Research Paper Proposes an Accounting-Adjusted Way to Read CAPE
A September 18, 2026 manuscript by Dino Palazzo proposes an accounting-adjusted version of the cyclically adjusted price-earnings ratio, called CAPE-H. The paper asks whether changes in financial reporting have made today's reported earnings less comparable with earnings from earlier decades.
The manuscript focuses on two items: research and development (R&D) expense and special items such as restructuring charges and asset impairments. Its argument is that these accounting effects can depress reported earnings, raising a price-to-earnings ratio even when the change does not represent a comparable decline in underlying earning capacity.
What the paper changes
Traditional CAPE compares an index price with a ten-year average of inflation-adjusted reported earnings. The FASB's summary of Statement No. 2 says the 1974 standard requires R&D costs to be charged to expense when incurred.
Palazzo's proposed CAPE-H adds R&D expenditures back to reported earnings and removes special items before calculating the long-run average. The author describes this as a measurement-consistency adjustment. It does not estimate the value of each company's research program, build an intangible-capital asset, or determine an ideal amortization schedule.
What the manuscript reports
The paper reports that CAPE-H identifies recurring high- and low-valuation episodes over its historical sample. It also reports improved out-of-sample forecasts of cumulative stock-price appreciation at three- and five-year horizons compared with traditional CAPE. Dividend growth, by contrast, remains difficult to predict under either measure in the paper's analysis.
Those are empirical results from the manuscript's data and model choices. They do not establish that the proposed measure will forecast returns in every future period, and a forecast of lower long-run returns is not a forecast of the date or speed of a market correction.
What the result does not establish
CAPE-H is one research proposal, not a generally accepted replacement for CAPE. Adding back R&D may improve consistency across accounting eras, but it does not prove that every R&D dollar creates an asset of equal value. Removing a special item from an earnings series also requires judgment: a charge may be unusual for one company but common across an industry or repeated over time.
The manuscript identifies Palazzo's Federal Reserve Board affiliation and states that its views do not necessarily reflect the views of the Board or its staff. The linked SSRN record is a working-paper record, not a Federal Reserve policy statement.
For a plain-English overview of CAPE, the accounting issue and the limits of return forecasts, start with CAPE, Accounting and Long-Term Stock Returns: A Plain-English Pillar Guide.