SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Strategies & Market Trends : Ask DrBob

 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: broadstbull who wrote (38402)6/30/2001 12:46:43 PM
From: xu, b.  Read Replies (1) of 100058
 
Peace,

"Notice how the band were not that effective in March, due to the extreme decline"

You made a good observation. In fact, the probability of SD band works when the sample data has a normal distribution. Financial data have been proven that their distribution is not normal. It has a longer tail than the normal distribution. You observed one example of this effect.

Another example is LTCA fiasco in 98. FED had to bail them out. It was a result of unexpected rate spike (as a result of Russia defaulted some debts) that caused a good model (developed by ex-FED Vice Chair and Nobel Prize Laureates) to fail. What really caused problem is their money management rules. Money management rules dictate that no robust model can show very high annual return unless one completely ignores the risk. Check the link below and you can see the results of several models with > 1000% annual return (no position size rules).

Speaking about probability, anyone remember 10/19/87? It was a "brief" pain but the probability for that event to happen (according to normal distribution assumption) is << 1%. It happened only 5 years after the index future was introduced.

You can find a free site that discusses statistical trading method at geocities.com.
Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext