Chapter 42: Key Economic Indicators
Core idea
Economic indicators are top-line numbers that summarise the state of the economy. The most useful classification splits them into three groups by timing: leading indicators change before the economy turns and so help forecast it (housing starts, durable-goods orders, consumer confidence); coincident indicators move with the economy and confirm what is happening now (GDP, industrial production, personal income); lagging indicators move after the turn and confirm what has happened (the unemployment rate, the prime interest rate, business inventories that have already cleared). No single indicator is reliable on its own — but a basket of leading, coincident, and lagging numbers read together gives you a far better real-time picture than any one number.
Author’s argument: Numbers in isolation are noise. Numbers in combination — over time, against benchmarks — are signal.
Why it matters
Forecasting requires the right kind of number
People constantly try to predict the future using the unemployment rate, which lags the economy by several quarters. By the time unemployment is rising fast, the recession is already happening. If you want a forecast, you need leading indicators (housing starts, ISM new orders, the Conference Board Leading Economic Index). Knowing the timing classification of every indicator is what separates real-time analysis from rear-view mirror analysis.
Markets react to indicator releases
Stock and bond prices move sharply on indicator announcements — non-farm payrolls, CPI, retail sales, FOMC decisions. Traders who know which release matters this week, what the consensus expects, and what a beat or miss tends to imply have a real edge. Even non-traders benefit from the same calendar awareness: it tells you why bond yields jumped on a Friday morning.
Some “weird” indicators work because they capture behaviour numbers miss
The Big Mac Index gauges currency over- or undervaluation by comparing what the same burger costs in different countries. The Cardboard Box Index tracks shipping demand. The Library Indicator rises in recessions because people substitute free entertainment for paid. None of these is rigorous, but each captures real behavioural information that the official data takes months to confirm.
Key takeaways
Key takeaways
- Three timing classes: leading (predict turns), coincident (move with the economy), lagging (confirm a turn after the fact).
- Leading: durable-goods orders, housing starts, consumer confidence (CCI), new-business permits, stock-market indices, ISM new orders, yield curve.
- Coincident: real GDP, industrial production, personal income, manufacturing-and-trade sales — they move with the economy in real time.
- Lagging: unemployment rate, average duration of unemployment, prime rate, ratio of inventories to sales, change in unit labor costs.
- Indicators are most useful in context — trends over time, comparisons across countries, and against explicit benchmarks (the Fed's 2% inflation target, the 4–5% 'natural' unemployment rate).
- Offbeat indicators (Big Mac, hemlines, lipstick, cardboard boxes, men's underwear) sometimes work because they pick up consumer behaviour faster than official statistics.
- The Big Mac Index uses a constant good across countries to assess relative purchasing power — useful for gauging if a currency is over- or undervalued.
Mental model — the indicator taxonomy
Read it as: indicators sorted by when they move relative to the economy. Leading turns first, coincident moves with it, lagging confirms. Offbeat indicators are unofficial but sometimes informative — useful for triangulation rather than headline forecasting.
Mental model — how the indicators arrive in time
Read it as: time flows left-to-right. Leading indicators move while a turn is still coming; coincident indicators move at the turn; lagging indicators only move once the turn is done. Using a lagging indicator (like unemployment) as a forecast is a category error — by the time it moves, you’ve missed the call.
Practical application
Building a personal dashboard
Reading an indicator the right way
-
Check the trend, not the level. “Unemployment is 4%” matters less than “unemployment has risen from 3.4% to 4% over six months.”
-
Compare to a benchmark. Inflation at 3% means one thing if the Fed’s target is 2% and another if the target is 4%.
-
Look at the consensus expectation. Markets price in the expected number; what moves prices is the surprise.
-
Combine timing classes. A leading indicator turning while a coincident indicator stays strong tells you a turn is forming. Both turning together is a much stronger signal.
-
Watch for revisions. Initial releases get revised — sometimes substantially — in subsequent months. The first print is a draft.
Example: spotting a turn in 2007
Run the indicator playbook against the months leading into the 2008 recession.
- Mid-2006: the yield curve inverts (leading). Most commentators dismiss it.
- Late 2006 to mid-2007: US housing starts roll over from 2.3M annualised to 1.4M (leading).
- Mid-2007: ISM Manufacturing dips below 50 (leading).
- Summer 2007: Consumer confidence drops sharply (leading).
- Late 2007: Real GDP growth slows but stays positive (coincident — not yet confirming).
- December 2007: NBER later identifies this as the recession start, but at the time only payroll data is wavering (coincident).
- 2008 throughout: unemployment climbs from 5% to 7% (lagging — by the time this is unmistakable, the recession is six months old).
- Late 2008: unemployment hits 10% (lagging — confirms what leading indicators flagged 18+ months earlier).
The story isn’t that the recession was easy to call — there was real uncertainty in real time. The story is that the leading indicators were doing their job all along. Anyone watching housing, the yield curve, and consumer confidence in 2006–2007 had the information; many simply discounted it.
Caveats
Related lessons
Jump to…
Type to filter; press Enter to open