By Justin J. Carbonneau (@jjcarbonneau) —
A few weeks ago my partner, Jack Forehand (@practicalquant) wrote a piece titled “The Mechanics of Value Investing“. In it, he highlighted the various ways in which value investing can be defined. This topic was recently brought to light again in a recent tweet (see below) by Tom Psarofagis, ETF Analyst at Bloomberg. As Mr. Psarofagis points out, the underlying holdings in “value” ETFS can often be very different. As an investor, understanding why there can be such big differences between value portfolios can be a very important question to answer. To get at that answer, at least partially, I thought I’d look across the spectrum of value models run on Validea to see how value stock selection strategies can differ and what investors may be able to learn from this.
The “Why” Behind Value
As recent research has shown, value investing works over time partly because value stocks are riskier and partly because of the behavioral biases investors have. If you would like a substantive overview of these concepts and what is behind the value premium, Larry Swedroe’s article, Deep Dive into the Value Factor, is a great place to start.
Value stocks, and more specifically concentrated value portfolios, will have significant deviations from the market for periods of 1, 2 or even 3 years. We’ve been running focused model portfolios on the Validea site since 2003 and we’ve seen strong and sustained positive periods for value and very bad periods as well (we’re in one of those prolonged periods of underperformance of value now).
Validea’s Value Set
Of the 22 distinct strategies on Validea, many include value related investment criteria but the ten in the table below are the models that have the highest degree of emphasis on selecting stocks with value characteristics.
|Model Name||Guru||Source (book or academic paper)|
|Value Investor||Benjamin Graham||The Intelligent Investor|
|Contrarian Investor||David Dreman||Contrarian Investment Strategies|
|Book/Market Investor||Joseph Piotroski||Value Investing: The Use Of Historical Financial Statement Information|
|Private Equity Investor||Validea||Leveraged Small Value Equities|
|Value Composite Investor||James O’Shaughnessy||What Works on Wall Street (4th Ed)|
|Acquirer’s Multiple Investor||Tobias Carlisle||The Acquirer’s Multiple|
|Price/Sales Investor||Ken Fisher||Super Stocks|
|Low P/E Investor||John Neff||John Neff on Investing|
|Patient Investor||Warren Buffett||Buffettology|
The first six models in the table, from Graham through Tobias Carlisle’s Acquirer’s Multiple, would be considered the deepest value models. These five approaches seek out the cheapest, most beaten down stocks in the market based on metrics like Price/Earnings, Price/Book, The Acquirer’s Multiple (Enterprise Value / Operating Earnings) or other value factors.
The Fisher and Neff based models also have a key set of value criteria, but for a host of reasons they don’t pull in the absolute cheapest parts of the market as much as the other models. The Greenblatt Magic Formula model has a strong value component due to its earnings yield criteria, but quality (a sort on return on invested capital) carries just as much weight in the strategy The Buffett model is not looking at a value metric specifically as much as it is looking at the future return on a stock given the firm’s valuation and where earnings may be in the future.
The table below showcases the main value drivers within each of the models. No strategy uses the exact same criteria and as result no two models hold the exact same stocks. There is certainly overlap between value strategies, but no stock in our universe passes all of the models below with a score above 80%. To test this, I used Validea’s Guru Stock Screener and looked for stocks that pass at least 8 models with “Strong or Some” Interest. No stocks score that high across the full set of value models.
|Model Name||Guru||Key Value Criteria|
|Value Investor||Benjamin Graham||P/E below 15, P/B x PE below 22|
|Contrarian Investor||David Dreman||Needs to be in the bottom 20% of the market based on two out of four value criteria: 1) P/E 2) P/CF 3) P/B or 4) P/D|
|Book/Market Investor||Joseph Piotroski||P/B in the bottom 20% of the market|
|Acquirer’s Multiple Investor||Tobias Carlisle||Rank on Acquirer’s Multiple: Enterprise Value / Operating Earnings (EBIT)|
|Private Equity Investor||Validea||Enterprise value to EBITDA bottom 25% of the universe.|
|Price/Sales Investor||Ken Fisher||P/S below 1.5, but the lower the better.|
|Low P/E Investor||John Neff||Total Return (EPS + Yield) / PE|
|Value Composite Investor||James O’Shaughnessy||Value Composite: (Price/Book, Price/Sales, Price/Earnings, Price/Cash Flow, EV/EBITDA and shareholder yield)|
|Patient Investor||Warren Buffett||Expected Return vs. Current Price|
|Earnings Yield Investor||Joel Greenblatt||Earnings Yield (earnings divided by price)|
While there may not be any stocks that meet all of the value models at the same time, the one thing we do see is that the performance of the portfolios of many of the value strategies are correlated with each other. For example, the deepest value models have all struggled considerably over the past 12 months (and few years) while some of the others that are less focused on the absolute cheapest parts of the market have held up better.
On the flip side, some of the very best performance of our value stocks has come after large declines. For instance, in 2009 at the start of the bull market after the financial crisis some of the very best performers came from the concentrated deep value approaches.
|Guru Based on||Portfolio||# of Stocks||Rebalancing||Return||S&P 500||+/- S&P 500|
|Tobias Carlisle||Acquirer’s Multiple Investor||10||Monthly||153.90%||23.50%||130.40%|
|Validea||Private Equity Investor||20||Monthly||78.30%||23.50%||54.80%|
|Joseph Piotroski||Book/Market Investor||20||Annual||72.20%||23.50%||48.70%|
|James O’Shaughnessy||Value Composite Investor||10||Monthly||67.60%||23.50%||44.10%|
The correlation in performance of some of our value models also would indicate an implied sector bet or avoidance. For instance, most value models aren’t long technology stocks or other growth industries, and instead prefer places like financials, energy and areas like the beaten down retail industry. In a market that is rewarding technology and growing firms, you would expect many value strategies to struggle. Financial advisor Lawrence Hamtil wrote a very good piece on value stocks, inflation and the sector exposure (How Inflation Makes the ‘Value’ Factor a Sector Bet). His conclusion is that value stock performance is very closely tied to how well energy and financials do, and inflation helps those sectors. Without inflation, traditional value stocks in these sectors may not produce the performance many expect.
A Few Considerations for Investors
So I think there are a few takeaways for investors.
When you hear the word value, keep in mind there are many ways to express value with a stock selection model.
Second, value does not imply less risk. In fact, value is more risky. That is part of what makes value investing work over time.
Third, value strategies typically are making a bet on a sector or specific part of the market at various points in time. A focused value model may have been heavy in energy after the huge decline in energy names in 2015/2016, moved into retail a year or so ago and could be targeting a completely different area of the market today. It just depends on where the value is and the specific criteria being used to uncover that value.
Photo Credit Copyright: 123rf.com / niserin