Capital Allocation Takes More Than Good Instincts
Only 30% of multi-business companies are worth more than the sum of their parts, and the best allocators get there by overriding three leadership instincts: they concentrate capital instead of spreading it evenly, back future potential instead of past winners, and hold investment steady through dips.
Summary
BCG compares nearly 600 listed multi-segment companies (2015-2024) with sum-of-the-parts benchmarks; the 30% that beat them earn 1.6 points higher returns on investment. They concentrate capital by giving each business an explicit portfolio role, with an 18% wider spread in investment intensity. They decouple funding from past cash generation (27% lower correlation) using milestone-gated tranches with kill criteria. And they hold through downturns, with 30% less volatile investment, cutting a declining segment in about half of cases versus seven in ten for weaker allocators.
Editor's opinion
Methodology (EV/Sales sum-of-parts vs. peer medians by 3-digit SIC) is sound given segment-reporting constraints, but 3 issues qualify the findings: 1) Skill vs. segment quality conflation – EV/Sales ignores margin/growth gaps within a SIC code, so some of the "outperformer" premium may reflect structurally better segments, not allocation discipline, inflating the behavioral effect sizes. 2) Correlation, not isolated causation – "actual value" is public market EV, which also prices governance, complexity discount and general management quality; these aren’t separated from allocation skill, so part of the value link is likely confounded by good managers doing everything well, not just capital allocation. 3) Sample/period bias – thin pure-play peer sets under coarse SIC codes, survivorship bias ($5B+ public-only sample), and a 2015-2024 window dominated by a V-shaped COVID recovery all bias toward validating “hold through the dip” and may overstate how generalizable the magnitudes (1.6pp, 18%, 27%, 30%) are outside this regime.
Key concepts
Sum-of-the-parts benchmark · Portfolio roles (growth-share matrix) · Milestone-gated tranches · Structural-break markers