Welcome to NexusFi: the best trading community on the planet, with over 150,000 members Sign Up Now for Free
Genuine reviews from real traders, not fake reviews from stealth vendors
Quality education from leading professional traders
We are a friendly, helpful, and positive community
We do not tolerate rude behavior, trolling, or vendors advertising in posts
We are here to help, just let us know what you need
You'll need to register in order to view the content of the threads and start contributing to our community. It's free for basic access, or support us by becoming an Elite Member -- see if you qualify for a discount below.
-- Big Mike, Site Administrator
(If you already have an account, login at the top of the page)
I need a suggestion as what to do here regarding my strategy.
On one day, one setting works, the next day, the opposite settings work. I am talking about Fast and Slow parameters.
So to avoid 'gambling' the values, I put both strategies (one opposite of each other) at the open to test which one turns green (because obviously the other one turns red) and keep the green one running because that's how the market algos are for that particular day.
Anyone else can have any other recommendation regarding this? I don't have to change any values because they work fine, but the only problem I have is to determine which values can work for any selected day.
Without more specifics, I can only offer some general advice. I would try to see if I could find a correlating factor that makes one better than the other. Maybe the day's cumulative data or the trend direction and strength. Maybe it's volatility (ATR). Are you getting stopped out or are you picking the wrong direction?
I wrote a little strategy that was based simply on the recognition of a pattern. My hypothesis was, when this pattern occurs, go long.... it produced unexpectedly bad results when back testing. Just for kicks I reversed the order, the opposite way I would have intuited. That returned an $8000 profit for the same one day trading period. I ran it back two weeks. The results were still good, but not proportional to that first day. Then, I noticed that some of the patterns weren't being executed properly... like at all! Further investigation revealed that my algo had an error in it that was executing somewhat random orders. Once I fixed the error, I improved my consistency, but it was bad EVERY day now. LOL. My point is that just because an algo is successful over the course of your data pool doesn't mean that it's based on any kind of causal event. The fact that you have two almost identical algos with just a few parameter changes leads me to believe that your algo may not be based on sound causal events?
That being said, some algos do better during up trend days. Some do better during down trends. Some do better during consolidating markets. There is absolutely no reason to not include them all in a single algo that only trades one of the strategies after having identified which market you're currently in. Just be careful... whatever the markets personality, it will change at some point (and may be inconsistent throughout the day) so I like to include some code that halts executing orders if I've had some number of consecutive losses or I'm down a certain amount. Then, it continues to monitor the market to see if there's a change (ex. trending up is now consolidating) before re-entering orders.
Different days are different regimes.
A day can be trending ( going from A to B )
or a day can range bound.
In a trending day a swing algorithm will work fine, once you are in the trade, it will run enough to make a profit.
In a range bound day a mean reversion algorithm will be more successful.
It is perfectly normal that not all strategies work on all days.
the trick is to recognize the type of trading day that is coming or that is evolving into
and then drive the right strategy for that type of market.