Tricks of the Portfolio Trade

Relative Strength by Numbers

By Matt Blackman and Mike Carr, CMT

Relative Strength, or RS, has become a popular technique to compare the performance of one stock or mutual fund against the overall market. Everyone from individual traders and investors with a few thousand dollars in their accounts to portfolio and fund managers with billions of dollars under management use it.

Correction: Many try to use it, but few ever manage to do so effectively.

“A large number of academic studies support the validity of RS,” notes Darrell Jobman, Senior Market Analyst for www.TradingEducation.com, a free trader’s resource. “It’s a method that’s been used for years to find the best performers over the recent past that have the highest probability of continuing to perform well in the near future for one simple reason: It works. However, the challenge for the vast majority of traders is using RS to find trading candidates and trade them effectively as part of a comprehensive trading methodology.”

Relative Strength – What it is… and isn’t

A common mistake newer traders make is confusing relative strength with Welles Wilder’s Relative Strength Index or RSI, which is an index that compares the current strength of a stock to its own historic performance to determine when it is overbought or oversold. Although similar in name, the techniques have very different applications.

Relative strength is relatively easy to compute by dividing the price of a stock by that of another stock or index. For example, by dividing the price of IBM by that of the S&P 500 Index, the trader is able to measure IBM’s performance against the broader market quickly. A rising line indicates that IBM is outperforming; a declining line indicates underperformance.  In a falling market, if IBM declines less than the market, its RS line rises, indicating positive relative strength. Conversely, if IBM is appreciating more slowly than the S&P 500, the RS line will fall, meaning IBM is underperforming the market (see Figure 1).

Generally, RS is determined for each stock in a universe, and the ratios are sorted into percentiles. A percentile ranking allows an individual stock to be compared to all other stocks in the investment universe. This is the basis of the RS ranking popularized by Investor’s Business Daily. In most scales, 99 is the highest percentile and includes the strongest performers; 1 is the lowest percentile and contains the weakest stocks.

Perhaps the earliest paper supporting this investment approach was published in 1967, when Robert Levy’s “Relative strength as a criterion for investment selection” was published in the Journal of Finance. More recent work, especially that of Narishimhan Jegadeesh and Sheridan Titman, has confirmed the persistence of RS over certain time horizons.

Although it is common practice to apply RS to individual stocks, it can also be applied to mutual funds, exchange-traded funds (ETFs), futures or any group of tradable securities. To attain a diversified portfolio, many investors look to mutual funds or ETFs, which are  especially suitable for RS analysis. Traders can buy the strongest security and sell it when it weakens, investing the sales proceeds into the strongest security at that time.

Here is a program that makes the process of finding the best relative strength candidates and trading them a whole lot easier.

Manufacturer – Harvest Advisors, Inc., PO Box 2375, Bigfork, MT United States, 59911
Phone – 1-877-606-6253
Web Site – http://www.monoclesystems.com
Version Tested – Monocle II Software Version 2.50.
Cost – Software system: $325. Data licensing fee (mutual fund/stock database updating): $75/calendar quarter. 30-day free trail available.

Recommended System Requirements:
·    Minimum 486-class IBM-compatible computer. (No Macintosh version available)
·    Windows Version 3.x or later
·    9600 Baud Modem or higher (may not work with satellite ISPs due to signal delays)
·    4 MB RAM (8 MB recommended)
·    One high-density floppy disk drive and one hard disk drive
·    Approximately 20 MB of available hard disk space.
·    Math coprocessor highly recommended
·    800 x 600 or greater resolution highly recommended

Getting Up and Running

Installation from the CD took only a few moments, and the initial data update was accomplished quickly using the included Monocle Downloader program. In less than 15 minutes, we had the program running:


Figure 1 - Chart showing the Monocle II screen with relative strength of IBM compared to the S&P 500 Index. Chart provided by Monocle II.

The Help icon opens an easily understandable, fully indexed user manual. One option is to read the manual onscreen, as you would read a book. Skimming most of the text will give advanced users all they need to know to get started. Those less experienced in using investment software packages will want to read most of the manual, setting it up with a split screen, so that you can read while clicking through the software.

To effectively test the capabilities, we decided to develop a trading strategy for an employer-sponsored pension plan. The employer offers eight funds for investing, and switches are allowed after seven days. The goal is to find a system that doubles the performance of the S&P 500 with half the risk. For this example, we define portfolio risk as account equity drawdown.

Developing a strategy

Clicking on Analysis and reviewing available funds, we found that several of the plan choices were not included in the database. Monocle’s default download includes data on 2,500 funds. Using the Help function, the user learns that more than 9,000 fund symbols are available for downloading, along with any stock symbol to be added to a portfolio. Opening the Downloader application and following the instructions, we were able to add the necessary symbols and download complete histories for each.

Returning to the Monocle application, we continued with Analysis and Add a group, which was a family of funds in our case.
After that, we click on Dynamic, Add and begin developing our system. Each option on the screen is explained clearly in the help manual. We name our strategy “Pension Plan” and accept the default values in the upper half of the screen. We add the “Pension Plan” group and begin testing systems, starting with “Alpha.”

Within Monocle, four basic strategies are available, and each can be adjusted for risk by incorporating standard deviation into the calculation: 

• Alpha is a statistic defined in Modern Portfolio Theory as a measure of the risk-adjusted return of a mutual fund or stock compared to the overall market. The calculation of Alpha is rather complex, using the least-squares regression of the fund or stock’s return against the return of the market, usually measured by an index such as the S&P 500. Alpha measures the past outperformance or underperformance of the fund or stock – in effect, factoring out the market return.
  

• Alpha / SD divides the Alpha statistic by the standard deviation, a commonly accepted risk parameter. Standard deviation is used to measure the variability of return by comparing the actual return to the average return over a specified time period. In finance theory, it is usually assumed that a larger variability of returns means greater risk was required to achieve the return. Incorporating standard deviation into the calculation accounts for this risk and should lessen portfolio drawdowns.

• Return simply measures the total change in price of the fund or stock over a specified timeframe. This calculation identifies the best performers and is one way to implement the familiar Wall Street adage that “the trend is your friend.”

• Return / SD incorporates risk into the calculation.

• Relative Strength Momentum measures the difference between a fast-moving average of RS and a slow-moving average of RS, mathematically defining the acceleration of RS. The greater the acceleration, the higher the RS Momentum value. The software developers point out, “This statistic is highly effective for locating funds with the greatest potential for future short-term performance.”

• Relative Strength Momentum / SD again accounts for risk by adding standard deviation to the calculation.

• Multi-Period Return evaluates the performance of a fund or stock over multiple time periods. By specifying up to four different return periods with variable weightings for each period, you can emphasize or deemphasize more recent performance. If stock returns are mean reverting, as some studies conclude, applying lower weightings to more recent performance should help to eliminate extended funds or stocks and provide less volatile portfolio performance. On the other hand, if a trader wants to emphasize funds or stocks breaking out from RS basing patterns, they can overweight recent performance.
  

As an example, applying the following weightings should identify strong recent performance:
    1 week return        50%    
    4 week return        20%    
    6 month return        15%    
    9 month return        15%                
The Multi-Period Return option summarizes this calculation into a single value, which is then used to rank the fund or stock.
·    Multi-Period Return / SD evaluates risk associated with this measure of RS.

Running through each strategy, Return/SD offers the best return with the least risk. We then optimize our system using the embedded software features and find the best parameters, settling on a 10-day lookback period.

Other options allow the user to select how many securities to hold at any one time, define a minimum holding period and establish sell criteria. Often overlooked, but singularly important, is the sell decision. When should a trader take profits or cut losses? Within Monocle, there are three settings to define when to sell:

1.    If the trade is wrong, an initial stop loss preserves capital. Monocle allows users to enter this as a percent of allocated capital, for example, you can sell if the price declines by more than 5% from your buy point.
2.    Sometimes a security may continue rising, but another fund or stock may be even stronger. Given Monocle’s ranking system, this will lower the rank of the security being held. Levy defined this as a “cast off” decision in his classic paper. Casting off a fund or stock that is beginning to show signs of slower price growth in favor of a stronger security is the best outcome to any trade. In this example, we find that buying the single strongest fund and selling when it declines to below the third strongest increases profitability.
3.    A shortcoming in RS strategies is that general stock market declines tend to punish the strongest, most volatile stocks the most. Defining sell criteria based upon the amount a security’s price declines from the peak price in a trade helps lock in maximum profits. As an example, Monocle allows us to sell a fund that is 8% off its highest price since purchase.

Within two hours, we were able to learn how to use the software and finalize a trading strategy that has a strong logical basis and demonstrates superior performance in back-testing. The results are noteworthy. Beginning in January 1989 through December 2006, this strategy would have averaged a return of 26.2% per year, compared to a 9.4% annualized return on the S&P 500. Equity drawdown with this strategy was 22.5% versus a worst drawdown of nearly 51% in the S&P 500. The strategy required an average of less than one trade per month.

Impressive results but another built-in software capability has the potential to allow us to even improve upon this performance. The equity curve can be charted along with a moving average of the equity curve (see Figure 2).

 
Figure 2 - Chart showing equity curve for the strategy, with a 50-period moving average. Chart provided by Monocle II.

By switching to cash when the equity curve declines below the moving average and only taking signals when the equity curve is above the moving average, we should be able to limit account equity drawdowns even further. Users of Monocle will want to explore this concept in greater detail, given the potential it offers to preserve capital in declining markets.

Conclusion

“Since the trader’s objective is to buy stocks that are outperforming in an uptrending market and sell stocks that are underperforming in a down market, relative strength is an extremely useful tool for those who know how to use it properly,” says trader Jim Wyckoff of www.traderblog.com

This is where Monocle comes in. For the average investor, trading relative strength strategies has largely been a matter of faith. Commonly available trading software does not usually include the ability to backtest portfolios of securities. Monocle II accomplishes this task while incorporating a broad range of RS analysis techniques. The strategy testing and optimization features of this package should enable small investors to move from faith-based systems to soundly tested investment strategies.

Monocle is an affordable, yet powerful, software package that can be used by small individual and large fund managers alike. In addition to providing RS portfolio backtesting capabilities, it can be used as a stand-alone charting package for those who use relative strength as their primary analysis method. The program also comes with more than a dozen common technical pre-programmed indicators as well as additional indicators. The user can also program additional strategies.

A limitation of the software is its specialized focus. Monocle is designed for trading relative strength portfolios. Those seeking a robust backtesting platform for use in trading other techniques should look elsewhere. Those seeking fundamental or technical screening packages will not benefit from this software.

The developers of Monocle have chosen to trade relative strength, and they do that one thing very well. It is important to note that Monocle II is unique – most certainly in its price category – in offering portfolio backtesting capabilities together with its ability to perform relative strength analysis.

To review this software, we tackled the common problem of allocating pension fund assets. RS offers a simple solution to this problem. We could have created a sector rotation strategy using ETFs or developed an ETF-based international investment allocation portfolio just as easily. Given the increased availability of ETFs and the capabilities of Monocle software, small investors can become as sophisticated as hedge funds in deciding how to diversify.

Traders have long been aware of the value of RS, and they now have the ability to incorporate this concept into their trading. Thanks to its affordability and ease of use, Monocle brings individual traders one step closer to enjoying the powerful tools previously available only to the big boys (and girls).