The New Magic Formula: Further Testing And Evolution Of Optimal Model

The QCF model, based on ROCE and P/FCF, is said to consistently outperform the S&P 500 in backtests when using a 15% ROCE threshold. The article argues that adding extra filters such as revenue growth, debt, dilution, or momentum reduces both returns and consistency. The takeaway is that a simpler quality-focused framework appears to produce better risk-adjusted results.

Analysis

The important takeaway is not that “quality wins,” but that the market is likely underpricing cash-flow durability as a scarce factor in a slower-growth, higher-rate regime. A 15% ROCE hurdle effectively isolates businesses with reinvestment runway and pricing power, which should keep outperforming while capital remains expensive and dispersion stays high. The second-order effect is that capital may keep migrating away from revenue-growth stories toward businesses that can self-fund buybacks, acquisitions, and capex without external financing.

The counterintuitive point is that adding “common sense” filters can destroy edge because many of them are already partially embedded in price, or they narrow the universe too far and increase style crowding. Momentum and growth screens often push investors toward the same names at the wrong phase of the cycle, while leverage/dilution filters can exclude distressed turnarounds that still compound free cash flow. In other words, the cleaner the signal, the less exposed it is to narrative noise and regime-specific overfitting.

For competitive dynamics, this favors high-ROCE incumbents over capital-hungry challengers, especially in software, semis, and industrials where reinvestment efficiency matters more than top-line growth. The losers are companies relying on repeated equity issuance or aggressive M&A to manufacture growth; if investors rotate toward simple quality, their cost of capital rises and strategic flexibility shrinks. The main risk is valuation compression if the market crowds into the same quality basket, particularly if real yields rise or earnings breadth improves and beta reasserts itself over the next 3-6 months.

Contrarian risk: this model may be most effective precisely because it avoids trying to time cycles, but the flip side is that it can underperform during sharp reflation or stimulus-driven rallies when low-quality cyclicals rip. If macro liquidity improves materially, the backtest advantage could flatten for several quarters even if it remains intact over a full market cycle. The key question is whether investors are paying too much for complex signals when the marginal dollar still rewards simple, durable free cash flow.

Key Decisions for Investors

• Build a long-only basket of high-ROCE / low-P/FCF names versus the S&P 500 over a 6-12 month horizon; use equal-weighted exposure to reduce single-name valuation risk and harvest the structural factor premium.
• Pair trade: long quality cash-flow compounders (e.g., MSFT, ADBE, SNPS) / short expensive growth names with weak FCF conversion; target 10-15% relative outperformance if rates stay elevated for the next 2 quarters.
• Avoid adding leverage or momentum overlays to quality screens in live portfolios; the backtest suggests these filters reduce breadth and increase regime sensitivity, likely worsening hit rate over the next 12 months.
• Consider selling out-of-the-money puts on a diversified quality ETF or basket during risk-off periods; the model implies pullbacks in high-ROCE names are more likely to mean-revert than structurally break down.
• If rotating out of beta, stage entries on quality names after earnings rather than chasing breakouts; the edge appears to come from fundamentals, not price confirmation, so wait for post-print dislocations.