Help · Knowledge base · Concept

Business cycle & market phases

From the platform knowledge base — the same entry the platform's AI agent cites in its answers. Educational reference, not advice.

Business cycle & market phases

Definition

The business cycle is the economy’s alternation of expansions and contractions, officially dated in the US by the NBER’s Business Cycle Dating Committee — which identifies peaks and troughs from a basket of monthly indicators (employment, real income, industrial production, real spending) with announcements that arrive many months AFTER the fact. Markets trade the cycle’s phases in anticipation: equity troughs historically LEAD economic troughs, sector leadership rotates with the phase, and the phase-conditioning of factor and asset-class returns is among the most-used frames in allocation. The KB carries the frame with its central honesty requirement: real-time phase identification is genuinely hard, and NBER’s own dates are retrospective.

How it works / structure

  • The dating machinery (NBER): no fixed rule — “significant decline in activity, spread across the economy, lasting more than a few months,” judged from the indicator basket; announcements lag turning points by 4-21 months (documented), which is why trading systems use PROXIES, never the official dates.
  • Real-time phase proxies (each labeled): yield curve for late-cycle (macro-yield-curve), credit spreads for stress onset (indicator-credit-spreads), ISM/PMI above-below-50 with direction (the most common practitioner four-phase grid), unemployment’s Sahm-rule trigger for recession onset, LEI composites — none individually reliable, and their disagreements ARE the hard cases.
  • The market’s lead: equity bear-market troughs have historically preceded recession-end by ~4-6 months on average (documented, dispersed) — “waiting for the all-clear” has meant buying well off the lows; conversely equities peak BEFORE recessions begin, with leads too dispersed for timing.
  • Phase-conditioned patterns (conventions, labeled practitioner): early-cycle — small caps, cyclicals, credit-sensitive names lead; mid — broadest participation; late — energy/materials and quality defensiveness; recession — staples, utilities, long duration bonds (strategy-sector-rotation carries the full grid and its evidence limits).

When it applies

Allocation tilts and sector-rotation frameworks (the phase grid is the input); recession-risk sizing (phase probability conditions drawdown budgets); earnings-cycle context (fa-guidance-estimates — estimate revisions ride the cycle); macro-data interpretation (event-jobs-report, event-cpi releases matter through their phase implications).

Risk profile & failure modes

  • Retrospective-date illusion (the core trap): backtests keyed to NBER dates use information that did not exist in real time — the documented gap between dated-phase returns and real-time-proxy returns is large (quant-backtest-hygiene point-in-time discipline).
  • Phase-identification disagreement: 2022-23 produced “rolling recession,” “no landing,” and recession calls simultaneously from standard proxies — the frame degrades exactly when positioning depends on it most.
  • Cycle-shape drift: post-1980s cycles lengthened (the Great Moderation, documented); service-economy and policy-response changes mean historical phase durations are weak priors.
  • Rotation-grid overconfidence: the sector-phase conventions are practitioner lore with mixed formal evidence — the KB requires them labeled as such wherever used.

Evidence & limits

NBER’s own documentation defines the dating procedure and its lags; the equity-lead and phase-rotation regularities are documented with wide dispersion. The frame’s value is CONDITIONING (what usually leads, what usually lags), not prediction; every use inherits the real-time identification problem.

Falsifiable-thesis examples

Illustrations only, not signals:

  • “Equity troughs lead NBER-dated recession troughs by 3+ months in the next recession (market-lead check)” — falsified by the dated sequence.
  • “A Sahm-rule trigger (unemployment 3-month average +0.5pt off its low) coincides with an eventual NBER-dated recession start within 6 months (real-time proxy audit)” — falsified by a false-positive trigger (2024’s near-trigger is the live caution).

Cross-references

  • The leading signals: macro-yield-curve, indicator-credit-spreads
  • The trading application: strategy-sector-rotation
  • The regime siblings: regime-rate-environments, macro-inflation-linkages
  • The data feed: event-jobs-report

The agent cites this page.

Inside the platform, this entry is live context. A signed-in citation opens the in-app view of the same id.

Inquire about founding membership