What are leading economic indicators and how reliable are they?

The short answer

Leading economic indicators are time series that, on average, change direction before the broader business cycle does. The Conference Board LEI, the slope of the Treasury yield curve, building permits, ISM new orders and consumer expectations are the canonical examples used by central banks and macro forecasters.

Their reliability is regime-dependent. They worked relatively well during the manufacturing-dominated cycles of the 1960s-1990s, but the financialization of the economy, the rise of services and the unusual post-COVID dynamics have generated several false positives that did not translate into NBER recessions.

Composite indices typically outperform any single indicator, but no leading signal is a substitute for cross-validation across financial, real and credit measures.

→ New to macro indicators? Financial education hub

What the data shows

Research from the Federal Reserve Bank of Chicago documents that composite leading indices have done better than individual indicators at signaling recessions up to one year ahead, with the slope of the yield curve standing out among financial measures.

Key documented facts (Conference Board, NBER, Fed Chicago, 1970-2026):

• The Conference Board LEI is built from 10 components and uses a 3D rule: a six-month annualized growth below roughly -4.3% combined with a diffusion index below 50.
• Recessions have occurred in only about 12% of months since 1971, making naive accuracy a poor metric — area-under-the-ROC-curve is the preferred measure.
• The 10y-3m yield curve has inverted before every US recession since 1955, with one notable false signal in 1998.
• Building permits have substantially declined before 8 of the last 9 recessions since 1970.

The exception that nuances the picture: the LEI showed a 14-month consecutive year-over-year decline ending in early 2026 without an NBER recession occurring — among the longest non-recessionary stretches of LEI weakness on record.

→ Dataset: Yield curve spread 10y-3m

Why it happens — the macro mechanism

The classical idea is that some sectors of the economy turn before others. If we can identify those sensitive sectors, we can build composite signals that anticipate turning points by several months.

Channel 1 — interest-sensitive sectors lead. Housing, durable goods and capital expenditure react first to monetary policy because they are credit-financed. Building permits and new orders therefore typically decline well before aggregate GDP. See how building permits predict housing downturns.

Channel 2 — financial markets price expectations. The yield curve, credit spreads and equity prices reflect collective bets about future growth. They embed information that has not yet shown up in lagging hard data. The angle that matters most: their predictive content has been documented to weaken since the mid-1980s, as Dotsey (1998) and subsequent research show.

The transition between regimes matters. Manufacturing represented over 25% of US GDP in the 1970s; today it is around 11%. Indicators built primarily on goods-sector data therefore carry less information about a services-dominated economy.

Channel 3 — diffusion across the economy. When weakness spreads from one sector to many, recession probability rises. The Conference Board’s diffusion sub-index aims to capture this dimension explicitly.

Synthesis by regime: in the 1970s-1980s, with manufacturing-heavy cycles and bank-dependent credit, traditional LEI signaled recessions with relatively few false alarms; in the 1990s-2010s, with disinflation and increasingly globalized supply chains, signals retained directional value but lead times became more variable; in the post-2020 regime — characterized by fiscal-monetary interaction, services dominance and supply-chain distortions — the LEI generated its longest false-positive streak on record while the labor market remained resilient.

Leading indicators are not crystal balls — they are statistical bets that the next cycle will resemble the average of past cycles, which is exactly what each new regime tends to challenge.

→ Framework: Economic cycle phases and market implications

What it means for different economic actors

Households and savers can use leading indicators to put news flow in context, but treating any single signal as a binary buy-or-sell trigger ignores how often these series have produced false alarms.

Investors typically combine multiple signals: yield curve plus credit spreads plus PMI plus the labor market (Sahm rule). Cross-validation reduces the risk of acting on a single false positive — see recession indicators accuracy.

Policy makers rely on these indices for nowcasting and short-term forecasting; central banks like the ECB explicitly document the predictive accuracy of their PMI-based models and adjust them across regimes.

A common error is to read a single negative print as a confirmed recession signal. The literature shows that depth, duration and diffusion all need to align before the historical hit rate becomes meaningful.

Practical observation

Go deeper

Related questions

• Why does the Conference Board LEI sometimes give false signals?
• How do PMI surveys predict GDP changes?
• How do building permits predict housing downturns?
• What is the NBER recession dating methodology?

Frequently asked questions

Are composite indices really better than single signals?

The Chicago Fed research using AUC measures finds that composite indices and the yield curve dominate other measures in signaling recessions one to two years ahead. Single indicators can be informative but are more vulnerable to one-off shocks. Composite construction averages out idiosyncratic noise and improves robustness, though it does not eliminate regime breaks.

Why has the LEI’s reliability changed across regimes?

The composition of the economy has shifted. Manufacturing is now a smaller share of GDP, services are larger, and credit dynamics have changed with the rise of shadow banking. The LEI weights, fixed by historical correlations, may give too much importance to sectors whose macroeconomic footprint has shrunk — explaining why a deep manufacturing contraction in 2022-2023 coexisted with services-driven expansion.

Which leading indicator has been most stable historically?

The slope of the Treasury yield curve has the longest documented track record, having inverted before every NBER recession since 1955 with only one significant false alarm in 1998. However, even this signal has come under scrutiny — see why the yield curve inverts. The 2022-2024 inversion combined with no recession through early 2026 is rekindling the same debate.

Last updated — 15 May 2026