BioComp Profit 8 - Trading the ES

Thanks. This system is an end-of-day system, so it only takes inputs once a day. Of course, the Profit software doesn't care about that, and you could feed it bars twice a day -- or twice an hour if you wanted to. But when I set this system up, it was done on an end-of-day basis, and all the Profit models were generated based on that, so I can't change it now. However, your basic point is a good one, and if you wanted to spend more time in front of the computer (I don't), then I'm pretty sure you could trade more often and generate higher profits.

That's how I came up with the present values when I started. But maybe they could be further tuned using the trades since I started this thread.

 

It was pointed out to me that I made a minor math mistake in my 11/18 post. When I accounted for the -$254 loss on that date, I got the closed equity amount wrong by $30 (high). So the correct closed equity amount now should be $30 less, or +$4,617.

Thanks to you guys for keeping me honest.

 

As the results of the system I’m trading here have varied over time, I want to introduce a measure of system quality that will allow me to monitor the system’s performance and detect when it is not performing well.

The most obvious and easiest approach is just to monitor the amount of profit (in either points or dollars) being made per trade. However, that has a serious disadvantage, particularly these days. The problem is that as volatility increases, the amount of profit/trade will increase, and conversely as volatility drops. Therefore, changes in volatility can mask the results of model performance.

I believe a better solution is to use the percentage of perfect that the system is achieving. 100% percent of perfect is defined as the amount (in dollars or points, it doesn’t matter) that a system would make if it was long every day the market went up, and short every day the market went down. Because this system is trading at the close, the daily calculation of 100% perfect is measured as the absolute value of the change from yesterday’s close to today’s close. If a system was perfect, it would always generate that amount of profit.

But since real systems aren’t perfect, the profit they produce will be less. So to calculate percent of perfect achieved by a system, just take the profits it generated over a given period of time, and divide that by the amount that a perfect system would have generated. This measure then cancels out the effects of volatility, and allows us to see how the system is doing over time.

What kind of performance levels should we expect? In his book “Design, Testing, and Optimization of Trading Systems”, Robert Pardo talks about this same concept, which he calls “model efficiency”. (I actually like that term better, but will stick with “percent perfect”, as that is how the BioComp programs report this metric) According to Pardo, “Experience has shown that trading models with model efficiencies of 5% and better are good based on closing prices.” My experience is that 5% may have been good back in 1992 when Pardo wrote his book, but I think it is possible with modern modeling technologies to have reliable systems that produce over 10% perfect for long periods of time.

And that brings up another key concept, which I’ve discussed in earlier posts in this thread. It is quite possible to achieve much higher model efficiencies, if one is willing to tune the model to a shorter window of time. The tradeoff in doing this, of course, is that the effective life of the model is reduced. It’s somewhat like a highly tuned dragster, which will perform well in a short quarter-mile sprint, but which would fall apart trying to complete the 24 hours of Daytona.

Turning now to the system I’m trading, it definitely leans towards the dragster end of the performance spectrum. It has high efficiency, but probably a relatively short life. (just how long remains to be seen)

I built the model on 8/28/08. Since then it has performed at a level of 31.6% perfect, neglecting commissions and slippage, which I consider to be trader-dependent. This is a very high level of performance, unsustainable over the long run, but I’ve been happy to have it so far. But while the aggregate performance level is interesting, I’m really more interested in how the system is doing lately, because that will tell me if its nearing the end of its life. In order to give more visibility to recent performance, I’ve started using a sliding window that measures model performance in terms of percent perfect over the last six trades. There’s nothing magical about the number six, it just seemed to feel right, but I might change it as we go on.

Using that 6-trade sliding window, performance has measured as high as 76% of perfect (in October), and as low as -20% of perfect a couple of weeks ago. That last number included a couple of large losses, which have now dropped off outside the 6-trade window, so the current figure is 33.44%. That is a very nice number, and I will happily take that, or even a number in the twenties, and continue trading.

The harder question is what to do when you get a low or negative number. For one thing, I think it is a signal to reduce the number of contracts you are trading, and keep your position size down until the performance rises again. But it is important not to jump to premature conclusions about the system’s performance, because systems often hit rough patches and bounce back. (As I write, the current system seems to be a good example of that). Keep in mind that a system that has a win rate such as this one can expect to run into 5 consecutive losses while operating within its normal parameters. That would be very rare, but it can happen, and doesn’t signal that there’s anything wrong with the system.

With that explanation out of the way, I will add this new performance metric to the trade reports going forward, and it will be interesting to see what we can learn from it.

 

Excellent info. I wish Ninja would show this metric, but it can always be calculated by hand.

 

Excellent contribution ,Chanelops...thankyou.

 

Hope everyone had a nice Xmas, I did. The model went SHORT at the close on Xmas eve (Weds), so I sold this morning at the open, with an execution at 866.50, for a loss of $1,237.50 per contract.

So our status is as follows:

Open Position: S&P mini (ESH9) - SHORT @ 866.50

Closed equity to date: +$3,409.50 per contract.

W/L=10/18 or 56%

18 trades down, 12 to go.

The model efficiency for the last six trades is 26.56% of perfect. Here's one related item that I forgot to mention in my longish post on this subject earlier: just to make the computation burden easier, I'm going to report the model efficiency as calculated using the prices reported by the Profit 8 software, which doesn't show any slippage that happens in actually executing the trade. That may result in small differences when compared to a calculation that uses the actual trade prices, but I don't think that matters.

For example, this morning's trade is actually shown by the software as taking place at Wednesday's close, with a price of 865.00. So in this particular case, the model efficiency will look slightly worse than if I had used the actual price I executed the trade at, 866.50. In some other cases it may make the model look slightly better, but I think it will average out over the long run.

Of course, the cash profit/loss and closed equity amounts continue to reflect the real prices I traded at, including any slippage.

 

Was there any new signal between the holidays? The model just gave back half of the profits. I don't like this EOD only signals, maybe there should be one midday too....

 

No new signals, and you're right that we've definitely given up some profits. However, that's pretty much the nature of EOD swing trading. I don't like it anymore than anyone else, for sure, but it comes with the territory.

I tend to agree with your point about a midday signal. But let me stress strongly that there's nothing about Profit that tells you how frequently your bars should occur. As I posted earlier, you could have 2 bars a day if you want, or 1 every hour, or even more frequently, I guess. Conversely, you could have a slower-trading system that uses weekly bars, which some people might prefer.

I picked end-of-day for this thread because my objective here was to have a system that can be traded by just investing 10 minutes or so once a day. (I'm talking now about day-to-day trading, not the time it takes to develop the original model) Adding more bars would make it a different animal.

That said, I still think the idea has some merit, and it's on my list of items to look into. I have been thinking of making the bars volume-based, rather than time-based, but that is going to take some serious additional research.

 

Even with an EOD system you could modify the system to use some reasonable intraday stop loss to get out of the bad trades before the next EOD signal gets you out and reverses at a big loss.

Holding on to a losing position for multiple days is not a good long term strategy.

 

2 observations:

1. I wonder how the system would have performed in the same period with a beginning of day signal, instead of an EOD?

2. What if the 2 signals are opening and EOD and not a midday? One signal at the open and the EOD signal tells you if you still should stay in that position or switch...