The problem with Larry Williams' patterns—I think I've studied them all—is that they have little statistical significance, and also that this difference masks the true power a pattern can have, although it's an interesting technique to use if you need it for your goals. It's always interesting to study the master Larry.
Really good article btw, look forward to more work of yours. Im constantly building strategies. Can get some good ideas. I also noted to pick up "Long-term secrets to short-term trading". Haven't read that one. Thank you.
Great article! Now, just put all instruments in a portfolio, and you will get a good sample size, a nice equity curve, a surprisingly good Sharpe ratio, and low drawdowns :)
Fantastic read! I really enjoyed how you broke down a classic Larry Williams strategy with such a thoughtful and data-driven approach, while keeping the tone engaging and humorous. The deep dive across multiple markets, along with the insights on lookback periods and the long vs. short asymmetry, were especially valuable. It’s refreshing to see someone test these ideas thoroughly rather than just accepting them at face value. Definitely sparked a few ideas for further testing on my end — great work!
Thank you for sharing - really enjoyed this post and it sparked some great ideas for me to test as well. Sample size is definitely an issue, I am thinking rather than trading it as a single instrument system it might be worth testing all of the indices in a portfolio so that there are a lot more trades and Backtesting and optimizing parameters using the whole portfolio rather than just a single instrument. This should make the resulting system more robust and allow a trend filter without reducing the number trades too much.
This is an awesome and in depth post, thanks for sharing! Most of these trades have low sample sizes of trades, but if they work in multiple markets provides more confidence in the strategy.
Nice work RQ
Thanks, appreciate it.
Great article! I like your style.
The problem with Larry Williams' patterns—I think I've studied them all—is that they have little statistical significance, and also that this difference masks the true power a pattern can have, although it's an interesting technique to use if you need it for your goals. It's always interesting to study the master Larry.
Congratulations on the article!
Quant Power!
Enjoy. ;)
Indeed.
This is my favorite gif. Instant follow.
what can I say other than…
Really good article btw, look forward to more work of yours. Im constantly building strategies. Can get some good ideas. I also noted to pick up "Long-term secrets to short-term trading". Haven't read that one. Thank you.
This is a really well written and engaging article. I have this book. I need to read it.
Great article! Now, just put all instruments in a portfolio, and you will get a good sample size, a nice equity curve, a surprisingly good Sharpe ratio, and low drawdowns :)
tks man.
Fantastic read! I really enjoyed how you broke down a classic Larry Williams strategy with such a thoughtful and data-driven approach, while keeping the tone engaging and humorous. The deep dive across multiple markets, along with the insights on lookback periods and the long vs. short asymmetry, were especially valuable. It’s refreshing to see someone test these ideas thoroughly rather than just accepting them at face value. Definitely sparked a few ideas for further testing on my end — great work!
tks man
Thank you for sharing - really enjoyed this post and it sparked some great ideas for me to test as well. Sample size is definitely an issue, I am thinking rather than trading it as a single instrument system it might be worth testing all of the indices in a portfolio so that there are a lot more trades and Backtesting and optimizing parameters using the whole portfolio rather than just a single instrument. This should make the resulting system more robust and allow a trend filter without reducing the number trades too much.
Thanks, Adrian. I really enjoy running these backtests bc every time I do, it sparks even more ideas to test. like a domino effect.
I also ran an intraday backtest but was too much info for one article. Will publish this on another post.
Me too - I love the creativity that the process sparks
yep
This is an awesome and in depth post, thanks for sharing! Most of these trades have low sample sizes of trades, but if they work in multiple markets provides more confidence in the strategy.
Tks! And you're 100% right. I address this issue about the sample size but yeah you can get confidence with multiple markets analysis