«

Feb 07

The Beauty of Back Testing

Fish-holidays-July-12-012

I’ve embraced my inner nerd this weekend, and spent a great deal of time with my face in front of the computer, writing out reams of data by hand.  You see, I had an idea to make my life a little easier and hopefully more profitable.

I’ve been mucking around with FX this year in an effort to diversify away from equities, and while I had a brilliantly profitable month in May, it turned to absolute junk in June.  The main issue with my method was that it relied too heavily on my own input to generate signals through the use of trend lines.

The thing with using a trend line as part of your trigger is that you can pretty much draw a trend line to create any outcome you desire.  Don’t want to enter?  Funny how that trend line you just drew neatly misses the mark.

Even if you don’t mean to, your subconscious can have its evil way and screw things up.

So I need an entry that is totally price derived, and not manipulatable(?) by me. On Thursday I noticed a set of circumstances that seemed to be an interesting trigger into a trade, so set about testing it to see how it panned out in a stop-and-reverse kind of scenario.

 

It was crap.  It lost money and had way too much noise.  But after a couple of filters were added it’s come off quite nicely. Thankfully – after manually testing every entry and exit on the $AUDUSD since 1998 I was starting to feel like my eyeballs would fall out.

The thing I like about doing a manual back-test is that you get to feel the pain.  Perhaps that’s a little sadistic, but I think it’s important to sit through the ups and downs your system would have generated.

You live through each entry and exit as it would have happened and when you hit a huge patch of losses you start you feel uncomfortable that it’s not going very well.  You start to doubt if the system will even be profitable after this stupid clump of losers.

With computerised back-testing programs, you don’t get that.  You get to see very painlessly whether your idea works, but you don’t have to sit through each and every loss with anything invested.  Even having your afternoon invested in the outcome is enough to make you want it to be successful, and when you’re recording it and it feels like it might not be, it’s really disappointing and feels like a huge waste of time.

But its not.  There’s absolute gold to be found in back testing, even if the system is junk.

The first thing is that at least you get to find out it’s crap before it loses you all your money.

And in the happy event it rakes in the dough hand over fist, you get to have a sneak peak into the personality of your system.

For example, within the 14 years of data I tested, there was a huge run of 14 losses.  And a break-even was in there too, so 15 trades without a win.  On a daily system this could translate into a totally win-free year – yuck!!

When I was testing, I could feel myself getting frustrated.  I worried about what kind of effect the draw-down would have on the systems profitability, and if I would even be able to sit tight and trade through that kind of situation.

You hear all the time that you never know what’s just around the corner, and this testing experience really brought that home to me.  You need to be prepared to trade through 15 trades from hell – well, purgatory perhaps, since we’re capping our losses – to get through to the other side.

Because just like you don’t know when a nasty patch is before you, you also don’t know what comes right after those losses.

The thing is, that run was an outlier.  Prior to that freak run, the largest hit of losses was just five – nearly anyone can handle that and if I’d quit after loss 10 and ditched the system I’d have missed out on what was to follow – namely, two monstrous money-making wins.

I still have a lot of work ahead of me with this method – I need to test it on some other pairs, check the correlations and I also want to see if the method translates nicely onto shorter timeframes.

But it’s work I’m happy to do, because every minute I spend now figuring this out will translate into a method I know inside out, and have enough confidence in to pull the trigger time and time again.

***

Did you know you can get Rogue Traderette posts sent direct to your in-box? Sign up for FREE here!

You can also follow me on Twitter Facebook., or StockTwits.  I know, I’m everywhere. ;)

And if you really liked the post, please spread it around a little – the share buttons are right below.

Leave a Reply