Leading, lagging and coincident indicators

If the peak and trough are only visible in hindsight, how does anyone position ahead of a turn? By reading a battery of indicators sorted by when in the cycle they move relative to the overall economy. This taxonomy (leading, coincident, lagging) is the single most useful organizing idea for real-time cycle analysis, and the Conference Board packages the US versions into published composite indexes (the Leading Economic Index, LEI; the Coincident Index; and the Lagging Index). Learn the logic, not just the lists.

Coincident indicators: these move with the economy; they are the cycle, roughly in real time. They tell you where you are now. The big four:

• Real GDP: the broadest measure of output (but quarterly and lagged in publication).
• Nonfarm payroll employment: jobs; the most-watched monthly coincident series, and one of the NBER's two favourites.
• Industrial production (IIP in India): physical output of factories, mines, utilities; monthly and timely.
• Real personal income less transfers and manufacturing & trade sales: income and the flow of goods through the economy.

When these are rising broadly, you are in expansion; when they are falling broadly, you are in contraction. They confirm the current state — they do not warn.

What are leading indicators?

Leading indicators: these move before the economy; they predict the turn. They are the analyst's early-warning system, and each leads for an understandable reason:

• The yield curve (10-year minus 3-month, or 10y−2y). The premier leading indicator (M7.03). An inversion (short rates above long rates) has preceded every US recession of the last half-century, typically by 6 to 18 months. The logic: the curve inverts when the market expects the central bank to cut rates in the future (i.e., expects a slowdown) even as the central bank holds them high now to fight inflation. It is the bond market forecasting the recession the central bank is trying to engineer. (Caveats matter. It has given false or very early signals, and the re-steepening after an inversion often marks the recession actually arriving. But no single indicator has a better record.)
• Building permits (and new housing starts). Housing is the most interest-rate-sensitive part of the economy, so it turns first when rates change: a builder pulls a permit months before ground breaks and a year before the house sells. Permits falling is an early growth warning; permits turning up is an early recovery signal.
• New orders for manufactured goods (especially new orders for nondefense capital goods ex-aircraft, the "core capex orders", and the ISM new-orders sub-index). Orders precede production precede shipments precede revenue; a firm's order book is its own near-future. This is why the PMI leads.
• The stock market (S&P 500; Nifty). Equities are a claim on future profits, so the market prices the expansion or recession before it arrives. Famously, "the stock market has predicted nine of the last five recessions" (Samuelson's joke about its false positives), but it does lead, typically by ~6 months. Falling equities are a (noisy) growth warning.
• Credit spreads (the extra yield on corporate bonds over government bonds; and bank lending-standards surveys like the US SLOOS). When lenders get nervous, spreads widen and standards tighten before the defaults actually arrive. The credit market smells the downturn early. Widening high-yield spreads are one of the most reliable stress signals (this is the credit cycle turning).
• Average weekly manufacturing hours and initial jobless claims (inverted): employers cut hours before they cut heads, and the first uptick in weekly jobless claims leads the rise in the unemployment rate (a lagging series) by months. Consumer expectations surveys round out the set.

What are lagging indicators?

Lagging indicators: these move after the economy; they confirm a turn that has already happened. They are the ones the untrained eye watches, and watching them is how amateurs get the cycle backwards:

• The unemployment rate. This is the classic trap. Unemployment peaks well after the recession has ended: firms only rehire once they are convinced the recovery is real, so the jobless rate keeps rising for months into the recovery. Waiting for unemployment to fall before turning bullish means buying the recovery a year late.
• CPI / consumer inflation. Inflation is a lagging indicator of the cycle: it peaks after the economy peaks (the economy overheats, then prices catch up) and troughs after the economy troughs. This is why the 1970s were so dangerous (see the Volcker shock) and why a central bank fighting inflation is always, in a sense, fighting the last war.
• Unit labour costs (wages per unit of output): wages are sticky and negotiated with a lag, so labour costs keep rising into a downturn, squeezing margins exactly when revenue falls.
• The average duration of unemployment, commercial & industrial loans outstanding, the inventories-to-sales ratio, and the prime/bank lending rate — all confirm, none predict.

How do you build a cycle read from them?

How to build a cycle read (the skill the app drills, and the worked example on a mixed indicator panel shows in full). You do not read one indicator; you read the pattern across the three families, because their alignment tells you the stage:

• Leading up, coincident up, lagging still down/flat → early expansion / recovery (the best time to add risk).
• All three rising, lagging now catching up (inflation and unemployment-duration turning) → mid-to-late expansion (add caution as leading indicators lose steam).
• Leading rolling over (curve inverts, permits fall, orders soften, spreads widen) while coincident is still strong and lagging (inflation) is high → the peak / late cycle, the most dangerous and most misread moment, because the numbers everyone watches (jobs, GDP, inflation) still look great even as the forward-looking numbers have already turned. This is where cycle analysis earns its keep.
• Leading and coincident both falling, lagging (unemployment) still rising → contraction / recession.
• Leading turning up (curve re-steepens, permits and orders bottom, spreads narrow) while coincident and lagging still fall → the trough, the second most misread moment, because the news is uniformly terrible exactly when the forward indicators say the worst is passing.

The through-line: the leading indicators turn first and are what you act on; the coincident indicators tell you where you are; the lagging indicators are what the crowd (and the headlines) fixate on, and fixating on them puts you a full half-cycle behind.