Hey guys, This week, I tested the enhanced strategy as suggested by Larry Connors in his "Short Term Trading Strategies That Work” book. It uses Cumulative RSIs instead of the vanilla RSI. They do offer a statistical edge (details are in the full article). The strategy achieves a 26% annual return with 1.18 Sharpe. The maximum drawdown is 37%, much better than the S&P 500's (57%) during the 24-year period but still too high. From now on, I decided to share a sensitivity analysis to prevent overfitting, something several people here talked about. The results below show close to 200 experiments obtained by varying the parameters in the parameter space: Equity curve, drawdown curve, and annual returns for the enhanced strategy 198 variations Statistics for the main indicators of the 198 experiments Sensitivity of annual return for maximum positions of 3 The rules in detail: Whenever the 2-day cumulative RSI (2 days) is below 10, we buy on the next opening; Whenever the 2-day cumulative RSI (2 days) is above 65, we exit on the next opening; We only trade a stock if its price is above its 200-day SMA; We trade multiple stocks in parallel to increase exposure; If there are more than 3 opportunities to trade in any given day, we sort them by market cap and prioritize the lower market caps; We restrict ourselves to trade only large & mega caps to reduce the risk of delisting; We limit to holding 3 positions at any given time, maximum; Whenever the stock closes below its 200-day SMA, if we have a position, we exit at the next open. To ensure we trade only liquid stocks: We will restrict ourselves to only trade stocks that have been traded in all sessions over the past 3 months from the day in question; We will only trade the stock if the allocated capital for the trade does not exceed 5% of the stock's median ADV of the past 3 months from the day in question. All implementation details and discussions are here. I'd love to hear what you guys think. Cheers!
Not to be overly critical but back testing is generally worthless and Larry Connors does not trade Larry Connors, his website doesn't have a particle performance or trade data - it's basically a book sales room. As we all know books are obsolete as soon as the ink dries.
This is great work. Ignore people that say backtesting is worthless, that a load of shit. Sure you need to be careful to not overfit (hence this is great analysis) which you did here. All of my trading which has been successful was originally based on a back test. My focus when backtesting is squarely on robustness. I want to see a strategy generally work in a range of sensible parameters for the indicators or techniques used. This seems to be exactly what you did here. It is a joke when people say negative things about back testing because it’s based on past data lol what other fucking data could we possibly have lol?
if it's no longer contemporary or updated how could it not be. Is yesterday's newspaper useful for today? It's last week stock summary useful for today? Most printed information degrades almost immediately. And do be adroit enough to recognize that absolute knowledge or data, like the atomic weight of copper, remains contemporary forever.
View attachment 342852 View attachment 342851 View attachment 342851 for those with thimble based intelligence on back testing see the information link below provided by Ninja. https://ninjatrader.com/support/helpGuides/nt8/NT HelpGuide English.html?discrepancies_real-time_vs_bac.htm. In terms of what data would we use otherwise it's called forward walk real time trading. Look at the performance reports the any 50 robots listed in iCannon systems and compare back testing to real time forward walk to cash trading and your eyes will be open and your mouth will be shut. I've included a simple one: back testing showed over $187,000 in gain (white bkg) but real time tracking $5600 loss (yellow bkg).
What is (white bkg) and (yellow bkg)? Squiggly line profit/loss line (does that ya know, up down) is blue. Are you disappointed its not straight to the moon, never dropping at all?
The white background is back testing +$295,000 that is to the moon by your thinking and the yellow background, just about six months, it's real time tracking with -$99 which would be not to the moon by your thinking but more like reality to the rest of us. As you can see I'm not disappointed but informed as you are now. I would be disappointed with a $22,000 drawdown, you?