I’ve spent the last week and a bit doing more research and studying. I need more strategies to supplement the two simple strategies I have running now – the continuation and the panic/retracement strategy.
It is said the market can do one of three things, Go Up, Go Down, or Go Sideways. Currently, I don’t have a strategy to really take advantage of any of those.
I believe to consistently make money in FX I need to considering the following strategies
- A trend following strategy for when the market is trending
- A mean-reversion strategy for when the market is in consolidation
- A scalping strategy that is designed to move back to the average, and should work in either trending or non-trending markets
- Event based strategies
- Market Panic and Retracement (already running)
- Big Move and Continuation (already running)
- Sessional Patterns (taking advantage of patterns related to different sessions)
- Seasonal Patterns (eg. Significant holidays, options expiry, IMM etc)
- Correlation based strategies
I did a little bit of work on sessional patterns producing the Omar_Sessions indicator but couldn’t see any obvious patterns.
I did a little bit of work on a reversion strategy producing the RangeFinder but couldn’t find a consistent way to identify a ranging period. Given the fractal nature of the charts, there are always ranges within larger trends and vice-versa.
After this, identifying trends seems to be easier so I have done some work on this and produced the TrendDirection indicator. This works by counting the number of bars in the past n bars that are above or below a configured moving average. This sum is divided by n to produce a ratio from -1 to 1 to indicate trend strength. It seems to be pretty good actually, perhaps I’ll use this to define a sideways market. One possible improvement would be to take the rate-of-change of this line as often the trend has already changed as the values move from one extreme to another.
From this, I produced a strategy that tries to follow the trend. It uses TrendDirection (H4, 50, 50, Simple) and a configurable threshold of the ratio (0.8) to determine if the market is trending or not. To time entries, it uses a faster MACD (m30, 5,15,4). So far the results are promising but I think there could be some improvements made in entry time. Also, I need to detect the recent swing peak to place better stops, and therefore also limits.