ImpliedRiskReport: 08/26/07

What is Implied Risk?

Implied Risk (IR) readings are really a translation of certain data elements produced by the listed options market. We use option implied volatility to produce expected trading ranges for issues, like equities, indexes and ETFs.

These trading ranges have unique characteristics that are based on the data's format. Option implied volatility is expressed as a function of the normal curve. Therefore, we can apply normal curve characteristics to our understanding of an issue's action within its own expected trading range.

The normal-curve characteristic that concerns us most is reversion to the mean. Here's how we look at mean reversion: The further away from the mean, the more likely the next move will be back toward the mean.

It breaks down like this. There are six standard deviations in the normal curve, three above the mean and three below the mean. Normal curve theory states that 2/3 of an issue's price points will be within one standard deviation -- up or down -- of that issue's mean price.

Take it to two standard deviations, up or down, and normal curve theory expects a full 96% of price points to be within that range. Price movements outside of two standard deviations of the mean are rare events.

Here's how we look at it: An issue that is two standard deviations from its mean price on a 90-day basis is a perfect candidate for reversion to the mean. So we look for issues that have reached extremes, both as trading/investment ideas and as a harbinger for where the market could be heading next.

Our Favorite Market Books

-- (2000) When Genius Failed: The Rise and Fall of Long Term Capital Management, Roger Lowenstein
-- (2000) Active Portfolio Management: A Quantitative Approach for Producing Superior Returns and Controlling Risk, Richard Gringold and Ronald Kahn
-- (1996) Against the Gods: The Remarkable Story of Risk, Peter Bernstein
-- (1994) Stocks for the Long Run, Jeremy Siegel
-- (1992) The Winner's Curse: Paradoxes and Anomalies of Economic Life, Richard Thaler
-- (1989) Market Volatility, Robert Shiller
-- (1986) Options As a Strategic Investment, Larry McMillan
-- (1973) A Random Walk Down Wall Street, Burton Malkiel
-- (1923) Reminiscences of a Stock Operator, Edwin Lefevre
-- (1922) The Stock Market Barometer, William Hamilton

Sunday, August 26, 2007

Implied Risk and Three of Our Favorite Books

Implied Risk data works -- helps people make more money and lose less -- because the theories surrounding the Normal Curve are perfect for investors who like to buy issues low and sell them high.

The best book on risk and the stock market is Against the Gods: The Remarkable Story of Risk ('96), by Peter Bernstein. The author traces the development of probability theory, which is based on the Normal Curve, to the observation of everyday life occurrences. Now probability theory has developed to the point that it's fair to say most of science is based on, or incorporates, some form of probability theory. The first reason Implied Risk data works is because it's based on the Normal Curve and probability theory.

Probability theory was introduced to the stock market by Harry Markowitz, who defined the concepts that William Sharpe turned into a risk measurement called "beta." (Both won a Nobel Prize for their efforts.) Beta measures an issue's risk relative to that of a constant, such as the historical volatility of the S&P 500 (SPX) or the rate of return from a three-month Treasury bill. In the absence of a market to provide true risk estimates, these relative risk estimates were a vast improvement on no risk estimates at all.

The listed options market debuted in 1973, first with calls only. Puts were added in 1974, paving the way for the theoretical option pricing model developed by Fischer Black and Myron Scholes in 1976. The Black/Scholes model introduced the world to the idea of option implied volatility, or the expected volatility of the underlying issue as implied by the price of the option. The key to this approach was the creation of the listed options market before Black and Scholes wrote their paper. Implied volatility is a product of the forward-looking listed options market, while beta is a product of the backward-looking calculator. That's the second reason Implied Risk data works:
Option implied volatility is produced in the marketplace by stock owners expressing risk concerns about the issues they own.

Another reason Implied Risk data works originated in Beat the Dealer ('66), a book about blackjack and card counting by Edward Thorp. He demonstrated how a seemingly random game could become predictable based upon immediately prior events (observations). The goal of card counting is to determine when the odds turn in your favor. Example: If a high number of face cards have appeared, you can accept new cards more readily because the chances of busting have dropped. Same with stock prices: If the stock has moved to the bottom of the market's estimated range for that issue, the chances of the stock dropping much further are minimal -- as long as you've chosen a fundamentally worthy issue. That's the third reason the Implied Risk data works: The imposition of a scale, along with rules attendant to the scale, permits the data to display an opinion concerning the expectations for further movement.

The key here again is the use of a scale, combined with the observation and analysis of immediately prior events. In blackjack, the scale is 52 cards X #of decks and the observations are which cards have appeared this time around the shoe. In stocks, the scale is three standard deviations from the current mean price and the observations are the stock's current price in relation to that current mean price.

Thorpe demonstrated that the imposition of a scale coupled with observations of events from the immediate past can increase the odds of certain actions occurring, turning a seemingly random game into one that can be somewhat predictable. Another work that continues the thought, The Winner's Curse ('92) by University of Chicago finance professor Richard Thaler, elucidates why opportunities -- he calls them anomalies -- should be expected to occur at either end of the Normal Curve. The anomaly with which we’re most concerned is called reversion to the mean.

Thaler and his colleagues in the burgeoning field of behavioral finance cite the prevalence of financial anomalies as proof that the theory of the fully rational investor is full of holes. One colleague, Yale economics professor Robert Shiller, has demonstrated statistically that markets "overreact" to events, an anomaly Shiller traces to behavioral factors. (Shiller coined the phrase "irrational exuberance" for Federal Reserve Chairman Alan Greeenspan.) Indeed, issues move to three standard deviations away from their 90-day means with greater frequency than the Normal Curve expects. But since Implied Risk data attempts to identify anomalies as interesting trading opportunities, the more the merrier.

One portion of Thaler's book examines the data supporting the conclusion that the phenomenon of reversion to the mean affects the stock market. "Indeed, stock prices do appear to be somewhat predictable," Thaler writes. "In particular, if one takes a long-term perspective (three-seven years) or examines individual securities that have experienced extreme price movements, then stock returns display significant negative serial correlation, in other words, prices are mean reverting." Implied Risk customers believe that stock price movements are mean-reverting over periods as short as three months, and they further believe that the options market provides a very reliable estimate of standard deviation from the mean. Maybe one day Thaler will study Implied Risk customers. Until then, that's the fourth reason Implied Risk data works: Reversion to mean is a fact of life among listed issues.

If reversion to the mean can be measured on a long-term basis of more than 100 years, we should be able to measure it on a shorter term basis such as three months. And if mean reversion works for the overall market, it must work for individual issues. Implied Risk data takes the theory of mean reversion and combines it with option market assumptions to produce useable information about the direction stocks will take in the immediate future.

The ultimate reason Implied Risk data works is because it's built upon sound theoretical assumptions. Two of the assumptions cited -- standard deviation as scale, and reversion to the mean as likely -- spring from the first assumption, that Normal Curve and probability theory apply to the stock market. The final assumption, market volatility as expressed through the options market, is what allows those other three theoretical assumptions to be fully realized.

How We Use Implied Risk to Trade

Implied Risk readings are used to denote percent deviation from trend. We're looking for issues that have deviated more than one standard deviation from trend, on the assumption that soon the rule of reversion to the mean will force the issue back toward its trend line

But the real value of Implied Risk becomes apparent only when you create compelling comparisons among issues that share similar characteristics. For example, we group together all the broad-based indexes, such as S&P 500 (SPX) and Nasdaq 100 (NDX), to give us a picture of the overall market. Then we group together all the sector indexes, such as Broker/Dealer (XBD) and Housing (HGX), to show us which sectors are rallying and which sectors are falling. Finally, we group issues by thier relation to thier means, so we can tell which ones have moved to lower or upper extremes.

Once we've established the relative strength of the market and its various sectors, we can determine the appropriate trading strategy. The most leveraged trades, such as buying calls or buying puts, are reserved for moves to extremes. Our most basic strategy, the buy/write, is applicable at all times except during a move to the upper extreme.

We look at the market from the top down. Market, sector, stock. We like to buy when all three are more than two standard deviations below thier means, and we like to sell when all three all more than two standard deviations above their means. We call it buying low and selling high.

Three Months Ago

Here's the front page from The DYR Report ($600/yr.) on June 18, 2007:
STRANGE REBOUND
Stronger bonds had the stock market on a strange rebound last week. It’s strange because most indexes and sectors are near Zone 6. There is so much liquidity that pullbacks don’t even reach 90-day neutral. Overall, the market’s action doesn’t match the data. We don’t like it.

We do like gold stocks. The golds gained about one standard deviation last week, and now they’re back to 90-day neutral. If the market does pull back significantly this summer, golds look good.

Here’s how crazy things have gotten. Posco (PKX), the fourth largest steel maker in the world, plans to expand its output next year by 30%. The company says it has orders for steel plates from shipbuilders. Things is, both PKX and the four largest shipbuilders are based in South Korea. Can you say corruption? Lots of risk here.

PKX, like the rest of the steels, has doubled over the past year. The stock remains more than 100% overbought on a 90-day basis, about the same as Steel Producers (STQ). We’re no economists, but shouldn’t the tightening of credit worldwide hurt these guys? Implied volatility (IV) expectations for steel stocks are among the lowest in the business, which makes long-term puts relatively cheap.

Among the sector indexes at the bottom of the DYR short-term rankings, the only one we like even a little is Corporate Bond Fund (LQD). Two other sector indexes are in Zone 2 – DJ Equity REIT (DJR) and Airline (XAL) – but we think they belong there. When there are no good oversold groups to buy from, don’t be buying.

Finally, we noticed that XAL has dropped below 50. When will it be time to buy again? We’re looking for a big sell-off, where maybe a big player says they’ll never own airlines again. It happened to the gold stocks just six years ago. We say buy long-term if XAL hits 25.

ImpliedRiskReport