hindsight by now I know, but sim it first next time. or had you already? if not I understand the sometimes idiotic need we as humans get to try something without testing first lol sorry about the L. you'll sort it out.
Index futures automation
- Thread starter hilmy83
- Start date
hindsight by now I know, but sim it first next time. or had you already? if not I understand the sometimes idiotic need we as humans get to try something without testing first lol sorry about the L. you'll sort it out.
Month ended negative. Testing overnight hours contributed to most of this downfall. I think after 450 trades in shows that it's a workable strategy, but it has some flaws.
Since December is a short month, I just want to spend going over these results and backtest few more scenarios. Mainly just narrowing down which markets I want to trade going forward and sessions.
I think ideally by January, I got all the tinkering done. As they say, back to the drawing board!
Since December is a short month, I just want to spend going over these results and backtest few more scenarios. Mainly just narrowing down which markets I want to trade going forward and sessions.
I think ideally by January, I got all the tinkering done. As they say, back to the drawing board!
sorry about the negative month. set those testing sessions on sim and keep working the ones you've proven to produce. you'll get it sorted. good luck.
For December, I'm going to skip messing with overnight hours. It's just not worth the slippage/commisions. MYM and M2K are basically dead during those times, too many gaps, moving sideways. So I will just trade the 5 core markets (mes, mnq, m2k, mym, mcl) just during US session.
I will also want to stress test a new price "filter" setting added to the algo. Breakout trading is susceptible to chops. Basically this filter "expands" the High/Low range to a fixed ATR of its starting range after a set losing trades in a row. My way adapting breakout in range markets.
Will reset the quantanalyzer stats from December forward so I get better look at how this performs vs last 3 months.
My updated algo is pretty small (just under 1000 lines of code). I actually have learned quite a bit about order management in MQL. I am going to actually spend some money to take classes to learn to program. I think it's a valuable skill to just be able to test something on the fly
I will also want to stress test a new price "filter" setting added to the algo. Breakout trading is susceptible to chops. Basically this filter "expands" the High/Low range to a fixed ATR of its starting range after a set losing trades in a row. My way adapting breakout in range markets.
Will reset the quantanalyzer stats from December forward so I get better look at how this performs vs last 3 months.
My updated algo is pretty small (just under 1000 lines of code). I actually have learned quite a bit about order management in MQL. I am going to actually spend some money to take classes to learn to program. I think it's a valuable skill to just be able to test something on the fly
Last edited:
love all of this. I can't tell you how many times over the years I've said "I should have just learned to code" lol. good for you.
After analyzing the performance of each instrument per hour, you can then set the trading permissible for each instrument.
My framework allows for a morning and afternoon session. For example: 9:30 start 11:00 stop, 13:00 start, 15:50 stop. This provides for enough flexibility IMHO.
When I let the breakout range increase per volatility as session goes, my probability for profit descreases and also increases time to profit. When I leave it, I get too many false breaks, but the TP hit rate is high. It's like you want to fix one issue and you end up dealing with another. You always have to compromise somewhere
Last edited:
Just looking the performance of individual symbols since I started. MNQ and MCL (lesser trades) have been very consistent compared to others. But not a lot of trades to go by since most of them just go straight to TP.
Maybe this strategy works better on these two markets or maybe it's just coincidence. I do wonder if I should just limit to oil and nasdaq.
MCL month to month contract makes it harder to look it at overall. Quantanalyzer allows only 1 or all market analysis.
I think I want to run forward test on just these two markets separately. See if the performance can hold.
Maybe this strategy works better on these two markets or maybe it's just coincidence. I do wonder if I should just limit to oil and nasdaq.
MCL month to month contract makes it harder to look it at overall. Quantanalyzer allows only 1 or all market analysis.
I think I want to run forward test on just these two markets separately. See if the performance can hold.
Whatever you are doing, keep doing it, because it is working. The market is not going to get any more volatile than it is right now in the future (unless we get a nuclear war), so you should be safe.