Markets · Strategy 11 April 2026 ~5 min read

Defensive Rotation: Why Quality Outperforms Late-Cycle

Earnings dispersion is widening while real rates stall. The historical pattern is clear: high-quality compounders extend their relative lead in the final third of the cycle.

The market is debating the wrong question. Most desks are still arguing whether we are in early-cycle or mid-cycle territory. The data — earnings revisions, credit spreads, the shape of the curve — has quietly resolved the question in favour of late-cycle. The more useful exercise is not to debate the label but to position for what late-cycle reliably produces.

Thesis

The quality factor — high return on invested capital, low leverage, stable margins — has historically delivered its strongest relative returns in the final 12–18 months of the cycle. The current setup is consistent with that regime, and positioning has not yet rotated.

What the data says

Three observations make the case mechanically. First, earnings dispersion across the index is at a multi-year high — the spread between top-quartile and bottom-quartile EPS revisions has roughly doubled since the trough. Second, real yields have stalled in a range that historically rewards balance-sheet strength over operating leverage. Third, credit-spread compression has plateaued: the easy beta is gone, and what remains is idiosyncratic.

Each of these, in isolation, is unremarkable. Together they describe a regime in which the market begins to discriminate sharply between businesses that compound through volatility and those that simply ride the macro tide.

Why quality, mechanically

The compounding spread widens

In a low-dispersion environment, the market pays for beta. In a high-dispersion environment, the market pays for the franchise — the business that grows EBITDA regardless of which way the macro turns. This is not a behavioural story. It is the consequence of cash-flow mathematics interacting with a rising cost of capital.

Defensive balance sheets get re-rated

When refinancing risk becomes a real input to fundamental valuation, the implied cost-of-equity for over-levered names rises faster than for cash-rich names. The relative DCF mechanics do the work without any narrative help.

Late-cycle is not a regime that rewards cleverness. It rewards the mundane: strong cash conversion, low refinancing exposure, durable pricing power. These are the boring inputs that always end up mattering most.

Implementation

The cleanest expression is at the factor level: long quality versus short the broader market, sized to neutralize sector and country tilts. For investors constrained to long-only mandates, the equivalent is an overweight to the top quintile on a composite score combining ROIC, net-debt-to-EBITDA, and gross- margin stability.

Two cautions. First, "quality" as labelled by ETFs and screens has drifted. Construct the basket from primary metrics, not from index methodology. Second, valuation discipline still applies — the quality factor underperforms when it becomes the consensus crowd. The signal here is that positioning has not yet rotated, which is what makes the trade interesting.

Risks to the thesis

None of these break the structural argument. They are mostly about timing — the question of when the rotation accelerates, not whether it occurs. Position sizing should reflect that distinction.