I am trying something a little different, normally I build day trading strategies, but find it very difficult to overcome slippage and commissions on the tight time frame. For instance, on ES if I have a strategy that is yielding $50 per trade. After adding in $5 (rounded) for commissions, and then 1 tick slippage in both directions, I am now left with only +$20 per trade expectancy and a horrible looking equity curve.

To combat this, I am having my "bot" now produce automated analysis on the daily time frame for S&P stocks. So far this seems very promising. For tomorrow for instance I have a handful of stocks to buy with an expectancy of several percent for a potential 5 day hold. I think changing time frames may make being profitable easier.. Thoughts?

I have a blog set up to keep track of my picks and results as I see how it works going forward. If you want to keep up with how its doing, check it out MyStockAnalysis.com.... I will also post some summary results in this thread.

For example based on historical results, I am expecting to earn +3.5% on NEM in the next 5 trading days.. Even after significant slippage, the equity curve looks great. Have you tried something similar? Good results? Bad?

The larger time frame, generally the larger the reward per trade (and the risk...). This helps you more easily overcome entry barriers like slippage and commission, as they become a smaller and smaller percentage of your profit from each trade.

Just make sure you aren't overly curve fitting. That is infinitely easier to do the larger and larger time frame you optimize against. Most strategies I am toying with these days are trading thousands of times a week, because I tend to think a strategy should be good at something a discretionary trader isn't -- such as high-speed, low latency, fast round turns, etc. This also generally means I personally focus on smaller time frames, not larger ones, with my strategies.

Are you already a profitable discretionary trader?

Mike