Cyclically Adjusted EPS

What Is Cyclically Adjusted EPS?

Cyclically Adjusted EPS (CAEPS) is a normalized earnings metric that averages reported or estimated earnings per share across a complete economic cycle — generally defined as 7 to 10 years — to strip out the distorting effects of cyclical margin expansion, one-time items, credit-driven demand surges, and recessionary write-downs. The concept is directly analogous to Robert Shiller's CAPE ratio (Cyclically Adjusted Price-to-Earnings), but focuses specifically on the earnings denominator rather than the combined price-to-earnings multiple, giving analysts granular control over the normalization process.

The calculation adjusts raw EPS for inflation — typically using the GDP deflator or CPI — averages the real earnings stream, and then re-expresses the result in current-dollar terms. Some practitioners augment this with operating leverage cycle adjustments, removing the mechanical EPS inflation from debt-financed buybacks that inflate per-share figures without improving underlying business economics. Others incorporate return on invested capital normalization, anchoring earnings to a firm's long-run competitive position rather than its peak-cycle pricing power. The result is a trend earnings estimate that better reflects the firm's or index's structural earning power, independent of where the macro cycle happens to stand.

It is worth distinguishing CAEPS from simple trailing twelve-month EPS or forward consensus EPS, both of which are acutely sensitive to current-quarter margins. Forward estimates in particular are well-documented to be procyclical — Wall Street analyst revisions systematically lag turning points, meaning peak-cycle EPS forecasts embed the very margin assumptions CAEPS is designed to discount.

Why It Matters for Traders

At cycle peaks, reported S&P 500 EPS routinely runs 20–40% above cyclically adjusted estimates due to margin expansion, fiscal stimulus pass-through, and financial engineering through share repurchases. Anchoring valuation to those peak figures leads traders to dramatically underestimate earnings-at-risk when the cycle turns. Conversely, at cycle troughs, reported EPS collapses below CAEPS, making equities appear expensive on trailing multiples precisely when long-run forward returns are historically highest — a dynamic that systematically punishes momentum-driven and price-to-earnings-anchored strategies.

For macro hedge funds and multi-asset allocators, comparing the P/CAEPS ratio to its long-run average provides a regime-aware valuation signal that is structurally distinct from the equity risk premium derived from near-term consensus estimates. Because CAEPS is largely immune to analyst forecast revision cycles, it offers a more stable long-horizon anchor for positioning decisions — particularly useful when structuring equity versus credit relative value trades or sizing cyclical sector tilts through sector rotation frameworks. Cross-sectional CAEPS analysis by sector reveals which industries are overearning relative to their structural capacity, enabling disciplined earnings revision cycle positioning before the inflection becomes consensus.

How to Read and Interpret It

• P/CAEPS above 25x for the S&P 500 has historically preceded below-average 10-year forward real returns, generally below 4% annualized. The 2000 technology peak reached approximately 44x on this measure; pre-2008 levels peaked near 27x before the financial crisis reset.
• P/CAEPS below 12x has historically coincided with the most attractive long-run entry points — seen briefly in early 2009 near 13x and during the 1982 secular trough closer to 7x, both of which preceded multi-decade bull markets.
• A rapidly widening gap between trailing 12-month P/E and P/CAEPS signals that current margins are elevated relative to structural history — a forward warning that EPS revision breadth is vulnerable to mean reversion, especially if nominal GDP growth is decelerating.
• Sector-level CAEPS deviations exceeding one standard deviation above their own 10-year trend are a reliable flag for cyclical overearning and potential operating leverage cycle reversal, particularly in capital-intensive industries such as energy, materials, and industrials.
• When CAEPS itself is accelerating — the 10-year average moving higher in real terms — it signals genuine structural earnings improvement rather than cyclical inflation, a meaningful distinction for long-duration equity positioning.

Historical Context

During the 2021–2022 profit boom, S&P 500 reported EPS peaked near $220 for full-year 2022, significantly above a 10-year inflation-adjusted CAEPS estimate of approximately $145–160 at the time — implying roughly 35–50% cyclical overearning driven by pandemic-era fiscal transfers, extraordinary pricing power, and inventory accumulation gains. The resulting P/CAEPS ratio at the January 2022 market peak, with the S&P 500 near 4,800, was approximately 30–33x — well inside the danger zone that has historically preceded multi-year real return disappointment.

When margins began compressing through 2023, the earnings revision cycle turned sharply negative in cyclical sectors including consumer discretionary, materials, and freight-sensitive industrials, fully consistent with the mean-reversion that CAEPS had flagged over a year earlier. Financials were a partial exception, where higher net interest margins temporarily supported above-CAEPS earnings — illustrating that sector-level decomposition is essential and that index-level CAEPS can mask significant dispersion beneath the surface.

An earlier episode makes the same point in reverse: in March 2009, with the S&P 500 at roughly 670 and trailing EPS collapsing toward $40 on massive write-downs, the index's P/E appeared dangerously elevated at 16–17x. But CAEPS at that moment was closer to $52–55 in inflation-adjusted terms, placing P/CAEPS near 12–13x — a historically cheap reading that correctly identified the generational entry point that trailing multiples entirely obscured.

Limitations and Caveats

CAEPS is a backward-looking construct that can materially misrepresent structurally shifted earnings environments. The sustained margin expansion of the 2010s — driven by technology sector mix-shift within the S&P 500, secular interest rate decline reducing financing costs, and globalized labor arbitrage — was not cyclical noise but a structural regime change. A mechanical 10-year averaging window running through 2014–2020 would have systematically underestimated normalized earnings capacity, causing CAEPS-anchored investors to appear chronically underweight equities during one of history's strongest bull markets.

Additionally, share buyback-driven EPS growth that reduces share count without improving economic earnings requires explicit adjustment; failing to do so overstates structural profitability, particularly for capital-light large-cap technology and consumer staples companies that have aggressively returned capital. Changes in accounting standards — notably the shift to ASC 606 revenue recognition and IFRS 16 lease capitalization — create discontinuities in long time series that require careful normalization before averaging. Finally, the metric is most powerful at the index level and progressively less reliable for individual securities, where idiosyncratic competitive disruption can permanently alter structural earnings capacity in ways no averaging window captures.

What to Watch

• Quarterly gap between consensus forward EPS and trailing CAEPS: a spread exceeding 30% on the S&P 500 is a reliable warning that reported earnings are vulnerable to mean reversion and that forward P/E multiples are flattering the market's true valuation.
• Sector-level CAEPS dispersion across technology, energy, financials, and industrials — cross-sector CAEPS divergence above historical norms identifies both overvalued and undervalued pockets for sector rotation positioning.
• Nominal GDP growth trajectory relative to aggregate corporate revenue growth, which anchors the long-run CAEPS trend and provides early warning when top-line support for elevated margins is fading.
• Real interest rate regime shifts: structurally higher real rates compress normalized multiples, meaning the historical P/CAEPS danger zones may need to be recalibrated when the neutral rate itself has shifted — a critical analytical caveat for the current post-pandemic cycle.