Backtesting Results: $EURUSD Weekly Outside Bars

In a previous post, I took a look at the weekly chart of the EURUSD and identified all of the Outside Bars, from 2001 to 2015.

I just have to make a quick correction to that post…I called the chart formation Engulfing Bars, when they are really Outside Bars.

Outside Bars only take into account the high and low of the candles, while the body of an Engulfing Bar must engulf the body of the previous candle AND have a higher high and a lower low. Also remember that Backtesting data is valuable, but it does have limitations.

Now that we have that out of the way, let's get to my Backtesting.

First, we need a method to test…

The Trading System

In my previous post, I only looked at Outside Bars that were in breakout and continuation (trending) situations. Based on a purely visual review, my hypothesis was that only trading this pattern in these situations would make it profitable.

I put in an order to buy or sell on a break of the Outside Bar in the direction of the trade. The stop loss was above/below the Outside bar. Once the trade was up 100 pips, I moved the trade to +5 pips.

Then I moved the stop up when a support/resistance level formed. I risked 2% on each trade. Here is an example trade on the chart.

Then I would let the trade get stopped out. That should work, right?

Let's see…

Results of Backtesting

The first test yielded a 356% return over the entire testing period of 170 months. This means it would have averaged 2.09% per month, over that period. Pretty great right?

But hold on…

As a result of working on the previous blog post, I got to know the EURUSD weekly chart really well. Was I doing well because I already knew what the charts looked like?

I decided to wait a few days, before doing another test. The second test yielded 142% over the same time period.

…yes, I was influenced by my previous work on the charts. But the results were promising.

So I tried the trading strategy on other currency pairs. The results were not as good.

They ranged from +20% to being overall net losses.  

The Debrief

Does that mean that Outside Bars are useless? No, not at all.

They can be powerful trading signals. But they need to be used in the right context. In this case, the EURUSD trends well, so a strategy that moves the stop up with each new level, looks like a good strategy.

But other pairs are much choppier, so using this method would not work well on those pairs.

Another thing to consider is that I may be trying to go for too much on each trade. What if I stuck to a 1R or 2R profit target?

That is a test for next time…

Bonus Backtesting Results

As I mentioned above, I also tested this method on other pairs. I filmed one of my sessions to help me see what I miss when I backtest. Here is my test on the GBPJPY.

There are a couple of things to take note of in this video. Sometimes I noticed that there was a look-back bias when I went too far and had to backtrack.

Since I already kind of knew what was going to happen, the results of the test were affected. But it didn't affect things too much because the results were still negative.

Conclusion

So this is a great example of a trading method that doesn't work. I think it is important to look at what doesn't work, as much as what does work.

There are so many sites out there that show fantastic gains, but they also hide the equally fantastic losses.

Remember that this doesn't mean that Outside Bars aren't useful in trading. They can be, it's just that this particular method wasn't profitable. Stay tuned for more tests, as I look for something that does work.