Monday, 4 November 2019

An Analysis of “Graham’s Net-Nets: Outdated or Outstanding?”


Introduction

In an earlier post we analyzed the prominent and often cited study on “net nets” conducted by Henry R. Oppenheimer from the Financial Analysts Journal (1986). In this post we analyze the article “Graham’s Net-Nets: Outdated or Outstanding?” by James Montier. The objective of the article was to examine the performance of securities that were trading at no more than two-thirds of their “net current assets” during the 23 year period from 1985-2007 globally and regionally (namely, in the US, Europe and Japan[i]).
Our objective was to analyse the article itself; determine its reliability, draw our own conclusions and glean, if any, actionable advice for the practitioner of the “net net” method of investing.

Summary

We simply replicate the summary provided by the author for ease of reference before conducting our own analysis.
Value Investing: Tools and Techniques for Intelligent Investment, Chapter 22, p. 230/231:

Parameters

·         Title: Graham’s Net-Nets: Outdated or Outstanding
·         Author: James Montier
·         Source: Value Investing: Tools and Techniques for Intelligent Investment, Chapter 22
·         Publication Year: 2009 (also 2008 Mind Matters 30 September 2008 by The Société Générale Group)
·         Data source: unspecified
·         Period studied: 1985-2007
·         Years in study: 23
·         Markets studied: US, Japan, Europe
·         Exchanges: Unspecified

Key Methodology

·         Valuation metric: “…current assets minus its total liabilities…”
The author did not expressly state whether preference shares were also deducted from Current Assets; however, we assume the formula used was as specified by Graham:
i.e. Net Current Asset Value per share = (Current Assets – (Total Liabilities + Preferred Shares))/Common Shares Outstanding
“Of course, Graham wasn’t contented with just buying firms trading on prices less than net current asset value. He required an even greater margin of safety. He would exhort buying stocks with prices of less than two-thirds of net current asset value (further increasing his margin of safety). This is the definition we operationalize below.” i.e. the study examines the returns to stocks trading at less than 2/3 of net current asset value.
·         Weighting: “equally weighted”
·         Purchase/rebalance date: unspecified
·         Holding/rebalancing period: unspecified

Reliability

While there are numerous biases/errors that can be made when conducting studies/back tests, below we have analysed those we deem most likely to impact a study of this nature:
1.       Survivorship bias
Given the data source was not specified and no mention was made of controlling for survivorship bias we have no way of knowing if the sample studied may have suffered from survivorship bias.
2.       Look Ahead bias
No mention was made with regard to the portfolio formation date, so we do not know if the back test suffered from Look Ahead bias, or not.
3.       Time period bias
The study spans 23 years and we classify this as a “more reliable” period.
For reference:
·         < 10 years; inadequate/unreliable
·         11 to 20 years; somewhat reliable
·         > 20 years; more reliable
·         > 40 years; most reliable
4.       Human error
Human error is always a possibility; however, we have little detail on the back-testing protocol implemented so cannot comment on the increased or decreased likelihood of human error entering the testing process.
5.       Journal rating/credibility[ii]
Given this was essentially an article we cannot assign it with the additional credibility afforded to publications appearing in well renowned finance journals.
Reliability Assessment: Given the absence of detail pertaining to the methodology and data source used to undertake the back test we deem the article to possess a low level of reliability. In addition, further analysis (refer below) causes us to question the reliability of the results generally.

Results and Analysis

“An equally weighted basket of net-nets generated an average return above 35% p.a. versus a market return of 17% p.a.”

“Not only does a net-net strategy work at the global level, but it also works within regions (albeit to varying degrees). For instance, net-nets outperformed the market by 18%, 15%, and 6% in the USA, Japan and Europe, respectively.”



We summarise the results in the table below (some of which were gleaned from Figure 22.2):


Global
USA
Japan
Europe
Annual Return (“p.a.”)
35%
42%
21%
17%
Market
17%
24%
6%
11%
Outperformance (absolute)
18%
18%
15%
6%
Outperformance (relative)
51%
75%
250%
55%

The author specified that the returns were “p.a.” i.e. per annum. It is not clear, to us, if “p.a.” represents the arithmetic mean or geometric mean i.e. the compound annual growth rate. This, in our opinion is of critical importance, especially given the length of the study and given it incorporated Japan, a market which experienced, to our knowledge, the greatest stock market bubble in recorded history during the period examined. But how could we find out?

Interestingly the author refers to the Oppenheimer study with which we are intimately familiar.

“In 1986, Henry Oppenheimer published a paper in the Financial Analysts Journal examining the returns on buying stocks at or below 66% of their net current asset value during the period 1970–1983. The holding period was one year. Over its life, the portfolio contained a minimum of 18 stocks and a maximum of 89 stocks. The mean return from the strategy was 29% p.aagainst a market return of 11.5% p.a.

Table V from the Oppenheimer study is reproduced below for ease of reference:



Using the above information, we can deduce the following with reasonable confidence:

The author derived “29%” by multiplying the monthly portfolio return of 2.45% by 12 i.e. 2.45% * 12 = 29.4% i.e. “29%”.

The author derived “11.5%” by multiplying the monthly market return of 0.96% by 12 i.e. 0.96% * 12 = 11.52% i.e. “11.5%”.

We know the compound annual growth rate in the Oppenheimer study for the full sample period was 28.2% (refer table IV, not reproduced here). Compounding 2.45% for 12 months yields 33.7% ((1+ 2.45%)^12-1). Therefore, we determine that the results in table V of the Oppenheimer study represent the arithmetic mean and not the geometric mean referred to in Table IV. Hence, based on the comparative statistics referenced by the author, we deduce, with a strong likelihood, that the author has calculated the arithmetic mean return.

The implication of calculating the arithmetic mean return vs the geometric mean return is stark. The arithmetic mean would, on the balance of probability, have materially overstated returns. Of particular concern are the returns stipulated for Japan which experienced, to our knowledge, the largest equity market bubble in recorded history in the late 1980’s which formed part of the period studied (i.e. 1985 to 2007). To illustrate the misleading nature of an arithmetic mean we refer to the “Summary Edition Credit Suisse Global Investment Returns Yearbook 2019” Table 1:


There is a material difference between the arithmetic and geometric mean achieved for all equity markets examined. For Japan the arithmetic mean is more than double the geometric mean! While the returns quoted are “real” returns and the time period different, our point remains, an arithmetic mean has, and does, overstate the actual returns achieved.

Moreover, a cursory glance at Figure 22.2 illustrates returns of 24% p.a. to the “USA Market” from 1985-2007; this also indicates an arithmetic mean has been used because, to our knowledge, the US market simply did not achieve a compound annual growth rate of 24% from 1985 to 2007!

To take our point to a theoretical extreme let’s examine a short series of returns calculating both the arithmetic and geometric mean:



Remarkably, it is theoretically possible to achieve an arithmetic mean twice that of the market (16.00% vs 8.00%; 2x); and simultaneously attain only a fraction of that return based on the geometric mean (4.99% vs 0.95%; 0.19x).

Conclusions and Practical Implementation

Given the relative dearth of studies examining the returns of net nets, especially those in Japan, we are grateful for this article. Nonetheless, due to the silence on the data source, details pertaining to the research methodology and the strong likelihood that the returns specified were calculated using the arithmetic mean, thus overstating returns, little can be relied on from this article.
While it may be disheartening to be left with something of a “null hypothesis”, it does provide us with the opportunity to be clichéd and conclude with a quote from none other than Charles T. Munger:

“Any year that passes in which you don’t destroy one of your best loved ideas is a wasted year.”


[i] Please note, the article did not state which exchanges were examined in the markets studied.

Friday, 1 March 2019

An Analysis of “Benjamin Graham’s Net Current Asset Values: A Performance Update”


Summary


·       The study examined the performance of securities that were trading at no more than two-thirds of its Net Current Asset Value (“NAV”) during the 1970-82 period in the US

·       Net nets, on a gross basis, more than tripled the returns of the market (as measured by the S&P 500 TR)

·       Net nets, on a net basis (i.e. after commissions and potential taxes) more than doubled the returns of the market (as measured by the S&P 500 TR)

·       The study used data that excluded dividends; consequently, a number of findings within the study may not be reliable

Objective

The objective of the original study was to examine the performance of securities that were trading at no more than two-thirds of its “net current asset value (NAV)” during the 1970-82 period in the US.
Our objective was to analyse the study itself; determine its reliability, draw our own conclusions and glean, if any, actionable advice for the practitioner of the “net net” method of investing.
Please note the terms “NAV” and “net nets” are used interchangeably throughout the analysis.


Parameters

·         Title: Ben Graham's Net Current Asset Values: A Performance Update
·         Author: Henry R. Oppenheimer
·         Source: Financial Analysts Journal, Vol. 42, No. 6 (Nov. - Dec., 1986)
·         Reference number: Journal, Vol. 42, No. 6
·         Publication Year: 1986
·         Data source: Security Owner’s Guide
·         Period studied: 1970-82
·         Years in study: 13
·         Markets studied: US
·         Exchanges: NYSE, AMEX, OTC


Key Methodology


·         Valuation metric:

 “…sum of all liabilities and preferred stock and subtracted it from current assets; this result was then divided by the number of common shares outstanding to give NAV.”

i.e. Net Current Asset Value per share = (Current Assets – (Total Liabilities + Preferred Shares))/Common Shares Outstanding

“Our investor bought a security if its November closing price was no more than two-thirds of its NAV”

·         Weighting: “equally weighted”

·         Purchase/rebalance date: “last business day of December of each year”

·         Holding/rebalancing period: “The securities were held for either 12 or 30 months, depending on the analysis.”

Reliability

While there are numerous biases/errors that can be made when conducting studies/back tests, below we have analysed those we deem most likely to impact a study of this nature:
1.       Survivorship bias
“All securities selected from the December 1970 through December 1972 and December 1978 through December 1982 Security Owner's Stock Guide were evaluated. For the remaining years, we evaluated all NYSE securities as well as random samples of about 20 to 30 AMEX and OTC securities.5”
Given the use of “random” samples and the Security Owner’s Guide (as opposed to a large computerized data base that might not include delisted stocks) the data source appears to be free from “survivorship bias”.
2.       Look Ahead bias
“Our investor bought a security if its November closing price was no more than two-thirds of its NAV… the date of purchase, the last business day of December of each year.”
Given the lag between when the “NAV” was calculated and the “date of purchase” this study does not appear to suffer from “look ahead” bias.
Furthermore, given the data was retrieved from a physical manual, logically, “look ahead” bias could not have occurred.
3.       Time period bias
The study spans 13 years and we classify this as a “somewhat reliable” period.
For reference:
·         < 10 years; inadequate/unreliable
·         11 to 20 years; somewhat reliable
·         > 20 years; more reliable
·         > 40 years; most reliable
4.       Human error
Given this study was created in a rather manual and laborious manner we think it prudent to assume some errors may have occurred. 
5.       Journal rating/credibility[i]
Appearing in the Financial Analysts Journal adds credibility to the study in light of its reputation.


Reliability Assessment: Based on the above and considering the clarity with which the methodology was explained the study appears to be reliable. However, following further analysis (refer below), we conclude that a number of findings in the study may not be as reliable as they initially appear.



Results and Analysis


Annualized Return Analysis

Table IV is reproduced below:



Below we have adapted the data from Table IV and looked at the performance in both absolute and relative terms to the S&P 500 Total Return (TR, i.e. including dividends):



Gross returns:
·         Compound Annual Growth Rate (CAGR) of 28.5% (with rounding impacting our calculations) vs the S&P 500 TR 9.3%; an absolute outperformance of 19.2% and a multiple of 3.1 relative to the S&P 500 TR
·         1 losing year (1973, which incidentally also possessed the equal smallest sample size during the study period (18 securities; refer Table I, p 41.)
·         2 negative years vs 4 negative years for the S&P 500 TR

However, as Yogi Berra said, "In theory there is no difference between theory and practice. In practice there is."

Consequently, we simulated reality by adding commissions and taxes. We included commissions of 2% (1% to buy 1% to sell) and in relation to taxes, we utilised the capital gains taxes rates of the corresponding years[ii] in the study and assumed our hypothetical investor resided in the US and sat in the highest marginal tax bracket but held on long enough to be eligible for the long-term capital gains tax rates. In contrast we assumed an investor in the S&P 500 bought and held, never realizing their gains (and for simplicity we did not adjust their dividends for tax).


Net returns (i.e. including commissions and taxes):
·         Compound Annual Growth Rate (CAGR) of 21.0% vs the S&P 500 TR 9.3%; an absolute outperformance of 11.7% and a multiple of 2.3 relative to the S&P 500 TR
·         4 losing years
·         2 negative years vs 4 negative years for the S&P 500 TR

In addition to commissions and taxes we also examined the impact of inflation[iii], which was rampant during the period to ascertain its impact on returns.


Net Real returns (i.e. net of commissions, taxes and adjusted for inflation):
·         Compound Annual Growth Rate (CAGR) of 12.8% vs the S&P 500 TR 1.4%; an absolute outperformance of 11.4% and a multiple of 8.9 relative to the S&P 500 TR
·         4 losing years
·         2 negative years vs 5 negative years for the S&P 500 TR



Valuation Return Analysis


Table VI is reproduced from the study below:

Note: It is unstated whether this “mean monthly return” figure is intended to be a geometric or arithmetic mean.  Oppenheimer “put[s] [the] results in a form more meaningful to an investor” by stating that “$10,000 invested in the NAV portfolio on December 31, 1970 would have grown to $254,973 (with monthly compounding) by December 31, 1983.”  (p. 42.)  However, that is a period of 156 months, and compounding $10,000 at 2.45% for 156 periods leads to a terminal value of $436,381, not $254,793.  The implied compound monthly return rate for 156 periods and Oppenheimer’s terminal value is 2.10%.  Consistent with this 2.10% figure, Oppenheimer states explicitly that the geometric mean annual return for the 13-year period was 28.2%.  (p. 44.) Therefore, we conclude that the table above displays the arithmetic mean.

With regard to Table VI the author states; “The conclusion is clear-cut. Returns and excess returns can be rank-ordered, with securities having the smallest purchase price as a percentage of net asset value having the largest returns. It appears that degree of undervaluation is important…”
This conclusion is consistent with the empirical evidence produced when examining the returns of various valuation metrics; i.e. the “cheapest” group outperforms the more expensive groups, usually in a sequence consistent with a stairstep, as in this case.
Notwithstanding the above, due to concerns over the data used to measure returns (which is discussed in detail in subsequent sections), based on this study alone (and not taking into account our wider body of knowledge), the conclusion “that the degree of undervaluation is important” is not as reliable as it initially appears (despite its likely accuracy). 


Earnings and Dividends


Table V is reproduced from the study below:


Dividends

When it comes to dividends the spread between “Positive Earnings and Dividends” (Panel D) and Positive Earnings and No Dividends (Panel E) is stark. No economic rationale comes to mind for why a net net paying dividends would underperform one that doesn’t pay dividends. Perhaps the spread is due to positive earning dividend payers possessing a shallower discount compared to positive earnings non-dividend payers. For instance, if positive earning dividend payers were consistently in the “most expensive” quintile (Table VI, Panel A) then their returns of 2.01% are similar to the 1.88% identified for the most expensive quintile. Similarly, if positive earning non-dividend payers were consistently in the “cheapest” quintile (Table VI, Panel A) then their returns of 2.88% are not materially different from the 2.95% identified for the cheapest quintile. That said, the aforementioned seems unlikely.
Interestingly in note 17 (p. 47) of the study the author specifies that An analysis of the 13 portfolios was also performed. The conclusions about earnings and dividends also hold for them” implying an element of robustness to the conclusion that “Firms having positive earnings and paying a dividend provided a lower mean return than portfolios of firms with positive earnings not paying a dividend. They also had a lower systematic risk. Finally, their risk-adjusted excess returns were not as high as those of the portfolio of firms with positive earnings but not paying dividends. Choosing only firms that have earnings and pay a dividend does not help the investor.”
It would appear then, that dividends, at least in relation to the period and market studied were a hindrance to returns…or were they?
Note 9 (p. 47) specifies the following:
all comparisons with the exchange benchmarks returns without dividends are used for both the security return and the benchmark return.”
That is to say, to our understanding, the dividends have been excluded in the returns displayed in Table V (among others), therefore the returns have been artificially reduced by the amount of the dividend paid out. On this basis it isn’t at all unexpected that the dividend payers would seemingly underperform their non-dividend paying counterparts.
For clarity we will demonstrate using an example between a non-dividend and dividend paying stock:

Stock A (non-dividend)
Stock B (dividend)
Price t0
$10
$10
Price t1
$13
$12
Price Return (excl. dividend)
$3 (i.e. 30%)
$2 (i.e. 20%)
Dividend
$0
$3
Total Return (incl. dividends)
$3 (i.e. 30%)
$5 (i.e. 50%)

From the table above Stock B (the “dividend payer”) has produced the higher “total return” (i.e. 50% vs 30%), however, if you exclude the dividend Stock A (the “non-dividend payer”) appears to have produced the larger return (i.e. 30% vs 20%). While perplexing, it does, to our comprehension, appear as though this rudimentary error (i.e. excluding dividends) has been made in conducting this study[iv].

In light of the above, we are of the opinion that no reliable conclusion can be drawn in relation to the impact of dividends on the return to “net nets” based on this study. Interestingly, it also means that returns of “net nets” may be generally understated as a result of excluding dividends.


Earnings
The author sees the differential between “Positive Earnings” (Panel B) and “Negative Earnings” (Panel C) as marginal. The author concludes that No clear-cut pattern emerges from an examination of these panels. If anything, firms operating at a loss seem to have slightly higher returns and risk than firms with positive earnings.”
However, in light of our findings with regard to the data used to measure returns (discussed in depth in the section titled “Dividends”) we deviate from the author’s assessment in relation to “positive” and “negative” earning net nets. We agree that “No clear-cut pattern emerges from an examination”; but do so for different reasons.
It is possible (probable in our view) that “positive earners” had a greater propensity to pay dividends. Therefore, if those dividends were not included in the measurement of their returns they would be artificially reduced. Of course, it is also possible (albeit unlikely in our view) that “negative earners” paid the greater share of dividends subsequently understating their returns. Irrespective, we are of the opinion that no reliable conclusion can be drawn in relation to the impact of earnings on the return of “net nets” based on this study. 


Holding Periods, January effect, Size effect


Holding Periods


Table III is reproduced below:


While it would have been interesting to analyse the returns for portfolios held for 12 and 30 months this is not possible as the annual returns to the 12 month holding/rebalance period per Table IV are geometric means, whereas the returns displayed in Table III above are arithmetic means. However, the author’s commentary does specify that “The results largely parallel those of Table II” [which displays the monthly arithmetic mean for a 12 month holding/rebalance period]. From a return (and wealth) perspective, the advantage of the NAV portfolios over the market indexes is consistently pronounced.”



January effect

January effect; the author tested and found the return to the NAV portfolios “while consistently smaller than those reported in Table III, still supported the conclusion that the NAV criterion provided satisfactory performance.” That is to say, the performance of the NAV portfolios was not solely explained by the “January effect”.

Size effect

Size effect; to demonstrate that the returns to the NAV strategy were not merely a result of the “size effect” the author provides analyses (p. 45.) examining the return to the small-firm index and the “MV1” portfolio which is even smaller than the NAV portfolios. While we don’t know the valuation of the small-firm index and the MV1 portfolio the conclusion that the returns to the NAV portfolios are not purely a result of the size effect seems reasonable and robust.


Additional observations


One noteworthy comment that stands out from the study is as follows: “Even though the NAV criterion is the valuation technique Graham is most famous for, it has been subject to relatively little research.' Oppenheimer has provided tests of its performance over the 1949-72 period, but his tests (which do not demonstrate consistent profits) are largely confined to data prior to 1958.” (p. 40)
Interesting; unfortunately to date we have not found these other “tests”.


Conclusions and Practical Implementation


Based purely on this study our conclusions and practical takeaways for a “net net” investor are:
·         Net nets, on a gross basis, more tripled the returns of the market (as measured by the S&P 500 TR)
·         Net nets, on a net basis (i.e. after commissions and potential taxes) more than doubled the returns of the market (as measured by the S&P 500 TR)
·         Net net returns are not explained entirely by the “January effect”
·         Net net returns are not explained entirely by the “size effect”
·         The impact of dividends on returns was measured using data that excluded dividends (i.e. “price return” data as opposed to “total return” data) and therefore the conclusion reached in the study, “Choosing only firms that have…. a dividend does not help the investor” does not appear to be reliable
·         As data excluding dividends was also used to measure the returns of “positive” and “negative” earning net nets, no reliable conclusion can be reached in relation to the impact of earnings on the return of net nets
·         Also as a result of using data that excluded dividends returns may be generally understated throughout the study






We assumed our investor was in the highest tax bracket
For 1978 long term capital gains tax rates were 39.875/33.85%, we adopted the lower of these rates
For 1981 long term capital gains tax rates were 28/20%, we adopted the lower of these rates

[iv] While we attempted to contact Mr. Oppenheimer to attain clarification on the matter unfortunately, he appears to have retired in 2016 (https://web.uri.edu/business/2016/02/10/professor-oppenheimer/)