Standard Deviation Question

I just realized I may have messed up on my pairs spreadsheet. I calculated 1 st dev with excel, and just multiplied it by 2 to get 2 st dev's. Is that how to get the 2nd st dev?


Dustin
 
Dustin; I don't know whether this will help you or not but FWIW the rule of thumb is;

1 std. dev covers 68.3 % of all possible outcomes
2 std devs. cover 95.4% of all possible outcomes
3 std devs cover 99.7 % of all possible outcomes.

perhaps you can figure a way to use these relationships to get where you want to go.
 
Originally posted by Dustin
I just realized I may have messed up on my pairs spreadsheet. I calculated 1 st dev with excel, and just multiplied it by 2 to get 2 st dev's. Is that how to get the 2nd st dev?
Yes.
 
first you calculate what a standard deviation is for your sample. One standard deviation is 1x, 2 standard deviations is 2x. The area under the curve will be significantly different though.
 
Hopefully a math/stats guy can shed some light on this question.
Does this standard deviation calculation apply to the stock prices?
(i think there is an assumption that there is a 'bell-shaped' normal distribution)...tks
 
The standard deviation is a component of the Bell curve which is used to determine probabilities. It can be used to analyze any data.
 
Originally posted by Rigel
The standard deviation is a component of the Bell curve which is used to determine probabilities. It can be used to analyze any data.

It's not a component, it describes the Bell curve. Of course the results of operations assuming a normal distribution and assumptions as to the probability significance of certain standard deviation multiples will only hold if the data sample in question in fact conforms to a normal distribution.

The problem with financial markets in this respect is that they may fit a normal distribution pretty well for much of the time, but usually when it's most critical (like in a crash) the assumption will fail due to kurtosis and fat tails.
 
Originally posted by dlincke


The problem with financial markets in this respect is that they may fit a normal distribution pretty well for much of the time, but usually when it's most critical (like in a crash) the assumption will fail due to kurtosis and fat tails.

OK, so the market does exhibit a normal distribution with the exception of major market crashes? So one can expect most of the price action to be contained within 2 std deviations the majority of the time.
 
Back
Top