Minyan Mailbag: Measuring and Trading Volatility
...no matter how many models you have, statistics you keep, or studies you do this is all a weird art-form.
How's it going?
I've been spending some time looking at option pricing models and attempting to learn how to measure and trade volatility. I'm beginning to understand the relationship between an option's price and its implied volatility, as well as the relationship between implied volatility and historical vol. However, given all the unknowns in pricing an option for the future based on its not-so-predictive past volatility, how would I approach the timing issue of the likelihood of said future (in this case, higher vol/option price) by expiry?
Say you absolutely knew the direction the underlying security would move in and the magnitude of the move, but you still couldn't time the movement. How would you approach this trade with options, if at all? If the asset were particularly non-volatile (by percentile) lately would you assume that current option prices were relatively cheap and that volatility should regress to its mean?
I guess, if so, this rate of regression would have to be some function of the previous time periods of fluctuating volatility (i.e. volatility has been steadily decreasing at a monthly rate to its yearly low and so should regress at a similar rate -- the volatility of volatility), right? In that case, since your predicted asset movement is likely correlated with an increase in volatility (this isn't necessarily true as it could simply move within its vol range, correct?) can you time your options by finding the established rate of change of volatility? Intuitively I would think this would increase the chances, but it's still not predictive.
Or, do you simply consider the price movement you expect to see, consider the asset's recent volatility, and then look for options with an attractive delta? Just eyeballing it doesn't seem like a solid approach for somebody with as little experience as myself. I feel like the correct approach might be developing a probability model for directional asset movement and synthesizing that with some way of calculating the probability of a significant change in volatility... just not sure where to even begin on the vol side of things.
I really appreciate any insight you guys can share, and sorry for all the questions.
I have written a great deal about the basics of what I do, trying to give people the understanding they need to use options strategically. I have refrained from much detail in volatility trading for personal reasons.
First of all no matter how many models you have, statistics you keep, or studies you do this is all a weird art-form. Subjectivity creates risk (I go through periods where I lose money, some significant), but also opportunity. I have explained this before: the option pricing models that exist (even proprietary ones) are very limited in scope. They do not explain much of asset price movement. So we even start from a flawed base.
We cannot predict volatility. There are some clues from both a macro and micro basis, but these are only probabilities. You can tell when an option is cheap, but it may become cheaper still. Managing risk is essential.
One hint: Diversification along optimal F is essential to successful volatility trading. This maximizes probabilities which is all one can do.
John's "weird art form" comments are another way of saying "the complexity associated with reflexivity" in asset price discovery. Which is another way of saying that, unlike static systems ranging from blackjack to earthquake prediction, what we think about the probabilities actually changes the eventual realization of the probabilities. If you could count cards perfectly, the probabilities would be known to the fifth significant digit and you would play those probabilities accordingly. But not so with stock or other negotiated financial markets. When competing with other traders in an exchange, what we think about the probability distribution actually AFFECTS the resulting probability distribution. If you think the probabilities of a stock going from $5 to $10 are high, you will act on that belief (note I am using the word belief here and all that conjures up in terms of the heuristics that humans use in their decision making) and your acting on that belief will actually increase the probabilities of it coming true. So when John refers to subjectivity, this is what he is talking about. And this is what makes prediction impossible. But prediction isn't a probability assessment.
But all is not lost from a probability assessment standpoint. The mechanism that best captures this idea of subjectivity is the concept of herding; traders, when faced with situations where they simply don't "know" the outcome, are wired to believe that someone else does. And they do what "someone else" does. On that score, you should know that herding (at least human herding in negotiated financial markets) has a fingerprint (of sorts - it's a robust fractal) that can be 'described' by a power law equation (among other tools that our firm works on every day). So with that insight - that subjectivity creates herding which creates robust fractal fingerprints that can be described with power law equations - you CAN assess probability distributions better (FAR better) than with a flawed linear model like (just to pick on Myron) the Black-Scholes option pricing theory. Remember what we said in Vail at Minyans in the Mountains: we are not in the prediction business, we are in the probability distribution business. And based on the explanation above, you now know why (and know some of the theoretical tools we use to boot).
Lastly, John referred to money management techniques as being critical to success irrespective of what probability assessment tools you are using. Neither of us could impart on you how important this is. John's reference to Optimal F is one money management technique that is out there. Another is the Kelly Criterion (I would strongly recommend you do as much research into it - history, development, mathematics, etc. - as you can bear). Both attempt to maximize the long term geometric mean growth rate of wealth while simultaneously decreasing the probabilities of total destruction of capital. It's opposite is called Martingale betting and Martingale betting is essentially what everyone does - and it's a near sure road to ruin in systems where probabilities fluctuate as wildly as they do in negotiated financial markets.
There are lots of keywords and concepts in there; as far as I am concerned, you can't read enough about them.
The information on this website solely reflects the analysis of or opinion about the performance of securities and financial markets by the writers whose articles appear on the site. The views expressed by the writers are not necessarily the views of Minyanville Media, Inc. or members of its management. Nothing contained on the website is intended to constitute a recommendation or advice addressed to an individual investor or category of investors to purchase, sell or hold any security, or to take any action with respect to the prospective movement of the securities markets or to solicit the purchase or sale of any security. Any investment decisions must be made by the reader either individually or in consultation with his or her investment professional. Minyanville writers and staff may trade or hold positions in securities that are discussed in articles appearing on the website. Writers of articles are required to disclose whether they have a position in any stock or fund discussed in an article, but are not permitted to disclose the size or direction of the position. Nothing on this website is intended to solicit business of any kind for a writer's business or fund. Minyanville management and staff as well as contributing writers will not respond to emails or other communications requesting investment advice.
Copyright 2011 Minyanville Media, Inc. All Rights Reserved.
Daily Recap Newsletter