Examining “An Empirical Analysis of Ben Graham's Net Current Asset Value Rule”
The focus of this post is the research paper “An Empirical Analysis Of Ben Graham's Net Current Asset Value Rule” by Joseph D Vu published in The Financial Review (1988). The objective of the paper was to examine the performance of securities that were trading at less than Net Current Asset Value (NCAV) during the 8 year period from 1977 to 1984 on the New York Stock Exchange (NYSE) and American Stock Exchange (AMEX).
Our objective was to analyze the study itself; determine its reliability, draw our own conclusions and glean, if any, actionable advice for the practitioner of the NCAV method of investing.
We have previously published a number of posts with regard to securities trading less than their Net Current Asset Value including:
An Analysis of “Benjamin Graham’s Net Current Asset Values: A Performance Update”
An Analysis of “Graham’s Net-Nets: Outdated or Outstanding?”
An Analysis of “Testing Benjamin Graham’s Net Current Asset Value Strategy in London”
Examining Greenblatts “How the small investor can beat the market”
“The NCAV is defined as current asset minus all liabilities including long-term debt and preferred stock.”
i.e. Net Current Asset Value = Current Assets – (Liabilities + Preferred Shares)
Stocks qualify for investment when they sell at (i.e. have a market capitalization) less than their NCAV. Unlike Graham, the author did not require any specific “margin of safety” relative to NCAV, e.g. 2/3 of NCAV.
Weighting of holdings: Equal weight
Portfolio formation: “stocks meeting the NCAV can enter the portfolio continuously at the end of each month, not on an arbitrary date such as December 31”. “This stock selection method … includes all stocks selling below NCAV at any time”.
Holding period: 2 years; “Each stock can enter the portfolio only once during a two-year period… For stocks that are acquired or liquidated, the final stock price is used in calculating monthly returns.”
Studies can suffer from a number of issues which reduce their reliability. Below we address potential concerns around sample size, return calculation methodology, data sources and common biases that may afflict research:
1. Sample size and firm characteristics
As is often the case, the number of firms selling below NCAV relative to the number of firms in a market is relatively low. In the examination period (1977-1984) 107 firms sold for less than NCAV on the NYSE and AMEX. Noteworthy is that fact that of those 107 firms, 50 were found in 1977 alone, 0 in 1983 and just 2 in 1984. By deduction, that means from 1978 to 1982, 55 firms sold for less than NCAV (an average of 11 per year).
The mean market capitalization of firms examined was “$51.3 million”[i]. However, no minimum market capitalization was specified. Consequently, the inclusion of the smallest firms may have biased the results as even when investing relatively modest sums the securities of the very smallest firms are virtually untradeable. The larger market capitalization relative to firms identified in other studies is likely a result of not including firms that traded over the counter as part of the assessable investment universe.
2. Return calculation methodology
“Raw Returns” are used to calculate investment performance. Raw returns are returns that are not adjusted for “risk” e.g. the Sharpe ratio is an example of a risk-adjusted return metric. In academic research (such as this) returns are often measured by calculating the arithmetic mean return (as opposed to the geometric mean return) - this appears to be the case in this study. It should be noted that in a dependant return series that exhibits volatility (like stock returns) the arithmetic mean will, as a matter of mathematical law, overstate returns relative to the more practitioner oriented geometric mean. The magnitude of the potential divergence between the two measures is unknown given the data made available in the study.
3. Survivorship bias
No mention was made with regard to controlling for “survivorship bias”.
However, given the authors used the physical “Value Line Investment Survey” from April 1977 to December 1984 (as opposed to a computerized data base that might not include delisted stocks) the data source would logically be free from survivorship bias.
4. Look Ahead bias
Controlling for look ahead bias was not specifically mentioned.
Prices were attained from the “Center for Research in Security Prices (CRSP) monthly return tapes at the University of Chicago”. While this is considered the “gold standard” for price data it doesn’t alleviate the possibility that prices retrieved from the database matched e.g. the date of underlying financial statements and therefore didn’t account for the delay in their public availability.
In the absence of an express statement addressing look ahead bias it remains a possibility.
5. Time period bias
The study spans 8 years and we classify this as an “inadequate/unreliable” period. However, as the authors mention that two stocks met the valuation criterion in 1984, and assuming the returns to those stocks were measured over the stipulated 2 year holding period, the study period may have been almost 10 years in length (i.e. April 1977 to December 1986).
· < 10 years; inadequate/unreliable
· 11 to 20 years; somewhat reliable
· > 20 years; more reliable
· > 40 years; most reliable
6. Data source and treatment
As mentioned previously the “Value Line Investment Survey” was used for fundamental data and the “Center for Research in Security Prices (CRSP) monthly return tapes” were used to retrieve pricing data.
The CRSP is considered the “gold standard” for pricing data; we assume therefore that dividends were accounted for in the pricing data, though, no mention of dividends was made in the study. Presumably, the well respected “Value Line Investment Survey” was a reliable source of fundamental information.
It should be noted that no industries were excluded; at times financials (e.g. banks) are excluded from “value screens” due to the different structure of their balance sheets.
7. Human error
Given this study required the manual retrieval of fundamental data by the researcher, we think it prudent to assume some error may have occurred.
8. Journal rating/credibility[ii]
This study was not, to our knowledge, published in a top tier academic journal and therefore cannot be granted the “additional credibility” that may come with such publication.
Reliability Assessment: Given the short time period (< 10 years), small sample size post 1977 , arithmetic mean return calculation (i.e. “Raw Returns”) and the absence of a minimum market capitalization requirement we would not place reliance on the results of the study from a practitioner’s standpoint.
Results and Analysis
For the 24 months before and after a stock met the necessary criterion (i.e. Market Cap < NCAV) the author calculated the monthly:
Cumulative Raw Return
Percentage of Positive Excess Return
Cumulative Excess Return
It will be important to emphasize the “excess returns” over the “raw returns” due to the sample of stocks being overwhelmingly concentrated in 1977. Focussing on excess returns will allow us to focus on the relative merits of stocks trading below NCAV. However, it should be noted that no definition of “excess returns” was identified in the study. Presumably it was the return above a broad market capitalization weighted index (such as the S&P 500) – an assumption we are forced to make to aid analysis.
Rather than reproduce the monthly statistics we calculate and summarise the key periods pre (t-24 to t-1) and post event (t0 to t24) along with our derivation of the cumulative market return (which was not reported in the original study):
It is stated that in “the pre-event period (month -24 to month -1), the average cumulative raw return is 21 percent with a t-value of 1. 76”. Reconciling this to the data we concur with the 21% return specified. Further, it is stated that the “post-event cumulative raw return of 61.7 percent is significant at the 0.01 level (t = 5.93)”. However, observing the return statistics it there appears to be a discrepancy; the post event cumulative raw return is actually 60.70% if month t0 is included – it makes sense to include month t0, otherwise the cumulative raw returns pre and post event do not reconcile. Such an adjustment does however leave us with a post event period of 25 months technically speaking.
Uncertainty over the exact pre and post event period aside, observing the cumulative return and cumulative excess return graphically reveals much:
Post event returns turn positive almost immediately. In the first year the cumulative raw return was 37.60% versus 22.60% for the market, an excess return of 15.00% (i.e. a 66.37% outperformance). In the second year the cumulative raw return was 23.10% versus 14.40% for the market, an excess return of 8.70% (i.e. a 60.41% outperformance). For the total two year post event period the cumulative raw return was 60.70% versus 37.00% for the market, an excess return of 23.70% (i.e. a 64.05% outperformance).
From the breakdown of the return statistics it appears that the first year post formation resulted in the bulk of return suggesting that a one year rebalancing period may prove to be return enhancing.
From April 1977 to December 1984 it appears as though firms trading below NCAV on the NYSE and AMEX demonstrated near perfect market timing capability – seemingly a magical time to be a deep value investor!
The study concludes with a foreshadowing… “Future research is needed to determine if the NCAV rule continues its past trend and to determine if certain adjustments for firm size can change the profitability of this trading rule.”
Indeed, the author along with Beni Lauterbach subsequently published a mathematically heavy research paper titled “Ben Graham's Net Current Asset Value Rule Revisited: The Size-Adjusted Returns”[iii]. The abstract states, “The study demonstrates how size controls can alter the outlook of an investment strategy. The Ben Graham net current asset value rule provides excellent excess returns according to traditional performance measures. Size-adjustment procedures, however, reveal that its size adjusted excess return is approximately zero.” We wonder what Graham would make of this assessment.
“An Empirical Analysis of Ben Graham's Net Current Asset Value Rule” by Joseph D Vu was an eight year study commencing in 1977 and concluding in 1984 that examined the returns to firms trading below Net Current Asset Value (NCAV) on the NYSE and AMEX exchanges.
Significantly the “raw returns” reported in the study were likely calculated as the arithmetic mean of returns; consequently, the reported returns would have overstated the actual returns achieved by an investor as measured by the more appropriate geometric mean. In addition, no minimum market capitalization was mandated for the firms examined, and therefore the study would have included even the smallest firms in the market. The inclusion of the very smallest firms may have biased the results as even when investing relatively modest sums, the securities of such firms are virtually untradeable.
Notes: [i] Approximately equal to an inflation adjusted USD 219m in 2020 (https://www.usinflationcalculator.com/) [ii] https://alphaarchitect.com/2015/04/01/where-to-find-cool-academic-finance-research/ [iii] Ben Graham's Net Current Asset Value Rule Revisited: The Size-Adjusted Returns, Beni Lauterbach and Joseph D. Vu, Quarterly Journal of Business and Economics, Vol. 32, No. 1 (Winter, 1993), pp. 82-108.