Options Course: On Saturday, November 10, the CBOE, Option Pit and OptionsProfits are hosting a full-day class, Using Volatility to Improve Directional Trading



I have been thinking a lot about differentiating signals from noise. Almost every month you'll hear about some new indicator, or quantitative study, or chart pattern that is supposed to offer better predictions about future returns. Most don't, but even the indicators that prove valuable over time ultimately fall into basic categories expressing well-known factors like momentum, liquidity effects, mean reversion (value investing) and so on.

The same thing goes for options markets. We can survey all the different types of information we get from options markets and how they are related, and think about which ones have the most predictive value. We could start with familiar pieces of data like the level of implied volatility in a weighted series of options, or the daily volume of options traded relative to some average. But any of those general types of information can be specified further: we can distinguish the absolute level of implied volatility from the rate of change (i.e. momentum) in IV, or from the realized standard deviation of IV (i.e. vol of vol), and so on with order flow and IV skew and many other inputs. Research departments and academics spend a lot of time tracking and defining individual variables and data points, and seem to spend much less time synthesizing the evidence we have thus far to see which sources of information have proven valuable since publication.

To help focus my own thinking and planning, I thought it would be useful to list the options-based factors that seem to have been the most informative in recent years:

1. the momentum of volatility

2. the realized volatility of volatility

3. the ratio of implied to historical volatility

4. the change in option open interest

5. ranked implied volatility skew

6. slope of implied volatility term structure

Again, each of these can be specified further and broken out into multiple complementary variations. For instance, we could look at the momentum of volatility across different time horizons, or at changes in open interest at different levels of moneyness, etc. I've included a link for each item to make it easier to find explanations already published here at OP and elsewhere.

There are some topics that seem relevant and even intuitive but about which I haven't seen enough research to warrant inclusion in the list above. For example, the market-implied correlation among equity index constituents is intuitively appealing, and it makes sense to use realized correlation for focusing on stock-picking versus indexing. Another topic I'd like to include is the relationship between options activity and corporate earnings, or the order flow around scheduled releases of important economic data. Most of the research I have mentioned so far concerns equity and index options, but as more investors become familiar with commodities and fixed income, there is a lot of room for research to examine which of these indicators have shown predictive power in other markets.

If you can think of an indicator or datum that isn't included above but should be, let me know in the comments.

OptionsProfits can be followed on Twitter at twitter.com/OptionsProfits

Jared can be followed on Twitter at twitter.com/CondorOptions