#1 Worst Mistake for Strategy Traders
Corona Del Mar, CA
Howdy Friend!
In today's modern world of fast computing (made even faster by cloud computing networks), the biggest problem facing strategic traders like you is...
...over-fitting the trading rules to the past!
I remember reading Way of the Turtle when it came out in 2007.
It actually had an interesting variation on the original turtle trading breakout rules.
Instead of buying on a new high, the author talked about getting into a position within a certain ratio BEFORE the new high.
For example, if a new 40-day high is equal to 0%, maybe you buy at a ratio of 20% from the high.
I ended up running a test on a basket of NASDAQ 100 stocks, stepping a bunch of parameters, then nervously walking away.
Would this strategy yield the riches I was craving?
The test took half a day as I recall.
When I looked at the finished results...I was gobsmacked.
The strategy was uber-successful: 70% historical growth rate, and a tiny max drawdown.
At the time, that was the best computerized test I had seen on a basket of stocks...it was like looking at the finest bottle of champagne.
I was ecstatic -- I had something amazing to look forward to when I got back from my tour of Italy.
When I go on longer vacations (this was a month), I choose not to trade.
Ahh...I'm missing Italy...especially during this lock down.
Anyhoo, I'll keep this message short.
When I got back to the States, I fired up my computer, updated quotes, ran my Quasi-Turtle strategy to see which stocks it said to buy...
...then my jaw dropped to the floor.
During that month I was gone, the strategy had performed the worst ever.
What I discovered was that due to that amount of variables I was optimizing, I was all but guaranteed to make a strategy that fit well to the past, but failed in the future.
More tests = more over-fitting.
Luckily I at least had a month of what's called "test" or "out of sample" data to show that the strategy was a steaming pile of you know what.
So the key is to verify with out of sample testing to solve the #1 worst mistake most strategy traders make.
But is there a way to reduce the odds of over-fitting in the first place?
Is there some sort of process?
Yes!
I invented a process called "Rotational Fitness" which has dramatically lowered the amount of over-fitting on the trading models invented by The Boss "SuperAi."
I talk about it in this free online training >>
In this webinar we're going to cover how to make money right now with these 3 strategies:
- A long/short “Pair Switching” strategy with a strange interest rate trigger for massive profits.
- A NASDAQ 100 stock trading strategy that uses the most bizarre non-price stop I’ve ever seen in my life.
- A high-speed swing trading strategy using what I call “Dropout Stocks.” It’s so deadly simple that I could hardly believe it’s so profitable.
- How artificial intelligence built all these strategies from scratch!
I'm also doing an in person Billy Mayes style demo of The Boss "SuperAi."
Last time it built a strategy with more than a 140% historical annual growth rate.
Come hang out with me LIVE!
I'll also be doing Q&A at the end, so don't miss out on your opportunity to ask me questions.
|
|
Responses