A portfolio consisting of 15 stocks which are selected using Benjamin Graham’s Stock Selection Criteria for the Defensive Investor (Graham, B., The Intelligent Investor, Chapter 14) delivers market-beating average annual returns with lower volatility than the market.
The following figure describes a simulation of the portfolio’s performance over a period of 18 years, from June 30th, 1999 to June 30th, 2017.
The Portfolio returned 11.48% a year on average, compared to 5.13% a year for the S&P 500 during the same period. Both the portfolio’s and the benchmark’s results assume reinvestment of dividends. Over the 18-years period we tested, the total portfolio’s returns were four times higher than the benchmark’s return, 607% vs. 146% for the S&P 500.
The volatility of the portfolio, as measured by standard deviation, was 11.64% vs. 14.62% for the S&P 500. Max drawdown for the period was 31.92% vs. 55.19% for the S&P. Thus, the portfolio’s volatility is significantly lower than the market’s. Higher returns and lower volatility were recorded for the period of the last three years as well.
Asset Turnover stands at less than 20%. Holding a 15-stocks portfolio, it would mean that upon every yearly rebalance, between two and three stocks on average need to be sold and replaced. The low turnover results in low transaction costs and a very low tax liability (for taxed accounts). Maintaining Graham’s defensive portfolio should not take more than an hour, once a year.
A notable attribute of Graham’s portfolio is its low correlation to the S&P 500. Let’s first explain what correlation means. A correlation of 1 between the portfolio and its benchmark would suggest that the portfolio behaves in unison with the S&P 500. The lower the correlation figure, the less related their price behavior is – The portfolio may appreciate when the S&P depreciate or depreciate when the S&P 500 is appreciating. A correlation of zero would imply that there is no statistical relationship between the portfolio and its benchmark.
Graham’s portfolio correlation with the S&P 500 for the full 18-year period was 0.59, a low figure compared to other quantitative methods we have designed or researched. The value for the last three years was 0.36, a meager figure, which indicated the low correlation between the portfolio and the S&P 500 in recent years.
The following figure presents the drawdown (how deep have been its declines from peak levels) and the percentage of cash invested:
Most of the times, the portfolio have dropped less than the S&P 500. Nevertheless, there were several occasions where the portfolio has fallen significantly more than the market and underperformed for several months. Those periods were November 1999 to November 2000, July 2016 to March 2017, and a few shorter periods between 2013 and 2015. Holding on to Graham’s Defensive Investor strategy would have been challenging during those times. Let’s examine the nature of this challenge in more depth. The following graph zooms in into 1999-2001.
Imagine that you are an investor starting Graham’s portfolio in July 1999. You would see your portfolio in a constant decline through April 2000, while the S&P 500 meddling along sideways during that period (and internet stocks booming). The paper losses on Graham’s portfolio would reach -30% by March 2000, full nine months after starting the portfolio, while the S&P 500 returns remain positive. I would imagine that many investors would lose faith in Graham’s strategy long before March 2000, condemning it to have “stopped working” and its advocates being charlatans. They may say that “times have changed” and Graham’s simplistic strategies do not work well anymore. They may refer to the new normal and call to abandon the old analysis & valuation techniques. If you were to hold on to the portfolio through its decline and recovery, you would see its losses turn into over-performance before the turn of the year 2000. Let’s look at another challenging period:
For most of the 2-year period of 2013-2014, the portfolio lagged the S&P 500. Often, it had more substantial drawdowns.
The takeaway from these two examples is that the merit of a strategy cannot be judged during a short period of several months, or even a year or two. An investor must develop conviction and hold on to a strategy for the longer term. Experts say, and we concur, that an investor with an investing horizon of fewer than five years should avoid stocks altogether.
Let’s examine yearly returns and monthly returns for the last 12 months:
The portfolio lagged the market during 1999, 2003, 2009, 2010, 2012, 2013, 2016 and 2017. Except for 1999, in every year that it lagged the market, it still made positive returns. On a monthly basis, the portfolio lagged the market for half of the time and exceeded it during the other half.
Those results are NOT an indication of excessive risk nor volatility. Rather, it is an indication that the correlation to the S&P 500 is relatively low. Investing in Graham’s portfolio is not riskier or worse than investing in an S&P 500 ETF. On the contrary, Graham’s portfolio is better in every respect than the S&P 500 – it delivers higher returns with lower volatility. It just behaves DIFFERENT than the S&P500 on a month to month basis and on a year to year basis. Holding it for a long enough period, though, reveals its merits.
Lastly, we performed a rolling test. Our rolling test was conducted by creating a new portfolio every month since July 1999 and letting it run for one year. The first portfolio was set up in July 1999 and ran for a year. The second portfolio was created in August 1999. The process is repeated for September 1999, October 1999 and the consecutive months, up to the period June 2016 – June 2017. For every such test, we recorded the returns and Sharpe ratios and compared them to the S&P 500 portfolios of the same duration. We repeated the process for test durations of two years, three years and five years. The results for one year holding period are as follows:
The figure shows the Excess returns of the portfolio over the S&P 500. Positive values indicate that Graham’s portfolio’s returns were higher than the S&P 500’s, and negative value indicates lower returns than the market. One can see that even for a short holding period of one year, Graham’s portfolio returns were higher than the benchmark’s for most of the time. The Sharpe excess in the following figure measures risk-adjusted returns. Sharpe is higher when the total returns are higher or when the volatility (as measured by standard deviation) is lower.
Rolling tests with 2-year holding periods:
The longer the holding period, the higher the chances of beating the market, both by absolute returns and by risk-adjusted returns. Rolling tests with 3-year holding periods:
Rolling tests with 5-year holding periods:
The results show that except for portfolios created during 2012, whenever Graham’s portfolio was created and held for at least five years, it had gained excess Sharpe, meaning that it had delivered better risk-adjusted returns. For portfolios ending in 2014 and on, absolute returns for Graham’s portfolio were sometimes lower. Nevertheless, they were gained with less volatility, hence the positive excess Sharpe.
The relative degradation in performance during the last several years is noticeable. Has Graham’s defensive strategy lost its magic touch?
We do not think so. During the last few years (as of 2017) the S&P 500 was soaring, with growth stocks leading the pack. FANG (Facebook, Amazon, Netflix, Google) stocks are now the name of the game. While growth starlets are booming, Value is left behind. This phenomenon has happened in the past. It will happen again. Learning stock market history, we know that Value always returns to shine sooner or later. History had taught us that the best times to buy Value is when its underperforming growth. We’re witnessing such a period right now, at the end of 2017. Now may be one of the better times to start Graham’s defensive portfolio.
To summarize, Graham’s portfolio has outperformed the general market for over 70 years since it was published in 1949. Albeit being known to practically all professional investors and most individual investors, its merits have not waned. Our specific implementation of Graham’s defensive strategy was tested in simulation through a recent 18-years period. It shows that a portfolio consisting of 15 stocks beats the market while doing so with lower volatility and with a very low asset turnover. We are inspired by Ben Graham’s insight and fortitude, recommending a strategy so simple to implement, and yet so powerful. The fact that after 70 years it still works so well, never stops to amaze us. It gives us the confidence to invest in Value strategies and encourages us to seek simplicity in our endeavors.