Guide

Earnings yield explained

Harbor Capital launched a domestic value sleeve in late 2025 with a simple rule: buy stocks where the P/E ratio was below 12. On paper, that implied paying less than twelve years of earnings — a classic Graham-style bargain. Eighteen months later, the sleeve lagged its benchmark by 420 basis points annualized. Post-mortem: six holdings had trailing P/E below 10, which translated to an ostensibly rich earnings yield above 10%. But those yields were built on cyclically peak EPS in chemicals, freight, and homebuilding. Normalized earnings cut the yield roughly in half; three names cut dividends within a year. The screen treated low P/E as synonymous with cheap when, for cyclicals at the top of the cycle, low P/E often means peak earnings — the exact moment when trailing earnings yield looks best and forward economics are worst.

Earnings yield (E/P) is the inverse of the P/E ratio: earnings per share divided by share price, expressed as a percentage. It answers “what fraction of my purchase price did the company earn last year?” in the same language bond investors use for coupon yields. This guide covers trailing vs forward earnings yield, enterprise-level variants, comparison to Treasury yields and FCF yield, sector context, the Harbor Capital refactor, a technique decision table, pitfalls, and an investor checklist.

What earnings yield measures

If you pay $50 for a share earning $4 per year, your earnings yield is 8% ($4 ÷ $50). That is the same information as a P/E of 12.5x, just inverted. Yield framing helps when comparing equities to fixed income, ranking a universe of stocks by “profit return on price,” or building factor portfolios that overweight high E/P names.

Earnings yield is not cash in your pocket. Unlike dividend yield, it includes retained earnings reinvested in the business. A 12% earnings yield with a 2% dividend yield means most profit stays inside the company — which may compound well or disappear into bad acquisitions and stock-based compensation.

Core formula (trailing):

earnings yield = EPS (trailing twelve months) ÷ share price × 100

Forward variant uses consensus or management-guided next-twelve-month EPS. Always label which EPS you use; mixing trailing price with forward EPS (or vice versa) is a common data error.

Earnings yield vs P/E: same data, different intuition

Mathematically, E/P = 1 ÷ (P/E). The choice of display is psychological and practical:

P/E — intuitive for growth investors (“how many years of profit am I buying?”). Breaks down at negative or near-zero earnings.
E/P yield — intuitive for value and asset-allocation comparisons (“what return on price does accounting profit imply?”). Handles high-multiple stocks gracefully (a 50x P/E is 2% yield; easy to compare to a 4.5% Treasury).

When screening, many quant value strategies rank on earnings yield rather than P/E because the distribution is more stable at the high-multiple end and combines naturally with other yield metrics in composite scores.

Trailing vs forward earnings yield

Variant	Numerator	Best for	Main risk
Trailing E/P	TTM reported EPS	Historical fact, credit-style analysis	Cycle timing — peak EPS inflates yield
Forward E/P	NTM consensus EPS	Prospective value, growth-adjusted screens	Stale or optimistic analyst estimates
Normalized E/P	Mid-cycle or adjusted EPS	Cyclicals, turnarounds	Subjective normalization assumptions
Shiller E/P (CAPE)	10-year real avg EPS	Market-level valuation	Not for single-stock timing

Harbor Capital’s mistake was trailing-only. Adding a rule that forward earnings yield must be within 150 bps of trailing (unless documented turnaround) flagged cyclical traps early. Pair forward E/P with PEG when growth differs materially across names.

Earnings yield vs other yield metrics

Investors often stack yield measures to see where accounting profit, cash, and distributions diverge:

Earnings yield (E/P) — accrual net income return on equity price. Fast, widely available; subject to GAAP noise.
FCF yield (FCF/EV or FCF/market cap) — cash after capex relative to enterprise or equity value. Better for capex-heavy firms; see our FCF yield guide.
Dividend yield — cash actually paid to shareholders. High E/P with low dividend yield implies retention; verify reinvestment quality.
EBIT/EV yield — capital-structure-neutral operating profit yield; useful when leverage differs across comparables.

A healthy value candidate often shows high earnings yield and high FCF yield within a reasonable spread. When E/P is 11% but FCF yield is 2%, investigate working capital swings, capex underinvestment, or SBC dilution before calling the stock cheap.

Equity risk premium and bond comparison

At the index level, strategists compare the S&P 500 earnings yield to the 10-year Treasury yield. The gap is a rough equity risk premium (ERP) — the extra return investors demand for owning volatile equities over default-free bonds. When the ERP compresses toward zero or turns negative (stocks yielding less than Treasuries on trailing E/P), history suggests below-average forward equity returns — though timing is imprecise.

Single-stock E/P vs Treasuries is weaker logic: company-specific risk dominates. Use bond comparison for allocation context, not as a standalone buy signal for one industrial name.

Sector benchmarks and interpretation bands

Absolute earnings yield thresholds mislead across sectors. Approximate trailing E/P medians (illustrative, vary by cycle):

Large-cap tech / growth — 2–5% (high P/E norms)
Consumer staples — 4–7%
Industrials / diversified — 5–8%
Financials — 6–10% (EPS quality varies by cycle)
Deep value / distressed — 10%+ (often reflects risk, not bargain)

Compare within sector and history. A bank at 12% trailing E/P during an credit upswing may be cheaper on normalized EPS than a utility at 7% with rising leverage and a 95% payout ratio.

Harbor Capital refactor

Harbor replaced the blunt P/E < 12 rule with a layered earnings-yield stack:

Forward earnings yield ≥ 7% (top tercile within GICS industry group, not global absolute).
FCF yield ≥ 5% on enterprise value, or documented temporary WC investment with line-of-sight recovery.
Trailing-forward E/P gap ≤ 150 bps unless flagged turnaround with two quarters of estimate revisions upward.
Earnings quality screen — exclude names where accruals > 10% of assets or one-time gains > 20% of net income (see earnings quality).
Leverage gate — net debt/EBITDA ≤ 3.5× unless utility/regulated pass-through.

Outcomes over the next four quarters: value-trap incidence in the sleeve fell from 34% to 9% of holdings (defined as >25% forward EPS downgrade within 12 months); sleeve tracking error vs pure low-P/E benchmark dropped 180 bps; median holding forward earnings yield stabilized at 8.4% vs a misleading 11.2% trailing median under the old rule.

Technique decision table

Approach	Best for	Weak when
Trailing earnings yield alone	Quick rank, mature steady-state businesses	Cyclicals, turnarounds, one-time EPS spikes
Forward earnings yield	Prospective value, estimate-driven screens	Thin coverage, post-IPO, fraud risk names
E/P vs Treasury (index ERP)	Asset allocation, market timing context	Single-stock buy/sell decisions
E/P + FCF yield combo	Value with cash verification	Early-stage, pre-profit growth
EBIT/EV yield	Leverage-diverse peer sets	Capital-light platforms with minimal EBIT
Normalized E/P	Commodity and industrial cyclicals	Requires defensible mid-cycle assumptions

Common pitfalls

Peak-cycle value illusion — highest trailing E/P at the top of the cycle; forward EPS collapses.
Negative earnings — E/P is meaningless; use forward or EV/sales with explicit loss horizon.
Ignoring dilution — use diluted EPS for yield, not basic, when options and RSUs are material.
Mixing price dates — Friday close price with Wednesday EPS release creates phantom yield moves.
Confusing yield with payout — high E/P does not mean high dividends or buybacks.
GAAP vs adjusted EPS wars — “Adjusted” E/P can hide recurring costs; prefer cash cross-check.
Currency mismatch — ADR price in USD, EPS in local currency without FX alignment.

Investor checklist

Compute trailing and forward earnings yield; note the gap and explain it.
Use diluted EPS; confirm TTM vs NTM definitions match your data vendor.
Cross-check with FCF yield on the same valuation date.
For cyclicals, run a normalized EPS scenario before ranking on E/P.
Compare to sector median and five-year own history, not a global cutoff.
Read accruals and one-time items — high yield from low-quality EPS fails the screen.
Pair with balance-sheet metrics (leverage, interest coverage) before sizing.
Document whether yield includes or excludes minority interest / preferred dividends.
Recompute after material estimate revisions, splits, or buyback acceleration.
Log thesis: what has to be true for this yield to persist or grow?

Key takeaways

Earnings yield is the inverse of P/E — same data, better for yield comparisons and factor ranking.
Trailing E/P peaks at the wrong time for cyclicals — Harbor cut value traps by adding forward and FCF gates.
Stack yields, do not stop at E/P — FCF and dividend yields reveal whether accounting profit converts to cash and payouts.
Index E/P vs bonds informs allocation — not individual stock timing.
Sector-relative ranking beats absolute cutoffs — 8% means different things in software vs regional banks.