backtesting is stupid

[alfobs]

Basically I meant a system with some robust core (algorithm) which is not optimizable and which gives you some positive expectancy, in other words the quality ingredient. Then I'd be looking to enhance it with quantity gradient which can be optimized. So, in short, if you set not the best parameters you'd still get some profit in long run, maybe with fractional profit factor like 1.01 (well honestly in practice I would not go with 1.01 but maybe with 1.5). Then the goal of your optimizable parameters is to make the profit factor acceptable. If you made a poor choice with it or badly over-optimized it you'd still save you account. No matter how good a system could be I always expect it may fail at any time therefore a backup system should be at hands as well until you figure out how to fix the original system. That's what was mentioned here as observing, learning and adjusting in real time. Just my point of view derived from practice.

Quote from Arthur Deco:

Well, duh! Your 5 years of 15 minute is not that much more than my half-year of one minute. Thanks for your comments on my time period. But I am having trouble wrapping my old head around "if you even enter at least favorable parameters you'd still be positive in long run." In what I do I try to ruthlessly drive out "least favorable parameters" to reduce the number of losers relative to winners, even if that means a lower net profit. For example I have one system where I could net 20% more but I would gag at increasing losers by a factor of 30% (roughly four per win rather than three).

 

[Arthur Deco]

Thanks for explaining. I would never have guessed. Your "core" algorithm is indeed the holy grail, a basic concept so good that it has positive expectancy without optimization. The only ones I have ever seen were cured by looking at longer data histories. But I am still looking. I agree with you about backups. One should always have them. For women and systems.

 

[Trader666]

Earnings trades are difficult... there are top-line and bottom-line expectations and even if a company beats or falls short, guidance can totally override the numbers. And then there's the market's mood at the moment.

So you can lose even if you correctly forecast revenue and net. But you started off at a disadvantage because you "would rather wing it" when even the simplest analysis shows that the results of the three companies you relied on are useless for forecasting IBM's numbers.

Quote from jonp:

well as your indicators told you that earnings would not be good, i'm sure you made a ton of money shorting it.

on a serious note this stings for a few reasons. I had a profit which I did not take this afternoon and was one click away from selling above 130. also because I posted it here. also because my analysis was just wrong, i figured intel just beat, amd did well, oracle did well so it was ibm's time to make 5-6% after hours. anyway now I am stuck with this shit