Return on invested capital, often abbreviated ROIC, measures how productively a company converts each dollar of operating capital into operating profit. High ROIC is the quantitative fingerprint of business quality — a company that earns 25 cents of operating profit on each dollar of invested capital is structurally different from a company that earns 6 cents. Earnings yield is the inverse of the price-to-earnings multiple — a stock with a 10 multiple has a 10 percent earnings yield, and earnings yield is the parsimonious quantitative fingerprint of price-cheapness.
| Factor combination | What the strategy targets | Why the combination matters |
|---|---|---|
| High ROIC alone | Quality businesses regardless of price | Tends to identify durable franchises but does not protect against paying any price for quality |
| High earnings yield alone | Statistically cheap stocks regardless of business quality | Tends to identify Graham-style cheapness but does not protect against value traps where the business is genuinely deteriorating |
| Combined two-factor rank | Quality businesses available at a reasonable price | The combination protects against both ends — paying any price for quality and buying cheapness without quality — and is the structural source of the strategy's long-run edge |
The mechanical implementation is unusually simple. Rank every investable stock in the chosen universe on each factor independently, sum the two ranks, buy a basket of the top names — typically 20 to 30 holdings — hold each name for roughly a year, then refresh the ranking and rebalance. Greenblatt's preferred implementation excludes financials and utilities and applies a minimum market-cap floor; small variations in the construction rules do not materially change the long-run results.
Combined Rank Score = Rank On Earnings Yield + Rank On Return On Invested Capital
The well-known underperformance windows. The strategy's long-run edge is not a steady year-over-year drip; it shows up as a cycle of strong years interrupted by multi-year stretches of underperformance. The 2010 to 2014 period is the most-documented modern example — the strategy lagged the broad market by several percentage points per year for several consecutive years, primarily because the post-crisis bull market was led by quality growth names whose multiples were already high enough to keep them out of the high-earnings-yield half of the rank. Earlier multi-year underperformance windows show up in the late 1990s late-cycle technology run and in shorter stretches throughout the 1980s. The frequency of multi-year underperformance is structurally embedded in the strategy and is not a sign that the strategy is broken when it occurs.
What the underperformance windows do to professional implementations. Professional managers running the strategy inside an institutional vehicle face quarterly or annual performance evaluations from clients who often cannot tolerate three to four consecutive years of lagging the broad market, regardless of the strategy's long-run record. The structural mismatch between the strategy's cycle and the typical professional evaluation cycle is the load-bearing reason so few professional managers run the strategy in its pure form, even though the long-run evidence base is real. The strategy works most cleanly for investors with the structural ability to commit to a full cycle — individuals investing their own capital, family offices, or explicitly long-term institutional vehicles.
Sit with the ideas.
A reader is told that a two-factor rank-screen combining return on invested capital and earnings yield generated meaningful long-run excess returns in the academic literature and the practitioner record. The reader plans to implement the screen and asks why so few professional managers actually run the strategy in its pure form. Which response best captures the practitioner consensus on the strategy?