Comments on: Calculating Historical Volatility

By: Tom

Tom — Fri, 06 Jun 2008 10:25:50 +0000

John, for most of my HV calculations I use natural log (LN) to calculate the continuously compounded return and then a standard deviation at some time period for those cc returns. The HV calculations are different than calculating volatility for Implied Volatility.

By: john

john — Thu, 05 Jun 2008 20:04:21 +0000

Tom,
Being a mediocre quantitative finance student, i think that (not sure though) while calculating volatility, the choice of return depends on for what you are going to use it for.
Like, if you are going to use it in a b&s formula (continuous time), you have to calculate the log returns, but if you are going to work in discrete time, then you should calculate volatility from arithmetic returns. However, I dont think there will be a big difference..

By: Career Advice by Randy

Career Advice by Randy — Mon, 18 Jun 2007 19:27:59 +0000

festival of investing - June 12, 2007…

Welcome to the June 12, 2007 edition of festival of investing.
Tushar Mathur presents Financial Planner: Do I need one ? posted at Life of a Resident Alien….
Thomas Ott presents Calculating Historical Volatility posted at Neural Market Trend…

By: admin

admin — Mon, 04 Jun 2007 11:08:03 +0000

Daniel,
I checked out your blog and I like it. I’m traveling right now so I’ll read it in detail when I get back. Thanks!

By: Daniel

Daniel — Mon, 04 Jun 2007 04:50:53 +0000

Hey dude! Read your post from the blog carnival. check out this post http://israelispeculator.com/blog/estimating-future-volatility-of-a-stock-like-the-pros/

By: Tom

Tom — Sat, 26 May 2007 11:44:24 +0000

Cpptrader: Thanks, I came across something a few minutes ago that corroborates your statement. I’m glad I opened this up to the public for scrutiny.

I’m going to make changes to it and republish!

By: cpptrader

cpptrader — Fri, 25 May 2007 20:58:33 +0000

Unless there is something I am unfamiliar with, Wikipedia has the estimated vol equation wrong. The denominator is not sqrt(1/t), it is simply sqrt(t). A book I am reading (Introduction to High Frequency Finance) explained it using only t, and I found a couple sources online that equate it with only t in the denominator, not 1/t. For example, I found this site at the university of nebraska: http://www.math.unl.edu/~sdunbar1/Teaching/MathematicalFinance/Lessons/BlackScholes/ImpliedVolatility/impliedvolatility.shtml

Like I said, maybe I am missing something but I don’t think so. Just FYI

By: Tom

Tom — Thu, 24 May 2007 13:37:08 +0000

Ernst,

Interesting. We should compare notes maybe there’s something we can collaborate on? Would you contact me directly at desertroot -at- gmail dotcom? I’ll pass on some charts I’ve created using my model.

By: Ernst

Ernst — Thu, 24 May 2007 13:29:49 +0000

Hi: I am working on a similar model. I have decided to use the VIX for my volatility. I am also going to add the premium for the VIX ATM put strike.
The VIX is real money (big money) telling you how much movement you can expect in the (near) future.
For instance if you check the VIX of the last two months and project the premium for the ATM put strike. You see that fear is building up. In March and April every time the market went up a couple of day the VIX ATM put strike would cost you around 40/50ct. Telling you that the market expected expiry with a VIX around 11.50. The last couple of weeks after weeks of up days the market the premium does not come back into the put strikes. Telling me the market does not expect the VIX to come down soon.

By: Tom

Tom — Wed, 23 May 2007 17:17:11 +0000

Neuro: Thanks for stopping by! I’d love to learn C++ but I have no time for it. I end up building macros and programming (limited) in Excel!