What is a value trap, and why does Penman focus on it?
A value trap is the risk embedded in a classic value position: buying high earnings-to-price and high book-to-price stocks does tend to earn higher returns on average, but the position can get badly hit, as it did in 2008 and for years afterward. The purpose of the talk is to understand what risk you take on when you enter this position.
Historical evidence since 1963 shows that ranking on earnings-to-price yields higher average returns for high E/P portfolios, and a large return spread also appears when ranking on book-to-price within earnings-to-price groups. Penman cautions that higher returns may simply be compensation for taking more risk, which is the question the analysis investigates.
How are value and growth investing typically defined?
They are most commonly defined by trading on multiples: a high earnings-to-price or book-to-price ratio is labeled value, while a low earnings-to-price or a high P/E or price-to-book is labeled growth. Penman treats these as marketing labels that lack transparency about what is actually being bought.
Morningstar styles and industry practice rely on these labels, typically earnings-to-price and book-to-price as the main metrics, but Penman argues the labels are mysterious and asks what you are really buying under this discipline.
What does the earnings-to-price ratio actually represent?
Earnings-to-price is a required return minus expected growth. Using a simple pricing model where price equals forward earnings capitalized at (r minus g), a given E/P could reflect a high required return with high growth, or a low required return with low growth.
Apple at a P/E of 11.5 (E/P of 8.7) illustrates the ambiguity: it could be a 12% required return with 3.3% growth, an 8.7% required return with zero growth, or a low-risk stock expected to deliver negative growth. Because growth is risky and may not pay off, a P/E reflects both growth and the risk of that growth, not just growth.
For a given earnings-to-price, what does a high book-to-price indicate?
For a given earnings-to-price, a high book-to-price means buying a firm with a low book rate of return and, counterintuitively, more growth, and that growth is risky. This inverts the conventional labels: the data shows high book-to-price portfolios are associated with higher future earnings growth, not low growth.
The relationship follows from the identity book-to-price = E/P × (book value / forward earnings), where book value over earnings is the book rate of return: holding E/P constant, ranking on book-to-price is an inverse ranking on ROE. Penman's portfolio data confirms the low ROEs, that two-year-ahead earnings growth is higher in the high book-to-price cells, and that the standard deviation and interdecile range of growth also increase, meaning buyers face right-tail upside but can be pounded in the left tail if the growth fails.
How does the accounting explain why low ROE signals risky growth?
Because historical cost accounting does not recognize earnings until uncertainty is resolved: revenues are not booked until there is a customer, and risky investments such as R&D, advertising, brand building, startup and training costs cannot be capitalized, so they are expensed immediately. Expensing lowers current earnings and ROE but, if the investments succeed, produces higher future earnings with no depreciation against them, yielding very high ROE for successful firms.
The accounting principle is conservative: you cannot put a not-yet-existing product's development on the balance sheet, and Graham's mantra about not putting water in the balance sheet came from 1920s asset write-ups that evaporated. Penman notes that over 50% of a typical SG&A line are expensed investments, so many loss-making companies are actually investing rather than failing, and the induced accounting pattern makes low ROEs indicate risk while high ROEs indicate realized success and lower risk.
How does Amazon fit this accounting analysis?
Amazon is not an unprofitable company but a potentially very profitable one: its technology platform development spending (a huge percentage of sales) is expensed each year, driving the low ROEs and reported losses in e-commerce. If those investments pay off, huge earnings would arrive with nothing on the balance sheet, producing very high ROE, though the if remains material and Penman agrees with value-investing friends that it may be too expensive while declining to short because shorting is risky.
Penman's friends short Amazon and see only losses in e-commerce, with profits coming only from cloud services, but the income statement's technology platform development line for grocery delivery, video streaming, film-making, and drone software is what gets expensed and depresses reported profitability.
How does Coca-Cola illustrate the other side of the accounting relationship?
Coca-Cola shows very high profitability and low risk because its brand generates earnings through the income statement while the brand itself is not on the balance sheet, no amortization charges it. Its book rate of return is very high and its beta is 0.4, connecting high ROE to lower risk and lower required return.
Does this mean the high book-to-price strategy is a bad trade?
No, Penman explicitly says he is not denying possible abnormal returns and calls it a good trading strategy with impressive returns. His warning is only against treating it as risk-free: the spread between 2.2 and 28.8 partly reflects risk, so beware if you troll it pretending you are not taking on risk.
The historical US returns from 1963 to 2012 came from a period (the American century) when betting on risky growth paid off and investors landed in the right-hand tail; buying Japanese, German, or Chinese stocks would have looked very different. Penman applies this caution to fund managers who advertise returns from good periods when their growth bets happened to pay.
What is the takeaway about the value versus growth labels?
The labels have been misapplied: when book-to-price is examined in conjunction with earnings-to-price, it is the other way around, high book-to-price relates positively to growth and to the risk you will not get that growth. Profitability is accounting profitability, not real profitability, and expensed investment gives a very different picture of what a multiple means.