That sounds reasonable, but there is an important subtlety. If a firm buys inventory that it can sell tomorrow with high probability, it books it as an asset on the balance sheet, If a firm builds a plant to produce inventory that it can sell with high probability, it books the plant to the balance sheet. But if a firm invests in R&D that has produced no product of revenue as yet…and it might not be successful…should the investment be booked to the balance sheet? There is a lower probability of payoff, so should an asset be booked with the pretense that it provides collateral? We are told that 85% of R&D investment is unsuccessful.

The investor is interested in the payoff to investment, and th ecutting edge here is the probability of payoff. That is the criterion for capitalizing assets in the balance sheet under FASB and IFRS accounting standards. For the accounting for R&D under FASB Statement No. 2, the FASB requires the investment to be expensed due to the “uncertainty of future benefits.”  In IAS 38, the IASB applies the criterion of “probable future economic benefits” to distinguish between “research” (which is expensed) and “development” (which is capitalized in the balance sheet and amortized). Even IAS 16 on property, plant, and equipment requires benefits to be “probable” for the asset to be booked.

Reporting R&D just says an expenditure has been made, nothing about the likely payoff. So the criterion for booking any asset to the balance sheet is the probability of future benefits. That suits the investor who is concerned about the risk to payoffs. Indeed, we will see in chapter 8 on risk to value that the expensing of intangible assets to the income statement conveys important risk information to the investor. The accounting authorities might not have the precise calibration as yet, but the investor is warned: Buying a firm with R&D with relatively low probability of payoff (an R&D startup, for example) is risky. And that applies to other intangible assets. For expenditures on advertising or brand building, it is uncertain whether the customers will be hooked. For investment in human capital, the outcome is uncertain; the employees may leave and go to a competitor. And so for investment to develop supply chains and distribution systems, customer loyalty programs, software development, merger costs, start-up and organization costs, and more.

For more on this, see

Barker, R., A. Lennard, S. Penman, and A. Teixeira. 2021.Accounting for Intangible Assets: Suggested Solutions.” Accounting and Business Research Vol. 52 No. 6, 601-630

Penman, S. 2023. Accounting for Intangible Assets: Thinking it Through. Australian Accounting Review Vol. 33 No. 1, 5-13.

Penman, S. 2024. Empirical Research on Capitalizing Intangible Assets is Logically Incoherent. At https://ssrn.com/abstract=4982366.

The value investor focuses on growth and the risk to growth: Do I want to buy growth? Can I bear the risk of buying growth? Here is the price that the market was putting on growth for some large-cap firms involved with AI in January 2025. The calculations are those of chapter 4 with a 10% required return. The no-growth price is based on book value and analysts’ consensus EPS forecast for 2025. The growth price is a percent of the stock market price.

The growth price is the number to challenge. From an understanding of AI developments, you might choose one, all of them, or none of them. Without this understanding, you might consider the risk to growth is too high. Develop a growth-risk profile as in chapter 8. Beware of buying growth.

An example of sales generated by investment asset growth or not is same store sales versus sales from new store openings in retail. The P, Q, FX breakdown is discussed in chapter 10.

Here is the growth-return profile:

In December 2024, Apple’s forward ROE was 96.13% and its book-to-price was 0.016. Thus,

Here is the updated growth-return profile:

Do you see how the profile has shifted down over time as Apple’s stock price increased? In 2018, the profile indicated a good BUY, but the return for growth purchased declined significantly by 2024. Berkshire Hathaway reduced their stake in Apple during 2024.

A recent paper develops an industry-level estimator they call the “Weighted Production Price Index” (WPPI) score. The goal of the WPPI is to succinctly summarize the net effect of inflation shocks across all industries on a given focal industry. For each industry-month, the WPPI score weights the vector of industry-level Purchasing Poer Index (PPI), where the weights are determined by the historical relationship between returns of that industry and the cross-section of industry PPI indices. The WPPI differs from an industry’s PPI. An industry’s PPI is a variable reported by the U.S. Bureau of Labor and Statistics and represents the weighted-average of monthly price changes in that industry. An industry’s WPPI is a comprehensive measure that reflects the effect of all price changes across all industries on the stock returns of the firms in that industry. Each industry-month WPPI encapsulates the composite inflationary effect of monthly price changes across all industries on the firms in a given industry. An industry or a firm that receives a high (low) monthly WPPI score has just experienced positive (negative) inflationary news.

We have not worked with it, but it is worth a look. The paper documents a strong correlation between WPPI and contemporaneous earnings (ROE or ROA) in the cross-section, more than twice as high as that between PPI and same-quarter earnings. Similarly, WPPI has a stronger correlation with contemporaneous industry returns than own-industry PPI. The paper also reports that WPPI has strong predictive power for cross-sectional industry-level and firm-level returns, while own-industry PPI does not. In out-of-sample tests, an industry-level WPPI-based strategy that buys recent inflation winners and shorts recent inflation losers yields in a seven-factor alpha of 1% per month.

See Feng, J., S. Huang, C. Lee, and Y. Song. 2024. Inflation in the Cross-section: Separating Winners from Losers. At https://www.ssrn.com/abstract=4907871.

CPI is the U.S. Consumer Price Index.

The following are the returns and yields on a day that the inflation rate was announced:

2/13/2024:  Expected 2.9%; Actual 3.1%; S&P Down 1.4%; T yield up to 4.315% from 4.293%

Beware of assuming a probability distribution and of models that assume a particular distribution. Rather go through the fundamental analysis asking: What is the upside potential in this business. What is the downside? Value added (residual income) is at risk: What is the upside and the downside?

Why do theorists use assumed probability distributions in their modelling? To get what they call “closed-form solutions.”  That is, they want to get to a formula. But formulas are always suspect, as chapter 3 pointed; they can leave you into thinking that valuation is just a matter of plugging into formulas. Formulas sometimes help your thinking, but only if they are based on reasonable assumptions. The residual income model is one: The two assumptions underlying it are palatable. First, investors invest to get dividends: quite reasonable. Second, the clean-surplus equation by which successive book values are increased with earnings and reduced by dividends: That’s how accounting actually works!

The same applies to risk measures. In an enormous effort to get precision, modern finance has developed risk measures for investors, the CAPM beta being just one of them. It has not been a very successful effort. On reflection, it is unrealistic to think that the many risks that a business faces^2Mproduct risk and its many dimensions, supply risk, regulatory risk, currency risk, leverage risk (to name a few)^2Mcan be reduced to just one number, beta or similar. That can be done by making assumptions (a probability distribution), but that’s a cop out. To think we can say the cost of capital is 8% and not 7% or 9% is rather naïve. Beware.

In the 1990’s, John Meriwether, former head of bond trading at Salamon Brothers, got together with a couple of academics who had received Nobel Prizes for their work on risk and asset pricing to form a fund called Long Term Capital Management. The fund’s strategy was to find arbitrage opportunities where the prices of securities relative to each other could not be explained by difference in risk measures. The fund’s name turned out to be a misnomer. After three years of outstanding returns with the fund growing to $100 billion, the strategies went against them in 1998. The fund collapsed, nearly bringing Wall Street down with it if it weren’t for a rescue arranged by the New York Fed. The fund’s strategy turned out to sophisticated gambling under pretense of modern finance theory. Beware.

The episode minds us to wary of models, including the models of modern finance.

(The two academics were Robert Merton and Myron Scholes. Their theoretical work is wonderful, an important contribution to the theory of finance. In their work, these gentlemen taught us how to think about risk, and some of that thinking is in chapter 8. The two were also behind the thinking for Black-Scholes option pricing model that does work!)

See the book, When Genius Fails by Peter Lowenstein and the Wikipedia entry on the episode.

The formula also gives us a short cut for forecasting future residual income. If our financial statement analysis indicates that particular drivers are fairly constant, then forecasting is simplified:

• If ATO is constant, future ReOI is given by a forecast of Sales and PM
• If PM is constant, future ReOI is given by a forecast of Sales and ATO
• If PM and ATO are constant (and thus RNOA is constant), future ReOI is given by a forecast of Sales