Thursday, October 21, 2010

Odds for 10/22/2010

I'll be posting odds after hours from now on. Yesterday's prediction was a good one, 60% chance of an up day, and the range with the highest probability was 0-0.5%, which was hit. I'll post my other forecast later.

Tomorrow looks like a slightly negative bias and the highest probability is a down day from 0 to -0.50%.

To understand the second two tables, they are looking at the high and lows of the next day for all cases. Want you want to see if you're bullish is very high odds (let's say > 75%) of a "significant high", which I count as 0.50% or higher. This gives you the best chance of walking away with some day trading profits if you are long at the close. Vice versa for a "significant low". In this case, I don't see strong enough odds to think either side could make significant money tomorrow.


  1. Very nice! You should start to keep some kind of running score of how well your prediction is going (and also to give you an idea of in what ways your prediction is the strongest). Also, since you have historical data, you don't have to wait to see how well your system does ... you can just go back in time. For instance, how well could you have predicted what happened each day around the last April top. If this goes well, you will likely draw commercial interest ... or even start a subscription service if it would be of interest and helpful to you (using something like PayPal). Deja Vu is a great name for a daily prediction service! If you haven't already, you should see how to trademark it. If you are thinking you would like to go in one of those directions, I would not show so much about the techniques you are using. Keep up the good work. Best Wishes, Peter

  2. Well I doubt I'll ever ask for money for my trading systems. I think the stock market is so rigged against non-professionals that I consider anything edge I give to be a public service. However, you're right, I haven't fully tested this to see how it performed in a back-test. Over the weekend I might do a bigger sample to see the results.

    As of now, this is just pure information. Whether or not it has predictive value remains to be seen or tested. For me, it's just for fun.