December 1, 2001

Nearsighted

Is there merit to a short-term trading approach? T

According to Modern Portfolio Theory (MPT), security prices should instantly and accurately reflect inputs of new information. The reason? MPT's assumption that all investors are rational and possess all the information necessary to make informed buy and sell decisions. With such an informed, level playing field, stock prices must certainly reflect the true value of the companies they represent.

There's no doubt that more folks put stock in MPT before the bursting of the Internet Bubble than now. Regardless, most academics believe that the theory explains approximately 95% of stock market behavior. The other 5% represents the sliver of the investment pie that is the domain of short-term traders, those intrepid souls who focus their attention on the inefficiencies that occasionally show themselves in stock prices.

There are some compelling reasons why one might believe that a short-term orientation could offer some promise to investors. For starters, there are a lot fewer short-term traders now than in the past. Individual day traders, for example, are largely an endangered species these days.

Large investors have also snubbed their collective noses at short-term equity trading. The desire to slash expenses has led many Wall Street brokerage firms to shutter their proprietary equity trading desks.

This relative dearth of short-term players raises a tantalizing possibility--namely, that with so few traders exploiting minute market anomalies, such inefficiencies might become more persistent. If this is the case, the remaining short-term traders could have years of outsized returns to look forward to.

Based on my real-time experiment with short-term strategies--a project started two years ago and tracked in my daily column on the investment Web site www.worldlyinvestor.com--there may be some validity to such an approach. But advisors keen on redirecting their efforts to short-term trading should realize that such strategies are a time-consuming pastime fraught with many problems, including excessive trading costs and significant tax consequences for clients.

However, for investors who center their attention on absolute return of their holdings, including tax-exempt entities and citizens of foreign countries who do not face the same egregious taxation on short-term gains as do U.S. citizens, well-researched short-term strategies may indeed have some value (see Figure I).

The Inner workings of a short-term portfolio

The Quant View Portfolio comprises a number of different trades to add diversification and increase returns, including:

1. Event-based trades, triggered by the market's reaction to government reports.

2. Seasonal trades, which typically take positions based on a calendar anomaly. The best known of these is the month-end effect, which attempts to capture the tendency of stock prices to rise at the end of each month. QVP also takes positions around holidays, and in some cases at the middle of each month.

3. Volatility-based trades, which buy and sell short based on movements of the implied volatility of OEX options (also known as the VIX).

The Portfolio does not use leverage, but does trade both long and short. The trades are initiated using SPY (the S&P 500-based exchange traded fund) or the trader-friendly index funds offered by such firms as Rydex and ProFunds.

An Informational Edge

My interest in short-term trading has its basis at my alma mater, the University of North Carolina, home of the well- known "earnings surprise" approach to investing. According to that theory, stocks that shock the Street with better-than-expected earnings tend to put in a repeat performance, which causes exploitable price trends in equity prices. The approach is used by many money managers in the domestic and international markets, but I wanted to put my own original spin on the subject. This led me to focus on the reaction of market indices to releases of economic data, a sort of macro version of earnings surprise studies.

It was not difficult to decide which economic releases to start with. According to a Federal Reserve study, the bulk of the largest 25 daily price moves in the U.S. government bond market occurred on the days of the monthly employment and producer price index reports.

Although earnings releases and government reports share many similarities, there is one significant difference. Unlike earnings reports, which are typically quite accurate, economic data is typically revised afterward, sometimes significantly. In order to eliminate the need to examine the report details, I decided to focus exclusively on the market's reaction to the release. For example, if the market rallies and closes near the day's high on the day of an economic report, the release would be deemed "bullish." An economic report meeting with the opposite reaction would be categorized as "bearish."

I originally expected stock prices to react to earnings surprises and economic releases in the same manner--rising after a good report and falling after a bad one. To my surprise, the opposite was true; on average, bullish government reports were followed by pullbacks in the market, while bearish reports were often followed by rallies.

For the same reason that traditional portfolios are composed of a number of different stocks, the Quant View Portfolio is made up of a mixture of different short-term strategies .

There are two ways to implement the strategies in the portfolio. Since both long and short positions are taken, the securities used must be as easily bought as sold short.

Most traders prefer using the trader-friendly index products offered by such fund families as Rydex and ProFunds. Both offer short funds that rise in value when the market falls, making implementation of the QVP as easy as buying shares of a mutual fund.

Larger traders may opt for exchange traded funds (ETFs) that track the S&P 500 index. The most liquid is the S&P 500 Depositary Receipt, or SPDR (Amex: SPY).

Proof in the Pudding

At first glance, the Quant View Portfolio seems to be performing according to its mandate. The QVP has gained about 19% since its inception, handily beating the 20% loss of its benchmark, the S&P 500 index. Almost half of the gains can be attributed to short positions, with most of those occurring after big rallies on employment report days.

Since its inception, the QVP has produced 60% winning trades; the average winning position is 2.3%, which is one-third larger than the average loser.

Even though the QVP traded in each of its 24 months of existence, it did not have much market exposure because of its short-term orientation; since the average holding period was a scant three days per trade, the Portfolio was in the market an average of one day per week.

Even with these winning attributes, some traders who follow the Worldlyinvestor Web site find the approach wanting. After all, it's great to use the S&P 500 as a benchmark when the index rises year after year. But when bears takes control of Wall Street, there is not much to celebrate when one loses less than the market.

For this reason, a new version of the Quant View Portfolio with an objective of generating an absolute return of 10% to 15% per year, regardless of the direction of the overall stock market, will be launched in January 2002. Some of the differences between the new and the old version of the QVP (which will still be tracked on the site) are as follows:

More opportunistic. With the goal of making money in bull or bear markets, the Enhanced QVP must have the ability to extend its normal short-term holding period if an extended uptrend develops in the market.

A larger menu. Since the previous QVP was pegged to the S&P 500, its strategies were restricted to trading that index. With a focus on generating an absolute return, the Enhanced QVP will trade a variety of highly liquid indices.

More emphasis on cash management. The returns of the Classic QVP do not include the interest earned in money market accounts, which is where the money sits over 85% of the time. The Enhanced QVP, however, will root out the best places to park cash in those periods when the Portfolio is not active in the market.

Although there are many differences between the Classic QVP and Enhanced QVP, there are also similarities. All strategies will be quantitative in nature, with no discretion used, and only indices will be used in implementing the trades.

If you have any questions or comments on the Quant View Portfolio, please drop me an e-mail.

Reprints Discuss this story
This is where the comments go.