The business cycle
Drive the cycle, watch the indicators peel apart
Four phases run clockwise. Early expansion (top-right): output rising off a trough, slack everywhere, inflation low, central bank still easy. Mid expansion (bottom-right): growth at trend, hiring strong, inflation drifting up, policy approaching neutral. Late / peak (bottom-left): capacity constrained, wages accelerating, policy tight, curve flattening. Contraction (top-left): output falling, unemployment climbing, inflation rolling over, policy pivoting back to ease.
Drag the hand. The phase description below the dial updates. So do the eight indicators on the right.
The right column is grouped by leading, coincident, lagging. Each indicator's sparkline shows its profile across the full cycle; the gold dot marks where it sits right now.
The point: indicators don't peak at the same angle. Yield-curve slope peaks early — it's already flattening by mid-expansion. Unemployment peaks last — it's still rising during early recovery. This phase-offset is the entire reason watching one indicator is dangerous and watching a dashboard is sane.
Park the hand near the boundary between mid-expansion and late expansion — about two-thirds of the way around. Look at the indicators. Yield-curve slope is already near zero. Building permits have rolled over. Credit spreads are starting to widen. But: industrial production is fine, retail sales are strong, unemployment is at a multi-decade low.
This is the moment commentators say "the economy is incredibly strong, recession fears are overblown." They are watching the coincident indicators. The leading indicators have already turned. By the time the coincident indicators turn, the leading indicators have inverted hard.
Hit ▶ Play. Watch one full revolution. Notice the wave structure: leading indicators sweep their high and low first, then coincident, then lagging. The same rotation produces a different reading from each gauge depending on where you check it. The cycle is one thing measured eight ways.
Pundits regularly declare the business cycle obsolete. The Great Moderation (1985–2007), then COVID-as-transitory, then the soft-landing narrative of 2024–25. The cycle keeps coming back because the mechanism that creates it — agents extrapolating recent conditions and over-investing in capacity that takes years to come online — is human, not technological. Bigger central banks and better data shrink the amplitude. They don't remove the wave.
Different decisions belong to different parts of the cycle.
Buying a house when the labour market is at its tightest is buying at peak affordability illusion. Negotiating a raise in early recovery is impossible; doing it in mid-expansion is easy. Refinancing decisions cluster around late-cycle peaks in rates. The phase you're in is more informative than the absolute level of any one variable.