The Business Cycle Sequence: Testing EPB's Four Economy Framework

The economy doesn't move as one unit—it moves in a predictable sequence. The Four Economy Framework divides the economy into Leading (money, credit, permits), Cyclical (construction, manufacturing), Aggregate (total employment, GDP), and Lagging (services) sectors that peak and trough in order. We test 45 years of data to see if this sequence actually predicts recessions, explains why the 2022-2025 yield curve inversion failed, and whether you can trade profitably on it.

The Problem with Standard Recession Indicators

In previous articles, we've tested unemployment, the yield curve, and credit spreads as recession predictors. Each works sometimes but fails other times. The yield curve inverted in 1998 with no recession. It inverted in 2022 and we still haven't seen recession by late 2025. Credit spreads spiked in 2011 and 2016 without recession. Unemployment is a lagging indicator that tells you recession has already arrived.

The Conference Board's Leading Economic Index (LEI) tries to solve this by combining 10 indicators: average workweek, jobless claims, building permits, stock prices, yield curve, consumer expectations, and others. But LEI mixes forward-looking indicators (building permits) with coincident indicators (jobless claims) and even includes stock prices, which we know don't reliably lead recessions anymore.

What if there's a better way? What if the economy doesn't move as one unit, but rather in a predictable sequence where some sectors always lead and others always lag? This is the hypothesis proposed by Eric Basmajian at EPB Research, a private macro research firm. In a series of articles published in early 2025, Basmajian argues that the economy should be analyzed as four distinct layers that peak and trough in a mechanical sequence. Let's test whether this framework actually works using our own data.

EPB's Four Economy Framework: The Theory

Basmajian's framework divides the economy into four layers that supposedly move in sequence:

The key insight: you can't just track one indicator. You need to watch the sequence. If Leading indicators peak but Cyclical Economy never actually contracts, there's no recession (1998, 2016, 2022-2025). If both Leading and Cyclical turn down together, recession is nearly certain.

Test 1: Do the Four Economies Actually Move in Sequence?

Let's verify the most basic claim: do these four economies actually peak and trough in order? We'll measure each economy using a composite index and track the timing of peaks before each recession since 1980.

Detailed Views: The Sequence in Each Recession

Zooming into each recession period reveals the sequence clearly. Watch how Leading (blue) peaks first, followed by Cyclical (orange), then Aggregate (green), then Lagging (purple). The timing is consistent across all business cycle recessions.

Test 2: Lead Times—How Much Warning Does Each Economy Provide?

Knowing the sequence exists is useful. But how much advance warning do you get? Let's measure the lead time from when each economy peaks to when the NBER-dated recession officially begins.

Test 3: False Positives—Does the Sequence Eliminate Them?

Our yield curve article found 1998 was a false positive: curve inverted, no recession. Our credit spreads article found 2011 and 2016 widened spreads without recession. Does tracking the full sequence eliminate these false alarms?

This explains the 2022-2025 mystery perfectly. The yield curve inverted in July 2022. Credit spreads widened modestly. Leading indicators softened. But construction employment grew throughout 2022-2025, and manufacturing stayed flat rather than contracting. The Cyclical Economy never turned down, so no recession occurred despite the yield curve inversion.

Test 4: Recession Magnitude—Does the Cyclical Economy Predict Severity?

Beyond just predicting IF a recession happens, can the Cyclical Economy tell us HOW BAD it will be? EPB Research claims the magnitude of the Cyclical decline predicts recession severity. Let's test this.

Test 5: Comparing to LEI—Which Framework Works Better?

The Conference Board's Leading Economic Index (LEI) is the standard recession prediction tool. It combines 10 indicators: average workweek, jobless claims, building permits, stock prices (S&P 500), consumer goods orders, vendor performance, new orders for capital goods, yield curve, consumer expectations, and credit conditions. The LEI treats all components with fixed weights in an index. How does EPB's sequence approach compare? We'll test predictive accuracy for the same recessions.

Note: We plan to write a dedicated article analyzing the LEI methodology and reproducing it with our own data. For now, we're using the Conference Board's published LEI values for comparison.

Test 6: Can You Trade Profitably on This?

The ultimate test: does knowing the sequence help you make money? We'll backtest a simple strategy:

• Defensive Signal: Leading Economy declining + Cyclical Economy starting to contract → sell stocks, buy bonds
• Offensive Signal: Leading Economy recovering + Cyclical Economy bottomed → buy stocks, reduce bonds
• Hold: All other times

Why Construction and Manufacturing Matter So Much

A common objection: "Construction and manufacturing are only 15% of GDP now. Services are 70%+. Why focus on small sectors?" The Four Economy Framework answers this definitively:

1. Amplitude, not share, drives recessions: Construction/manufacturing can swing +10% to -20% in annual growth (see the Cyclical Economy Index charts above: 1981-82 fell from 108 to 82 = -24%, 2008 fell from 120 to 78 = -35% annualized growth rates). Services rarely move more than ±2% (see the Lagging Economy lines, which are much flatter). A 20% decline in 15% of the economy is 3% GDP impact. A 2% decline in 70% of services is 1.4% impact. The volatile sectors drive the cycle.
2. Services only decline AFTER Cyclical job losses: Services employment is a lagging indicator. Restaurants, retail, healthcare don't fire people until AFTER construction/manufacturing workers lose jobs and stop spending. Tracking services tells you recession has arrived, not that it's coming.
3. Credit feedback loops start in Cyclical sectors: Construction workers defaulting on mortgages and manufacturing workers defaulting on auto loans create the banking stress that then spreads to the broader economy.

The 2022-2025 Case Study: Why No Recession (Yet)?

The yield curve inverted in July 2022. Fed hiked rates from 0% to 5.5% in 18 months. Credit spreads widened. Leading indicators softened. Yet as of late 2025, no recession has occurred. The Four Economy Framework explains why perfectly:

Limitations and Caveats

The Bottom Line

The Four Economy Framework works. Tracking the sequence—Leading → Cyclical → Aggregate → Lagging—predicts recessions with 100% accuracy when requiring BOTH Leading and Cyclical confirmation. It eliminates the false positives that plague single-indicator approaches. It explains recession magnitude. And it provides a tradeable signal with better risk-adjusted returns than buy-and-hold.

The key insight: construction and manufacturing are the pulse of the business cycle, not because they're large, but because they're volatile and interest-rate sensitive. Services employment, which everyone watches, is a lagging indicator that tells you recession has arrived, not that it's coming. Leading indicators (yield curve, money supply, permits) give early warning, but only matter if they translate into actual Cyclical Economy contraction.

For 2022-2025, the framework explains perfectly why the yield curve inversion "failed": Leading Economy signaled trouble, but the Cyclical Economy never declined. Construction boomed throughout, manufacturing stayed flat. Without Cyclical contraction, there's no transmission mechanism to the broader economy. The theory didn't fail—the sequence just never completed.

This framework synthesizes everything we've learned across our previous articles: unemployment lags because it's in the Lagging Economy, the yield curve leads but produces false positives without Cyclical confirmation, credit spreads widen when Cyclical Economy contracts, and the 2022-2025 episode makes sense when you track the full sequence. The Four Economy approach ties it all together into a coherent, predictive system.