Tuesday, November 9, 2010

Odds for the Dow on 11/10/2010

The trajectory has changed rather dramatically, which does raise some questions in my mind about how useful this is for forecasting. I'm not surprised to see the forecast change -- it's adaptive in the same way forecasting the weather is adaptive. As conditions change, the forecast changes, and often times looking a few days out is difficult. In this case, I'm starting to see that the 4 or 5 period forecast is pretty unreliable. I'm not sure how it can be used. The one period, two period and three period forecast seem to be a little helpful at giving possible targets.

So how can this be used? I don't know. The trajectory algorithm is my own creation and it's a bit arbitrary. I think the odds table itself contains what you need to know. I only know the next day forecast has been pretty accurate, and the two-three day forecast has been giving reasonable price targets.

 Category Prediction Rate % Correct Notes Direction 11/14 79% 13/14 disregarding minor closing range Range 8/14 57% 12/14 disregarding minor closing range Significant High/Low* no instances Unfilled Gap** 1/2 50% Testing Tomorrow Next Day Gap** 2/2 100% Testing Tomorrow Gap Fill+ 1/1 100%
* A significant high/low occurs when the market makes a move greater than 0.5%. This test is measuring the following case: If the odds of a significant high/low are >= 70% in either direction, a significant high/low was made the following day.
** An unfilled gap occurs when the market gaps and never fills. The next day gap determines if the condition correctly predicted the gap.The unfilled gap and next day gap tests are measuring times the odds of a positive low or negative high were >= 14%.
+ When the odds of a negative high/pos low <= 3% and the market gaps in the low odds direction, the gap fills.

I have to post the trajectory and full odds table tonight because I'm mobile today.

Direction was correct today while range was wrong. The trajectory seemed to indicate that a large pullback was coming. It got the day wrong, but the price (around Dow 11,300) was hit today. Is the pullback over? Possibly. The odds for tomorrow are positive, but the odds of a negative high are as high as I've seen.

We'll be testing tomorrow:
• Since the odds of a negative high >= 14%, does the market gap down and never fill the gap?
• Since the odds of a negative high >= 14%, does the market gap down tomorrow?
We've seen consistent results on the next day gap metric, but not so much on the unfilled gap prediction, so I think it's very possible the market gaps down tomorrow. Let's see if we can get a bigger sample to find a rule here.The possibility of a gap up is also 10%, so I'm not so sure I'd jump in with both feet here. Remember, with 13/14 accuracy when you disregard minor closing range, I am getting the hint that you should not bet against the directional odds.

Being short here doesn't look all that favorable, because the odds of a significant low tomorrow are just 43%. Also, remember that the trajectory from a few days ago pointed higher a few days out (not sure how it looks now though). So even if the market does gap down, it's not likely to be a big drop. We'll see.