The Archaeology of Finance: The Value of Investing in History is an attempt to share with you our knowledge of historically successful investors their methods and approaches. Today we present a study by James Montier and Rui Antunes of Dresdner Kleinwort Wasserstein in London that provides further evidence that value investing works, and that stocks possessing the value characteristics described herein provide attractive returns over long periods of time.
While this conclusion comes as no surprise to us at The Wall Report, it does provide clients with empirical evidence that Benjamin Graham’s and David Dodd’s principles of investing, first described in 1934 in their book, “Security Analysis”, continues to work today, some 74 years later.
Aspects of the investment selection criteria described in Montier and Antune’s “Better Value or The Dean Was Right!” has been incorporated in The Wall Report’s investment screening and decision making process. So while the street wants you to buy the exciting story and high growth name, "the safer bet is to stick with value".
James is one of the smartest analysts and behavioral finance thinkers I have had the pleasure of reading and I recommend his book "Behavioural Finance: A User's Guide" published in 2003 to readers.
“Better Value or The Dean Was Right!”
Montier and Antune demonstrate that using a Graham and Dodd P/E ration can significantly enhance stock selection relative to a simple trailing P/E ratio. Their research is based on the earlier work of Graham and Dodd who argued - in the 1930s - that investors should cyclically adjust P/E ratios by using a moving average of earnings of “not less than five years and preferably seven or ten years”. Montier and Antune demonstrate that using a Graham and Dodd PE can significantly enhance stock selection relative to a simple trailing P/E.
Montier and Antune used the MSCI World index for the period 1980 to 2005 to test the effectiveness of Graham and Dodd P/Es when selecting stocks. At the start of each year decile portfolios were formed and their performance measured over the next 12 months. They excluded stocks with negative P/Es, as well as those stocks with Graham and Dodd PEs below one – the authors believing such stocks likely represent distressed companies and that it was unlikely that any value investor would have taken a risk investing in a stock trading at less than 1x five-year average earnings. If a company did not have a full set of earnings, they computed the P/E based on the maximum amount of data available.
Overall, the authors demonstrate that the potential gains that arise from moving from a simple trailing P/E to a Graham and Dodd P/E are significant – and worth investigating. Table 1 highlights the pattern they uncovered, which is that the value premium (the spread between the low P/E and high P/E deciles) increases monotonically as the time horizon for the calculation of the earnings variable is extended. That is, the increase in the number of years of earnings data used to determine the Graham and Dodd P/E is worth pursuing.
Table 1: Returns to Using G&D P/Es (1980 – 2005)
| Earnings time average (years) | Return from high PE stocks %pa | Return from low PE stocks % pa | Differential %pa. | 1 | 9 | 20 | 11 | 3 | 8 | 23 | 15 | 5 | 5 | 23 | 18 | 7 | 0 | 22 | 22 | 10 | -4 | 21 | 25 |
Source: DrKW Macro Research
The key results from this study are as follows:
• A strategy of buying low one year trailing earnings PE and selling high P/E stocks applied to the MSCI World Index generated an excess return of 11% p.a. over the period of 1980 – 2005, with stocks held for one year and annual rebalancing.
Extending the time horizon for the calculation of earnings could improve the value strategy, the results of which are:
• Using a 5-year moving average of earnings the differential between high and low stocks rose to 18% p.a. on average, with the return from low and high P/E stocks being 23% and 5% respectively.
• Using a 10-year moving average of earnings, the difference between high and low P/E stocks increased to an average 25% p.a., with low and high P/E stocks returning 21% and -4% respectively.
Montier and Antunes then illustrate the difference between value (low P/E) and growth (high P/E) deciles, with value stocks outperforming growth stocks by a greater margin as each additional year of earnings is added to the time horizon for the calculation of earnings. For instance P/E7 outperformance is greater than P/E5, which is itself greater than P/E#3.
The authors also provide the annual return for a long / short portfolio based on 10-year average earnings for the period 1990 to 2004; demonstrating that the return each year from buying the lowest Graham and Dodd P/E decile and shorting the highest Graham and Dodd P/E decile has been positive each and every year between 1990 and 2004, with an average return of 25% per annum. To put this into context, a $10,000 investment in such a strategy in 1980 would have grown to more than $2.6 million in 2005 (excluding rebalancing costs), whereas the MSCI would have returned around $46,500. Thus the long / short portfolio would have grown to be ca. 56 times that of the MSCI over 25 years – not a bad problem to have if you ask me.
The key conclusion one can draw from Montier and Antunes research is that value investing works, and that value investing following Graham and Dodd cyclically adjusted P/E works even better. To put this into context we need to compare the return from a buy and hold strategy for the MSCI, a buy and hold strategy for the MSCI Value Index, to that from using Graham and Dodd PEs for the 1980 - 2005 period. The results are quite extraordinary.
Table 2 illustrates the return for various investment strategies for the 25 year period 1980 to 2005. Specifically, a simple buy and hold strategy for the MSCI would have generated a 6.6% p.a return, compared to the 7.8% p.a. for the MSCI World Value index.
Table 2: Return and risk of various strategies (% p.a.) (1980 – 2005)
| Strategy | Annual Return (CAGR) | Risk (St.Dev) | Return/Risk (Sharpe) | Buy and Hold MSCI | 6.6% | 17.9% | 0.37 | Buy and Hold MSCI | 7.8% | 17.2% | 0.45 | Long-only G&D 10-year | 17.8% | 23.1% | 0.77 | Long/Short G&D 10-year | 22.4% | 16.7% | 1.34 |
Source: DrKW Macro Research
The Graham and Dodd P/E strategies outperformed by a significant margin. The long-only portfolio generated a 17.8% p.a. return, albeit it did have higher risk than the buy and hold MSCI strategies. However, if we equalised the return to account for higher risk, the return on the G&D portfolio would still have been 13.2% p.a., double that of the MSCI.
The star performer was the Graham and Dodd equal long/short portfolio which generated a 22.4% p.a. return over the period 1990-2004 (it also returned 25% p.a. over 1980-2005). Importantly, not only did it deliver stunning outperformance, it did so with less risk than the buy and hold strategies!
This study suggests to me that the study of past financial data - a term I call the archaeology of finance - can yield attractive returns, often with less or equivalent risk to that of investing in the broader market.
I will leave you with two comments from individuals far wiser than I. One is from Confucius the Chinese scholar; the other is from Montier and Antunes, the authors of this study who end their research with the following statement:
“Investors would be well served to remember Ben Graham’s advice on a wide variety of issues including such sagely words as “Buying a neglected and therefore undervalued issue for profit, generally proves to be a protracted and patience-trying experience. And selling (short) a too popular and therefore overvalued issue is apt to be a test not only of one’s courage and stamina but also of the depth of one’s pocketbook.”
which in itself reflects Confucius, who states:
“I for my part am not one of those who have innate knowledge. I am simply one who loves the past and is diligent in investigating it.”
I leave it up to you to decide if you are interested in The Archaeology of Finance – that being the search for and investing in undervalued companies - or otherwise appointing someone that is interested in such an approach to manage your affairs.
Kind Regards
Jonathan Wall
Founder and Editor, The Wall Report.
Posts
