Welcome to NexusFi: the best trading community on the planet, with over 200,000 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 -- discounts are available after registering.
-- Big Mike, Site Administrator
(If you already have an account, login at the top of the page)
When developing an automated strategy, in your opinion, how much historical tick data would you need in order to feel comfortable with the results? This particular strategy is an automated scalping strategy trading off of Euro tick data.
Can you help answer these questions from other members on NexusFi?
watching some documentary where they criticized LTCM for backtesting only 5 years of data. had they backtested a bit more they could get a "feel" of the 1987 black monday.
You need to also factor in the frequency of trades.
For my backtests, I look for the strategy to trade at least 3-5 times a day and I optimize over at least 2 years of tick data. It is still very easy to over curve fit to data, but strategies that only trade once a day or once a week, etc, are obviously much easier to curve fit.
I loved the video. I saw the movie that CNBC put out re: the Lehman collapse but I think the youtube video was even better. Anyway, the automated trading that I am doing places between 3 and 5 trades per day. That said, it sounds like I need to back test against 2 or more years of data.
They were backtesting on daily data. 5 years of data is equivalent to 1250 bars. This is insufficient anyway. 1 week of 5 min data of ES would be 1395 bars, which is about equivalent.
If the strategy makes 3 trades per day and you have two years of intraday data, then this would be roughly 1500 trades (750 per year), which is a good base. You can then divide the data into your backtest and the walk forward sample.
