Guide

Conference Board leading economic index explained

Harbor Capital's multi-asset sleeve rotated into industrials and small caps in early 2025 because the yield curve had steepened and payrolls still printed above consensus. The desk treated one-month GDP tracking noise as confirmation. What they skipped was the Conference Board's Leading Economic Index (LEI): by the time headline growth looked soft, the composite had already fallen 4.2% over eight consecutive months and the component diffusion index had dropped below 40 — a breadth collapse that historically precedes NBER recessions by six to nine months. Rebuilding the regime model around LEI level, momentum, and diffusion cut cyclical drawdown in the 2025–26 soft patch from −11.4% to −6.8% versus benchmark while preserving upside when CEI growth re-accelerated.

The LEI is a monthly composite of ten forward-looking U.S. data series published by The Conference Board. It sits alongside the Coincident Economic Index (CEI) — where the economy is now — and the Lagging Index (LAG) — confirming turns after the fact. Together they form the classic leading/coincident/lagging framework for business-cycle analysis. This guide covers LEI construction and components, how to read month-over-month and six-month changes, diffusion breadth, recession heuristics, CEI and LAG context, a Harbor Capital refactor, a technique decision table versus single-indicator models, pitfalls, and an analyst checklist — complementing our guides on recessions, the Sahm rule, and GDP nowcasts.

What the LEI measures

The Leading Economic Index aggregates ten component series that tend to turn before broad economic activity. The Conference Board standardizes each component (volatility adjustment, trend removal), assigns weights, and chains them into an index set to 100 in 2016. A rising LEI suggests expansion should continue or strengthen over the next three to seven months; sustained declines signal mounting recession risk.

The LEI is not a GDP forecast by itself. It is a composite leading signal designed to front-run turns in the coincident economy. Analysts pair it with CEI growth (current conditions) and LAG behavior (confirmation) rather than trading any single monthly print in isolation.

Release timing: typically the third or fourth business day after month-end, covering the prior month's component data. Revisions to underlying series (claims, orders, permits) can revise LEI history; archive vintages if you backtest trading rules.

The ten LEI components

Component weights and exact definitions evolve when The Conference Board re-benchmarks the index, but the current U.S. LEI generally comprises:

Average weekly hours, manufacturing — employers cut hours before headcount; ties to weekly hours in the payroll report.
Average weekly initial claims for unemployment insurance — the highest-frequency labor stress gauge in the composite; see initial claims.
Manufacturers' new orders, consumer goods and materials — Census factory orders, demand pulse for goods production.
ISM new orders index — survey diffusion of order books; complements ISM manufacturing PMI.
Manufacturers' new orders, nondefense capital goods excluding aircraft — core capex pipeline excluding volatile aircraft and defense.
Building permits, new private housing units — forward-looking housing construction; links to building permits.
S&P 500 stock prices — financial conditions and expectations channel; equities often lead earnings and capex plans.
Leading Credit Index — proprietary Conference Board composite of credit conditions (spreads, lending standards, quantity of credit).
Interest rate spread (10-year Treasury minus federal funds rate) — captures curve shape; inversions historically precede recessions.
Average consumer expectations for business conditions — sentiment from household surveys; expectations often move before spending.

Because components span labor, orders, housing, finance, and sentiment, the LEI diversifies single-series noise. When most components fall together, the signal is stronger than any one input — that is what the diffusion index quantifies.

CEI, LAG, and the three-index framework

Coincident Economic Index (CEI)

Four series that move with the present cycle: nonfarm payroll employment, real personal income less transfer payments, industrial production, and real manufacturing and trade sales. CEI growth approximates “where we are” in the expansion or contraction. LEI leads CEI; watching LEI fall while CEI is still rising is the classic late-cycle warning pattern.

Lagging Index (LAG)

Seven series that confirm turns after they occur: average duration of unemployment, inventory-to-sales ratio, labor cost per unit of output, average prime rate, commercial and industrial loans, consumer installment credit to income, and services CPI change. LAG peaks after CEI and troughs later — useful for validating that a downturn is mature, not for timing entries.

Reading the trio together

Healthy expansion: LEI rising, CEI rising, LAG rising (late expansion may show LEI flattening first). Recession onset: LEI fell months earlier; CEI turns down; LAG still elevated then rolls over. Recovery: LEI inflects first, CEI follows, LAG lags. Plot all three with year-over-year percent changes for regime dashboards.

How to read LEI momentum and diffusion

Month-over-month and six-month change

Headline LEI m/m percent change is volatile. The Conference Board emphasizes six-month percent change (annualized or simple, per release notes) and the six-month diffusion index — the share of components improving minus deteriorating over six months, scaled to −100 to +100.

Rule-of-thumb heuristics (not mechanical laws): sustained LEI declines over six-plus months and diffusion below 50 often precede recessions; sharp LEI rebounds with diffusion above 50 support recovery trades. Cross-check against GDP nowcasts for quarter-level magnitude, not just direction.

Contribution analysis

Each release publishes component contributions to the m/m LEI change. Negative contributions concentrated in orders, hours, and claims differ from a LEI drop driven mainly by the yield spread or equities. Decompose before narrating “manufacturing recession” versus “financial conditions shock.”

Recession probability models

The Conference Board and academic researchers publish probit models linking LEI declines to recession odds. Treat published probabilities as Bayesian updates, not certainties — COVID-era supply shocks and fiscal pulses broke several historical patterns briefly.

Harbor Capital regime sleeve refactor

Harbor's prior playbook used yield-curve steepening as an all-clear for cyclical beta. After the 2025 drawdown, risk managers added a three-gate LEI overlay:

Momentum gate — six-month LEI change below −1.5% halves cyclical overweight versus policy benchmark.
Diffusion gate — six-month diffusion below 40 blocks new small-cap and machinery adds even if CEI is positive.
Confirmation gate — LEI must rise two consecutive months with diffusion above 50 before restoring full cyclical tilt.

Gates paired with the Sahm rule approach zone (0.35pp) reduced false re-risking during soft-landing headlines. Backtests from 1990–2024 showed LEI-plus-Sahm cut peak-to-trough equity sleeve drawdown by 3.8 percentage points versus curve-only rules, with modest lag on re-entry (average +1.2 months versus CEI trough).

Technique decision table

Question	LEI composite focus	Single-indicator alternative
Recession timing 6–9 months ahead	LEI six-month change + diffusion breadth	Yield curve inversion alone (misses labor-led slowdowns)
Real-time labor deterioration	LEI claims + hours components	Sahm rule (faster, narrower; use both)
Quarter GDP magnitude	Supplement with nowcast models	LEI level (wrong frequency for Q/Q GDP)
Current activity for earnings	CEI growth, not LEI	LEI m/m noise
Confirm downturn is mature	LAG rollover after CEI peak	LEI (too early for confirmation)
Sector rotation (early cycle)	LEI inflection + orders components	Lagging unemployment rate level

Use LEI when you need a balanced leading dashboard that already embeds labor, orders, housing, credit, and financial conditions. Prefer dedicated indicators when the question is narrow (claims for weekly labor stress, Sahm for binary recession flag, nowcast for GDP tracking).

Historical context

The LEI lineage traces to the U.S. Department of Commerce leading indicators (1960s) before The Conference Board assumed publication. Major recessions since 1970 typically featured LEI peaks six to twenty months before CEI peaks, though lead times vary: the 2008 financial crisis showed a long LEI slide; 2020's pandemic produced an abrupt cliff inconsistent with slow-burn patterns.

Post-2022, LEI declined for an extended stretch while employment stayed resilient — a “rolling recession” in goods and housing while services held. That episode reinforced that LEI signals sectoral breadth stress even when headline GDP avoids two negative quarters. Pair LEI with industrial production and services PMI to see which side of the economy the leading signal is targeting.

Pitfalls

Trading single m/m prints — components revise; six-month trends matter more than one noisy month.
Ignoring diffusion — a flat LEI with 20% breadth deteriorating differs from broad-based strength.
Using LEI for GDP point forecasts — wrong tool; use nowcasts for quarterly tracking.
Mixing vintages — component revisions rewrite history; store release-date snapshots for backtests.
Extrapolating pandemic breaks — 2020–21 distort lead-time statistics; use long samples with annotation.
Forgetting index rebasing — 2016=100 base changes; compare growth rates, not levels across decades without adjustment.
Overweighting equities component — stock prices are volatile; check whether orders and claims confirm the LEI move.
Neglecting CEI — LEI can fall while CEI rises for quarters; define exit rules for false late-cycle calls.

Investor and analyst checklist

Bookmark The Conference Board LEI release calendar (monthly, ~third week).
Download LEI, CEI, LAG, and six-month diffusion from FRED or Conference Board data.
Plot six-month percent change in LEI with NBER recession shading.
Decompose each release: which components drove the m/m move?
Cross-check claims and ISM new orders against their LEI component contributions.
Pair LEI deterioration with Sahm approach zone and credit spread widening.
Use CEI for current-earnings regime; use LEI for 6-month risk budget shifts.
Archive vintage LEI values at release timestamp before revisions.
Stress-test portfolios on LEI −4% six-month scenarios, not just inversions.
Re-read Conference Board methodology notes after annual benchmark updates.

Key takeaways

The LEI composites ten leading series — labor, orders, housing, credit, equities, and sentiment — into one business-cycle early warning.
Six-month LEI change and the diffusion index measure breadth; single monthly prints are noisy.
CEI tracks present activity; LAG confirms turns — use all three indexes, not LEI alone.
Harbor Capital cut cyclical drawdown 4.6 pp by gating sector tilts on LEI momentum and diffusion, not yield-curve steepening alone.
Pair LEI with Sahm, claims, and GDP nowcasts — each answers a different recession-timing question.