A number of readers are using the scalper EA in live accounts. The number one issue that many of them cited is that my research focused solely on the EURUSD. Does it work on other forex pairs?
Absolutely. However, it doesn’t work on all of them. It’s important to follow the same logical process that explained why the expert advisor works so well on the EURUSD.
Analyze the scalper EA in Excel charts
We must dive back into Excel to evaluate the original hypothesis. My expectation was that the strategy should work on charts where the distance of the price from the 200 SMA forms a nice inflection midway through the curve.
The area right around the 0.5% marks the inflection point. As a reminder, you can think of the curve as being composed of two parts. There’s the steep part, which is where the price is highly likely move. Then there is the flat part. That means the price drifts instead of moves.
Think of slope as rate of change. A steep slope means a fast rate of change. The price is likely to be anywhere but here on the next bar.
Flat slopes make for slow rates of change. The price is in fact very likely to remain a similar distance from the SMA in future bars.
The strategy only works when price is likely to stay in the same spot. We are, after all, scalping. The opportunity only exists when the expert advisor can trade in the chop. The chop only exists when the slope of the frequency line is flat.
I used my experience on the EURUSD to infer that 0.75% would make for a natural starting point to evaluate for the moving average envelope. It’s far away enough from the inflection point to overcome spread costs, but close enough to yield a solid number of trading opportunities.
The initial results came out even better than the EURUSD. These results do not include slippage, commissions or spread costs.
The results are very much in line with the original idea. Percent accuracy stayed in the same ballpark, coming out to 81%. The profit factor jumped very nicely to 2.99, which is substantially better than the EURUSD performance of 2.16. The sample size consists of 113 trades, which is enough to infer a reasonable expectation of performance.
The final test is “does it make money when including trading costs?” The answer is yes. On a 2.5 pip spread, the total trading costs of standard lots on 113 trades is $25/lot * 113 lots (trades) = $2,825. That number is substantially less than the raw profit of $5,360. It makes sense to trade this strategy.
The final step of walking forward unfortunately doesn’t offer enough data points to draw a conclusion. It only placed 13 trades for the entire year. It broke even.
USDCAD scalping stats
USDJPY is a bad idea
The frequency graph for the USDJPY looks much, much different than the other currencies. Instead of being steep and mostly flat, it’s more like free falling and perfectly flat. The massive size of the tail and the severe contrast between the steep and flat portions led me to believe, correctly, that trading USDJPY would not be a good idea.
Although the areas near the inflection point are indeed the most profitable, the profit factor for USDJPY plummets to slightly above 1.0. When trading costs are factored in, it doesn’t make sense to trade.
Have you read the article explaining how and why the scalper EA works?
If you have any suggestions on how to make the rules apply to more currency pairs or instruments, then please share in the comments section below.