Guys,
Does anyone know how to model a volatility cone? or can point me in the right direction?
Thank you.
What do they mean by probability cone?
It looks like a useful idea would be a model for price with point-wise and simultaneous confidence (or prediction, depending on the application) bands.
If you can give me more background about what it is you would like to do, then we can work on a solution. Standard deviations might be totally inappropriate for what you want to do here.
Hey thanks for the reply.
This is what i’m trying to do. I have a stock historical prices charted in excel and would like to using the annualized standard deviation to model the price going-forward one year. Let’s say it’s a 20% s.d.
So the chart should model the price going out one year. The stock could go up to 120, stay at 100 or go down to 80.
Do you want any measure of statistical reliability attached to that? Using the SD alone will really only give you a description of the data set.
If you don’t want any measure of statistical reliability attached (which is implied by using only the SD) you could just calculate the returns at the interval (if you want monthly prices, then do monthly) desired for a few years and rank order them. If you want the 97.5th percentile and 2.5th percentile returns, then draw these out and use them to turn time t=0 price into t+1. You could/should do this base on past returns. The window used should change each period.
Another solution is to create a model and simulate the prices out to a certain period while maintaining a certain confidence level and the bands represent a specified level of confidence (95% for example), but this is different than just using the standard deviation.