Robert Drach, Trading with 95% Confidence?

A reader asked about Robert Drach’s “Basic Timing” Model Portfolio, initiated on 5/5/95 with the objective of demonstrating that market timing can beat the S&P 500 Index. The self-assessment of this portfolio (as of 8/19/09) reports 399 closed positions with an 89.7% win rate. The average closed position yields a 7.22% gain over 205 calendar days for an annualized return of 12.9%. The cumulative portfolio gain is 129%, compared to 93% for the S&P 500 Index. These cumulative returns “…are reflective as to capital capture and market price of current holdings… They do not include cash dividends, interest earned on cash balances, transaction costs, or anything else.” Robert Drach is publisher of the “Drach Weekly Research Report” (no web site). As explained in an article by Jon Markman, “he scales into stocks only when he believes there is a 95% likelihood of a successful result. …He focuses on buying only from a master list of 80 large stocks [with decent earnings predictability] that hasn’t changed much over the years.” Do Robert Drach’s results demonstrate market timing ability? We can address this question approximately by measuring the correlation of his cash balance (from the Cash Balance Ledger) with stock market returns. Using this cash balance data and contemporaneous S&P 500 Index data for the period 5/5/95-7/31/09 (171 months or about 14 years), we find that:

For a precise analysis of stock picking versus market timing, we should work with asset class (stocks, cash) portfolio weights rather than dollar amounts. However, the effort to reconstruct and mark this portfolio to market over a 14-year period is very large, so we use cash balance as an imperfect proxy for cash portfolio weight.

The following chart shows the average daily portfolio cash balance by month and the end-of-month S&P 500 index over the entire sample period. It is difficult to infer market timing from this chart, but there is some indication that Robert Drach’s model portfolio tends to hold more cash during bull markets than bear markets.

For greater precision, we relate the monthly S&P 500 Index return to average daily portfolio cash balance by month.

The following scatter plot relates the monthly S&P 500 Index return to the Drach model portfolio average daily cash balance during the same month from August 1995 through July 2009. We exclude the first three months of the overall sample period (May 1995-July 1995), because it appears that the cash balance during these months is high due to portfolio start-up. If portfolio timing is good, then the average daily cash balance by month should be systematically lower (higher) when the monthly S&P 500 Index return is stronger (weaker).

The chart shows, however, some tendency for cash to be higher (lower) when the stock market is stronger (weaker). In other words, the portfolio tends to be somewhat wrong-footed with respect to cash-stocks allocation at a monthly frequency.

For a broader perspective, we look at annual data.

The next chart shows the annual return for the S&P 500 Index and the Drach model portfolio average daily cash balance by year during 1995-2009. Calculations for 1995 and 2009 are for partial years based on the 5/5/95 start date for the portfolio and 7/31/09 end date of this analysis. If portfolio timing is good, then the average daily cash balance by year should be systematically lower (higher) when the annual S&P 500 Index return is stronger (weaker).

In general, the portfolio appears to be too much in cash during the good market years (especially 1995-1997 and 2003) and too little in cash during the bad market years (especially 2000-2002 and 2008). Holding losers until they become winners (emphasis on a high percentage of winning trades) may explain the latter periods. When the market goes down, Robert Drach buys and his cash balance goes low. If the market falls further and stays down, he has no winners to sell and his cash balance stays stuck on low.

The following scatter plot built from the same data confirms a conclusion of poor timing at an annual frequency. The Drach model portfolio cash balance tends to be lower (higher) during years when the stock market is weak (strong), giving the portfolio higher exposure to bad market years than good market years.

A policy of holding stocks until they show a gain appears to be at least part of the reason that the Drach model portfolio has a high (~90%) win rate. However, 23 of his 24 Current Open Positions, with buy dates extending back as far as August 2000, are under water (many substantially).

While Robert Drach reports an apparently attractive annualized return on model portfolio closed trades of 12.9%, this return on invested capital does not include:

1. The relatively much lower return on funds allocated to cash.
2. The negative net return on the 24 open portfolio positions.

The compound annual growth rate for Robert Drach’s Basic Timing Model Portfolio during 5/5/95-8/19/09, assuming that dividends and return on cash offset trading frictions, is about 6.0%. The comparable annualized return for the S&P 500 Index over this period is about 4.7%, and the comparable annualized return for buying and holding the investable S&P Depository Receipts (SPY) with reinvestment of dividends is about 6.3%. It may be that Robert Drach’s stock picking tends to compensate for the adverse general market timing found above.

With an average holding period of 205 calendar days, taxes on the capital gains in Robert Drach’s model portfolio would substantially dent net returns.

In summary, evidence from simple tests using an imperfect metric indicates that Robert Drach’s approach to timing the broad U.S. stock market depresses rather than enhances portfolio returns. The aggregate performance of his “Basic Timing” Model Portfolio is roughly the same as that of buying and holding SPY.

See Guru Grades for links to assessments of the stock market forecasting ability of other stock market experts.

You May Also Enjoy...

• Robert Taylor’s Peculiar Definitions of Up and Down
• Robert Prechter: 100-Year Bear?
• Does Consumer Confidence Predict Stock Market Returns?
• Robert McHugh Objects to Our Evaluation of His Commentaries
• Robert McHugh: Caution Is Warranted?