If you have a fairly high volume trade with at least 10 years of data, you could also consider optimising for groups of FX symbols instead of all symbols. For example I will sometimes use high and low trend pairs, or high (eg Cable), medium (eg Euro) and low (eg EURGBP). This provides a reasonable balance between over-fitting by optimising individual pairs and very generic optimisation with all tradeable pairs at once.
Crystal clear. I like to run my optimized parameters on all the available symbols and compare the results with other strategies. If a generic system cannot work on other symbols with the same currency (at the very least), then it's a very clear sign of overfitting for me.
Thanks for this video! I've always wondered how to avoid curve fitting my EA's and have tried a few different techniques but now i have a better understanding on how to make more robust strats. Makes ne feel good knowing I just finished an algo that trades the "ross hook" candlestick pattern and works well in backtesting across multiple pairs and 1 CFD including EURUSD, GBPUSD, EURJPY, GBPJPY and XAUUSD
Hi, Thanks for your video your videos are very beneficial. When we are testing with multiple timeframes, should we use same parameters?For example, when we use 350-500 period range for 1m, should we use 6-10 period for 60m and 70-100 period for 5m or should we use the same period range with 1m(350-500)?
Compared to MT5, running multi-symbol and multi-timeframe tests in Zorro is far easier. Walk-forward optimisation and retraining is far more powerful too.
![[#2024MAMA] G-DRAGON - 무제(Untitled, 2014)+POWER+HOME SWEET HOME+뱅뱅뱅+FANTASTIC BABY | Mnet 241123 방송](http://i.ytimg.com/vi/Ox29z5Nf1Uk/mqdefault.jpg)
If you have a fairly high volume trade with at least 10 years of data, you could also consider optimising for groups of FX symbols instead of all symbols. For example I will sometimes use high and low trend pairs, or high (eg Cable), medium (eg Euro) and low (eg EURGBP). This provides a reasonable balance between over-fitting by optimising individual pairs and very generic optimisation with all tradeable pairs at once.
I do believe this is a wise approach.
Crystal clear.
I like to run my optimized parameters on all the available symbols and compare the results with other strategies. If a generic system cannot work on other symbols with the same currency (at the very least), then it's a very clear sign of overfitting for me.
Thanks for this video! I've always wondered how to avoid curve fitting my EA's and have tried a few different techniques but now i have a better understanding on how to make more robust strats. Makes ne feel good knowing I just finished an algo that trades the "ross hook" candlestick pattern and works well in backtesting across multiple pairs and 1 CFD including EURUSD, GBPUSD, EURJPY, GBPJPY and XAUUSD
Very nice videos Martyn. Waiting for the next ones.
Thanks for the feedback Leon
Thanks!
Thanks for the video and I agree with your analogy. Can you recommend a broker I could use with MT5 that works with Crypto/FX/US Stocks/Futures?
great value. crystal clear information.
Very interesting, but could you activate the subtitles of this video and the previous one please?, Thanks!
What is the next video for this series?
Hi, Thanks for your video your videos are very beneficial. When we are testing with multiple timeframes, should we use same parameters?For example, when we use 350-500 period range for 1m, should we use 6-10 period for 60m and 70-100 period for 5m or should we use the same period range with 1m(350-500)?
Compared to MT5, running multi-symbol and multi-timeframe tests in Zorro is far easier. Walk-forward optimisation and retraining is far more powerful too.
What is best, testing multiple uncorrelated symbols/instruments or very correlated multiple symbols/instruments?
Gems thanks 🙏🏾
How do you optimize multiple time frames simultaneously?
does Amibroker support this multi-symbol testing?
Wow.
🦉🦉🦉