When Is A Trend Not A Trend?

If you do not keep records of your trading and trials, spotting patterns and trends can be difficult, if not impossible. I've always tried to do so and the past three days have reminded me of its importance.

I'm investigating a new strategy that makes use of the new automation facility in Fairbot. I don't intend discussing the strategy itself but it has immediately flagged a phenomenon (I'm reluctant to call it a trend after three days) which I have often seen in the past. Indeed, I'm sure I've posted about it before.

Since starting my life on Betfair, I have frequently ruined a good day by burning much, if not all of my profit in the last hour of racing. Whether I was scalping or swing trading; dobbing or laying the field it was remarkable how often I'd be doing quite well in the first 2-2.5 hours of trading only for one or two races to come along late in the card where I'd throw it all away.

I never really found an explanation for it other than my own stupidity. I put it down to tiredness on my part. Having spent all afternoon concentrating on the markets, I just assumed my concentration had slipped and I'd made a stupid mistake. This was so frustrating that I took to stop trading come 16:30, or at least I'd drastically drop my stakes. This latest trial calls that into question since the work is being done automatically and, for the moment, I'm using pretend money.

The phenomenon is weird because it seems to be restricted to dinner time. Specifically, between 16:45 and 19:00. Let me give you some figures to illustrate my point:

Using stakes of £50 the strategy has covered 74 races, losing 10 of them for a hit rate of 86.5% and a profit of £447. This is an excellent result and I'm delighted with it. However, when I looked through the results over the past three days I found that 80% of those losing trades happened during that dinner time period. Weird!

That time period produced a loss on each of those days. If I stripped those results out, the strategy had 2 losers out of 53 for a hit rate of 96.2% and a profit of £669. That's a huge difference! If not executing this strategy on these races produces such a huge upside, would it be possible to purposefully reverse this strategy for these races to boost things further? A very interesting question. 

Now whether this is truly a trend, only time will tell. It is certainly worth keeping an eye on and I'm very interested to see how it pans out as I continue my trial. If a long term trend does prove to be present I feel I'll still be at a loss to provide an explanation. One thing though; because of the automated nature of what I'm doing, I'll be happier that it isn't my fault. I'd still like to know why.

Early days for this strategy but it is important that you record your performance, irrespective of what strategy or trading technique you are employing. If you've never done it before, it can be a bit of a drag, but you'll soon get used to it. The potential benefits can easily outweigh the effort spent.

No comments: