Can Volatility Be Your Friend?

April 01, 2000

The author presents alternative ways to add alpha by exploiting observable extremes of behavior of market participants. He describes several overlay methodologies designed to do this using puts and calls that indexing-oriented plan administrators can implement at limited risk. He reports improved overall plan performance.


The goal for active managers is simple to state: add alpha - positive alpha, outperformance resulting from managers' decisions - but this goal has been difficult to achieve. As many recent articles have highlighted, the past decade has been especially challenging for active managers benchmarked to the S&P500. Both fundamental and quantitative investment managers have found their current methodologies have not produced the desired results. Investment managers are now beginning to believe what many academics have been saying for years - the large-cap universe is simply too efficient to consistently add alpha. Logically, it is easy to deduce that the large-cap arena is so efficient because there are simply too many investment managers tracking the same securities. For large-cap investment managers, uncovering exploitable information to add alpha is the equivalent of finding the proverbial needle in a haystack. These inefficiencies have caused many plan administrators to index their large-cap allocation to the S&P500.

Since many plan administrators have incentives to place their assets in the hands of outperforming managers, indexing assets presents a conflict, and abandoning the possibility of adding alpha can be a difficult choice. This has led many plan administrators to seek alternative non-traditional methods to enhance the S&P500.

Most of these alternatives depend on skillful futures trading, credit and duration bets, or even the use of so-called 'market-neutral hedge funds.' While these methods have added excess returns, the benefits may have come at the expense of higher volatility or the acceptance by the manager of unforeseen systematic risks.



  S&P Return Index Market Value S&P + 1% Index Market Value
    100.00 $100,000,000   100.00 $100,000,000
1988 16.34% 116.34 $116,344,900 17.34% 117.34 $117,344,900
1989 31.22% 152.67 $152,665,218 32.22% 155.15 $155,150,845
1990 -3.13% 147.89 $147,889,240 -2.13% 151.85 $151,848,615
1991 30.00% 192.26 $192,259,265 31.00% 198.93 $198,925,026
1992 7.43% 206.55 $206,551,050 8.43% 215.70 $215,701,567
1993 9.92% 227.05 $227,048,556 10.92% 239.26 $239,264,159
1994 1.28% 229.96 $229,964,314 2.28% 244.73 $244,729,431
1995 37.12% 315.32 $315,315,569 38.12% 338.01 $338,008,053
1996 22.68% 386.84 $386,835,761 23.68% 418.06 $418,055,458
1997 33.10% 514.89 $514,892,324 34.10% 560.63 $560,627,420
1998 28.34% 660.80 $660,799,937 29.34% 725.10 $725,101,489
1999 21.04% 799.83 $799,832,244 22.04% 884.91 $884,913,857
Chart courtesy of Rampart Investment Management Co., Inc.


There is another approach to adding alpha to existing portfolios without using futures and hedge funds. In fact, overlay strategies using options can be implemented by utilizing existing plan assets to add alpha to both active and passive portfolios. This strategy can also be implemented without adding additional capital. Furthermore, alpha can be added without large increases in volatility and without large increases in the variability of returns versus the benchmarks.

Plan administrators are sitting on billions in assets. These assets can be made more productive by pledging small portions of the asset base as collateral to implement conservative overlay strategies to add 100-200bp. While many may discard the idea of attempting to add only 100bp, that 100 bp could be extremely valuable. Based on active managers' past performance, adding 100bp annually to the S&P500 would make any plan administrator a hero. Just how meaningful would 100bp be, on a $100,000,000 portfolio? As shown in the table above, the incremental value would be nearly $85 million in the period 1988-99.


This strategy involves substantially less trading than traditional methods. It simply takes advantage of behavioral biases of the stock market. As documented in numerous psychological and financial journals, human emotion often causes flaws in judgment and irrational responses. There are rare times, usually only a handful each year, when these flawed responses lead to temporary extremes in market prices that are unsustainable, and also identifiable. Knowing how to respond and when to react to irrational responses of market participants widens the opportunity to generate alpha. Rampart Investment Management has created and uses multi-factor models intended to identify these opportune times.


One of these models, (Rampart's short-term reversal model), is used to sell option premium, which often becomes quite large during such periods. The method used is to apply short-term out-of-the-money S&P 500 index spreads. (Selling call spreads during periods of excessive optimism and selling put spreads during periods of excessive fear).

A simple example to illustrate: Suppose the S&P 500 is trading at 1400. You sell a call option on the S&P500 exercisable at 1405 ('out of the money' because it is above the market price) and simultaneously buy a call at the next higher exercise (or 'strike') price, 1410, to protect yourself. Both command premiums but the long call with the higher strike price will be cheaper since it is less likely to be exercised. Suppose the difference is therefore $2.50 per call. If you are indeed near the market peak forecast by the volatility measures, it is likely prices will fall away or at least not rise enough to trigger exercise of the option you sold, and you will be able to pocket that difference (less transaction costs) when the calls expire. If you are wrong, your loss is limited to the difference between the calls. Obviously, for this strategy to be profitable, it is necessary to be right considerably more often than wrong.

The measure we have used, described briefly below, seems to meet that test. These credit spreads are sold during turning periods of measurably extreme market behavior giving these trades high success rates. (And by using index spreads, overlay investment managers and plan administrators can precisely see - and measure - their maximum risk and reward exposure at all times).

Market regulations, specifically price limits and coordinated circuit barriers, can add to the short-term reversal model's success. These artificial volatility barriers, first implemented in April 1988, provide market participants time to acknowledge and correct their irrational responses.


Rampart's short-term reversal model can be viewed as a market sentiment measure, recording the peaks and valleys of market consensus. To illustrate how sentiment can be acted upon, two commonly used sentiment measures are examined from 7/1/98 to 12/31/98. VIX, the CBOE's implied volatility index based on OEX index options, reflects the market's fear factor.

Traditionally, the implied volatility of index options has usually been slightly higher than the historic volatility, meaning the premium paid for options is typically a little more than the actual volatility of the S&P500 would indicate, using the traditional Black-Scholes pricing model. However, during times of market panic, this implied volatility increases dramatically. During these times of panic, investors look to protect their portfolios and become aggressive buyers of put options, consequently bidding up volatility - that is, widening the premiums, often sharply. More often than not, the VIX has signaled important turning points in the market. Multi-year VIX highs of 48.3 and 48.5 on 8/31/98 and 10/8/98, for exam ple, corresponded with respective lows on the S&P of 957.28 and 959.44.

Similar to VIX readings, Market Vane, a poll to determine bullishness of leading market advi sors, can signal opportune times to sell option pre mium, that is, take advantage of extraordinarily bullish market sentiment. As seen in the previous chart, extreme sentiment readings often signal key turning points in the market. Historically, readings of >70% bullishness are usually followed by dis appointing market returns and readings of <30% bullishness are usually followed by strong market returns. The sentiment readings below 30 on 8/31/98 and 10/8/98 coincide with VIX signals of a bottoming market.


The batting average of Rampart's short-term reversal model used to sell index spreads has shown it to have predictive power during the period 1988-1999. Results from this one model shown below clearly indicate these trades provide information content.

While the S&P 500's historical 29-day average return has been 1.77%, the average 29-day return on the underlying S&P 500 from the signal to sell out-of-the-money index call spreads has only been .33%. And while the S&Phistorical 25-day average return has been 1.53%, the average 25-day return from the signal to sell out-of-the-money put spreads was a remarkable 5.17%. These results do much to confirm that Rampart's model has successfully identified important market turning points.

Since 1988, backtesting shows the statistical probability of the market closing at a level to cause a loss on the model's trades has been 26%. In an efficient market, this would assume the model's batting average would only be 74%. The short-term reversal model, which takes advantage of market participants' behavioral biases, clearly demonstrates market inefficiencies with its 97% win ratio. With (typically) less than six trades per year, one doesn't need to be an active trader to add alpha, but one does have to be a smart trader. A carefully executed and conservative option overlay strategy employing this phenomenon could create value for plan administrators who want to augment returns on their existing portfolios.

  Number   Average Avg. Market Avg.Holding
Results from of Win Holding Move from Period Move
1988-1999 Trades Ratio Period Trade Signal Of S&P 500
Call Spreads 36 95% 29 0.33% 1.77%
Put Spreads 20 100% 25 5.17% 1.53%
Total / Average 56 97%      
Chart courtesy of Rampart Investment Management Co., Inc.


Just as sentiment models can be used to initiate the sale of index spreads, they can also be used simply to purchase put options at an indicated market peak and call options at an indicated market bottom. Similar to the sale of index spreads, option purchases also have defined risk, which is limited to the cost of the option. However, the return differential of the two strategies can be dramatically different. The use of longer-term option purchases funded by the successful capture of option proceeds from previous index spreads can be used to enhance the risk-reward scenario of an overlay program and provide further opportunities to add alpha.


There are many other strategies that can be used to exploit the behavioral biases of market participants. But unlike the previously mentioned strategies, these strategies are only for plan administrators who are willing to let their overlay managers go beyond predefined-risk trades, to where even greater alpha opportunities lie, along with concomitant risks.

One is exploiting what is called dispersion. Having employed option strategies for more than 16 years, Rampart maintains what it believes to be one of the largest option databases on the Street. Just as the VIX measures the implied volatility of the OEX 100, Rampart's Volatility Index measures the underlying volatility of each individual security in the OEX 100. By carefully monitoring the implied volatility of all securities in the index, Rampart can take advantage of discrepancies in implied volatility between the index and the securities that make up the index. This involves the sale or purchase of index options and the opposite transaction in options on the underlying index components. In large market declines (greater than 10%) the spread between the implied volatility of the OEX 100 options and the underlying index components becomes extremely narrow. As mentioned before, during these market declines, investors' buying puts drive up the implied volatility of OEX options dramatically. However, the rise in underlying index components' implied volatility is often dramatically less than that of the index itself. Since the OEX index is less volatile than the securities that comprise the index, the market is providing a unique opportunity to profit. For example, during the October 1999 low, the implied volatility spread between the index and its components was less than 5 volatility points. It is easy to conclude that the spread between the two indexes is often a signal of market turning points. Areturn to more normal relationships can then produce a profitable result.

Another way to add alpha by exploiting the behavioral biases of market participants is by trading volatility. This is accomplished by making swap agreements with Wall Street broker-dealers in which the two sides agree in advance that one will pay the other according to whether actual volatility in the marketplace proves to be above or below an agreed-on standard (broker-dealers often have ample means of hedging themselves against the risk they take on). Just as extreme levels in the VIX can be used as signals to initiate option trades, these extreme readings can also be used to initiate volatility trades. When volatility is at historically high levels it can be sold and when volatility is at historical lows levels it can be purchased. Volatility trades remove the need for having a directional bias on future market movements. Instead, your trade payoff is based on whether actual volatility is higher or lower in the future. As can be seen from the VIX chart (Figure 3), the market offers excellent opportunities to take advantage of these extremes. The contracts can be very technical in nature, and there are a lot of nuances to these sorts of deals, so caution is recommended to any plan manager when setting these swaps up.



Find your next ETF