90% winners - really?

So I saw this promotion recently where the promoter was touting a system that had 85% winners or something insane like that and I thought to myself - I can probably write one of those without too much trouble. I mean, if you don't have to limit yourself to real-world realities like execution slippage and stuff, sure, you can have a system with 85+% winners. So, here you have one on the daily ES chart with two rules.

1. Two days down - Buy ES on the close; Two days up - Short ES on the close.
2. Flip out the next open for 2 points.
3. Exit at end of day if not filled.

Over a 12+ year period, there were only 2 losing years with the model bringing in 60K on a single ES contract. Of course, no slippage, commission etc is included in this. And it looks like it may have deteriorated a bit over the last 3-4 years but is positive so far in 2012.

It's not a bad model - at least you know you have a 90% chance of a reaction up/down on the close on those days that meet the criteria but not something you can trade automatically and not something any beginner (the target of the AD I saw) will find any success with.

Setting the target to 3 points dropped the win rate to 83%. Setting the target to 4 points dropped it to 77%. Going for just 1 point gave you a win rate of a staggering 96%.

I wonder what I could sell this "system" for...ha.

If you want to play with the code, it's located here: https://www.structuredmarketseducation.com/MemberServices/Strategies/tabid/227/Default.aspx

 

Getting high percentage winners is easy. Trivial, even. Eg, go long 2% of your account everyday 90 minutes after open, with a 1% gain as the exit, and a -50% stop. You'll do quite a bit better than 90% win rate.

 

Since we trade the futures overnight as well, the system, while it may work, either leaves a lot on the table, or will get caught once in a while. Am I wrong? Not trying to be disparaging at all, just curious.

Do you have a plan for the actual reversals during the night?

Don

 

Don - are you talking to me or to Random Capital?

 

All responses welcome.

Don

 

The model I threw together above is on the 24 hour session. It works just as well on the day session only and some days give you nice wins with the gaps. But the model was only to show that you can get a high win rate and a nice $ profit but not be quite as good in the real-world as would be advertised. I think I was kinda on a small soapbox so getting off it now...

 

+1

All you have to do is skew the target / stop loss to change the win rate.

And of course, this skews the *size* of the profit / loss inversely.

Net net nothing.

 

That's ok. We have tried 100 different systems over the years, and have found that most tend to give back profits when some factors are left out. Not on purpose of course, just the fact that they may not have had the long term experience, or rely on "backtesting" for results.

In any event, just keep doing it when and if those exact scenarios come about, seems a bit limiting.

I used to play"fun Craps" with my Sister in law...we'd wait for 2 non field numbers and then bet the field. Of course there was absolutely no correlation, but it was fun.

All the best, and thanks for sponsoring on ET.

Don

 

A martingale system will give you a 100% success rate until it wipes you out.

 

There must be a reason you are giving away this system for free and are in the education business , no ?