The trouble is, we're in uncharted territory. You can't make a proper conditional probability such as- If the market sets new lows at 33 months, what are the odds it will have new lows past 37 months. Because there is no data set to draw from. It's never happened.
Now that is WELL SAID, and it is exactly the point. We have NO data upon which to make a statistical inference of what is to come.
Does that mean it can't happen? NO, but the BEST calculation I can make on it is that the odds are 10-1 it will make new lows. YES, those are the odds at the beginning of the cycle, but that is the only logical way to place probabilities on it that I have come up with.
Unfortunately, this is not the best calculation you can make. Based on what you said in the previous paragraph, and the point I was trying to make, this calculation is irrelevant. Actually, it is worse than that. Based on statistical theory, it is just plain wrong.
Since you appear to be a statistical professional, maybe you can help me find a better method. Is there a 'fudge factor' you can multiply by to estimate for non-existent data? Any suggestions?
The statistical professionals would be aghast at the idea of me being named one of them <ggg> I did have to learn a few things along the way, so I am sure of my ground on this issue.
Macavity did the proper calculation in a later post
Message 18548246
I have not looked at the "input" data he used, but based on your numbers it sounded reasonable, and the math is right. But the bottom line on this is that there is no statistical calculation, based on available cycle data, that will predict when this cycle will end. Your time would be better spent making your case based on other market and economic factors. |
