ghariton wrote:No. I think you will find broad agreement among financial analysts, etc, that what is relevant is the future volatility.
Why the obsession with future volatility to the exclusion of future value? To me volatility is a given if I am going to invest in equities, and even to some extent if I am going to invest in corporate bonds. The real question to me is not what volatility is going to be in the future, but is a current appearant discrepency between price and value a symptom of volatility or an indication that value has changed?
ghariton wrote:I said above that one should not mindlessly extrapolate the past. But, as a good Bayesian, I also believe that there is a lot of useful information in the past. One should not merely throw it away.
But what does that volatility measure tell me that is of any use? It does not explain why one security is more volatile than another, it simply tells me that it was. I don't see any practical value in it from a long term view of investing, and other than trying to pick the best price possible for a buy or a sell, I have little interest in technical analysis and to me volatility is really only of significance to that.
ghariton wrote:You are, of course, free to make up any definitions you wish. As you say, people who have developed a definition that they find useful for their purpose, will ignore your definition.
I have already beaten my head against the wall many times on the Agnostics mailing list trying to explain that even the dictionary claims that Atheists "believe" in the non-existance of god. However a large contention of atheists insist they have no such belief.
The problem is that the technical term disabuses a much more general concept and by those in the industry continuously using the word Risk to represent a particular metric that can be measured, they are likely to fall into a trap believing that this metric measures the probability or likely hood of loss in an investment which it in no way does.
ghariton wrote:In general, people on this thread should remember that volatility is a concept that includes the entire probability distribution of possible outcomes. Standard deviation is just one summary measure of that distribution, nothing more. Depending on circumstances, it is a more or less useful summary. But then that is true of all summaries. In particular, one should not rely on summaries for taking important decisions.
Ah, I can't see how volatility includes the entire probability distribution of possible outcomes. It can only be measured for those outcomes that happened not those which were unlikely but not impossible which did not occur. Am I missing something here? Have you a definition of volatility which looks forward to events that have never happened and assigns some level of probability to them?