exclude evidently poor trades that the machine would take. Here’s what Tony said about the Abraham Pullback:
I’m trading Abraham’s Pullback model now, but with my own discretionary filter on the set-ups. In the process I’m tracking the set-ups I reject as though I had traded them, so later (after 30 trades) I can see if my discretionary input is adding value. Out of 8 set-ups presented so far:
- I’ve rejected three, all of which were fairly quick losses.
- Of the five I’ve accepted: 1 didn’t trigger before invalidating, 1 hasn’t triggered yet, 1 loss , 2 still in.
- Potential R:R before the next evident support/resistance.
- The slope of the Bollinger Band that’s ahead of the move. (I rejected one because there was no real momentum and the BBs were contracting.)
- My assessment of the price structure;
- Agreement or not of my fundamental bias.
A human trader can add value IF (and only if ) he has experience and understands why he is doing what he is doing. Without experience, or with limited time or capital, the trader may be better off with a systematic model or with some kind of autosignals account. At FXRenew we are working on both initiatives. Stay tuned.
About the Author
Justin is a Forex trader and Coach. He is co-owner of www.fxrenew.com, a provider of Forex signals from ex-bank and hedge fund traders (get a free trial), or get FREE access to the Advanced Forex Course for Smart Traders. If you like his writing you can subscribe to the newsletter for free.
The post How to Manage Your Trades Part 7: Managing Systematic Breakouts appeared first on FX Renew.