I've been reading about volatility forecasting. All discussion relates to options.

[dtrader98]

Quote from Kovacs:

Yeah, I want to know enough to understand the academic trading papers and also the formulae in the practitioner books. I doubt I'll go into advanced territory.

I'm reading Calculus: An Intuitive and Physical Approach by Morris Kline. Fantastic for self-teaching.

The simplified version of calculus is you have two major functions introduced above prior math levels; the derivative and the integral.
The derivative is a fancy way to measure rate of change of a curve, the integral is a fancy way to measure cumulative sums or area under a curve.
What makes it complicated is the theory used to approximate properties of continuous functions.

BTW, if you haven't taken any statistics, I would argue that is more important to understanding his book, and trading math in general. There are very few concepts in his book that truly require calc, except maybe solving or reducing complicated integrals, which you likely will not need to do unless you are interested in proofs.


Good Luck.

 

[Kovacs]

Quote from dtrader98:

The simplified version of calculus is you have two major functions introduced above prior math levels; the derivative and the integral.
The derivative is a fancy way to measure rate of change of a curve, the integral is a fancy way to measure cumulative sums or area under a curve.
What makes it complicated is the theory used to approximate properties of continuous functions.

BTW, if you haven't taken any statistics, I would argue that is more important to understanding his book, and trading math in general. There are very few concepts in his book that truly require calc, except maybe solving or reducing complicated integrals, which you likely will not need to do unless you are interested in proofs.


Good Luck.

Thanks. I read more on statistics a few weeks ago when I encountered terms like "first moment", "second moment" etc. Turns out that Pearson applied jargon from mechanical engineering, and that they represent Mean and Std. Dev. respectively, with third and fourth meaning skewness and kurtosis.

Ignorance of terminology trips me up a lot because they make things seem more difficult than they are. So I try to remedy the problem by covering a wider range of topics.

Do you have a recommendation for a practical statistics book?

 

[dtrader98]

Quote from Kovacs:

Thanks. I read more on statistics a few weeks ago when I encountered terms like "first moment", "second moment" etc. Turns out that Pearson applied jargon from mechanical engineering, and that they represent Mean and Std. Dev. respectively, with third and fourth meaning skewness and kurtosis.

Ignorance of terminology trips me up a lot because they make things seem more difficult than they are. So I try to remedy the problem by covering a wider range of topics.

Do you have a recommendation for a practical statistics book?

There are many. If you are used to simple programming, I would suggest to jump start both your trading and update your skills by learning statistics through any introductory R book.

http://www.amazon.com/Introductory-...=sr_1_1?ie=UTF8&s=books&qid=1264289032&sr=8-1

A good basic statistics text is :
http://www.amazon.com/Statistics-4t...=sr_1_1?ie=UTF8&s=books&qid=1264289790&sr=1-1

I like the R books better, because by running examples, you can get a lot more hands on approach to practicing statistical examples, while learning a very useful trading tool.

You might choose to learn statistics along with finance (i.e. taking the most relevant parts of statistics with hands on applications) using a text like this great text that uses excel
http://www.amazon.com/Financial-Mod...=sr_1_5?ie=UTF8&s=books&qid=1264289173&sr=1-5

In some ways, I can see why you might have trouble following Vol trading book, because a lot of the terminology (while simple) requires some basic exposure to statistics. The entire beginning of the book expands on many methods and pitfalls to measure standard deviation and volatility, which are beyond the most fundamental approach to standard deviation. You really should understand basic properties of standard deviation from statistics first, to get an appreciation of his various citations and progression of approaches.

Another approach would be to take a complete trading book that is based on principles of statistics in practice, it doesn't cover systems but does a good job
of explaining how statistics is relevant to
evaluating trading systems.
http://www.amazon.com/Evidence-Base...=sr_1_1?ie=UTF8&s=books&qid=1264289615&sr=1-1

Or just buy this one to have fun understanding it via baseball...
http://www.amazon.com/Curve-Ball-Baseball-Statistics-Chance/dp/038700193X/ref=ntt_at_ep_dpi_3

Unless you are doing heavy quantitative stuff, I would argue to really learn statistics and R (which is free) first.
A second subject which is pretty important to learn would be linear algebra. You will see an awful lot of this as well in financial texts, it is a requirement to really understanding a lot of statistics theory
that you will see in papers (like regression matrix equations, for instance). The main relevant theory is covered very well in the excel Benninga book mentioned above.

 

[Johno]

Quote from psytrade:

problem with anything in trading is that correlation doesn't predict causality.

That being said some people have good models of volatility but I doubt it has anything to do with only simple volatility measures of implied vol.

The OP needs to determine whether he wants to develop an accurate model or simply persue academic questions. I can state for a fact that you don' need an indepth knowledge of calculus or algebra in order to create an accurate and reliable model that can be traded easily with very little stress! The answer for most people is not in price charts or indicators derivatived from these advanced concepts. Consider the principles ( based in creative thinking ) that ancient mariners harnessed to navigate through the unknown for most of us the answers can be found here.

Best Regards

Johno

 

[Kovacs]

dtrader98,

Thanks for the list of recommendations and the advice.

I've been reading the R book and I'm on the second chapter. I'll browse the other titles and maybe pick up a book on linear algebra.

 

[dtrader98]

Quote from Kovacs:

dtrader98,

Thanks for the list of recommendations and the advice.

I've been reading the R book and I'm on the second chapter. I'll browse the other titles and maybe pick up a book on linear algebra.

Good to hear. Hold off on the lin alg book. If you can get through an R stats book, you should be ok. Again, I'd suggest getting the benninga excel book to cover linear algebra (kill finance and lin alg in one stone)-- worth owning, or just pick up a lin alg book at a library.

Cheers,
dt

 

[intradaybill]

Quote from Kovacs:

But if I can forecast the volatility *range* of a stock for a given time interval, can't I use that information as a filter to identify trades on the stock directly?

Wrong approach. The correct approach is figuring out ways of dealing with unpredictable volatility so that risk of ruin is minimized and expected profit is maximized - both together. Volatility cannot be forecasted within a sufficient accuracy to provide an edge in trading. This is equivalent to knowing the future events in the world with a sufficient accuracy. It presupposes some kind of powers not available to any human beings.