Portfolio Strategy: Allocating Across Business Units

Your task as CFO is to impose the discipline that the external market would impose—but with the informational edge that headquarters uniquely possesses. That means answering one question for every unit, every cycle:

> Is the incremental dollar deployed in this business earning more than it would in the next-best alternative, adjusted for risk?

Note the word *incremental*. The historical returns of a unit are sunk. What matters is the return on the *next* dollar. This is where CFOs get trapped: a division showing 18% average ROIC can still be destroying value at the margin if its incremental projects earn 6% against an 9% cost of capital. Blended averages hide the marginal truth.

Divisional Cost of Capital: Stop Using One WACC

You already know how to compute WACC. The advanced move is refusing to apply a single corporate WACC across a multi-business portfolio. Using one hurdle rate systematically overfunds high-risk businesses and starves low-risk ones, because the corporate rate is a blend that fits neither.

The practical method:

1. Identify pure-play comparables for each business unit—publicly traded firms that operate only in that segment.

2. Unlever their equity betas using each comp's capital structure, average them, then relever to your target capital structure for that unit.

3. Assign a business-unit-specific cost of debt reflecting the unit's asset risk and cash-flow volatility—an infrastructure unit borrows cheaper than a software venture.

4. Build the divisional WACC from these inputs, and layer in country risk premia where units span geographies.

A media conglomerate might run a 7% hurdle for its cable infrastructure unit and a 13% hurdle for its streaming venture. Apply the corporate blend of 10% to both, and you underinvest in the safe annuity while pouring capital into a venture that isn't actually clearing its true risk-adjusted bar.

The Allocation Framework: Sorting the Portfolio

Once each unit has its own hurdle rate, you can mapmapUsing software to automate repetitive marketing tasks and campaigns, enabling personalisation at scale across channels like email, web, and social.Voir la définition complète → the portfolio on two axes that actually drive value: economic profit generation (ROIC minus divisional WACC, times invested capital) and incremental return on growth capital (what the next dollar earns).

This produces four archetypes, and each demands a different capital posture:

Compounders — High spread, high reinvestment returns

These units earn well above their cost of capital *and* can absorb more capital at the same high returns. This is where growth capital belongs. The discipline here is not restraint—it's *ambition*. CFOs often under-fund their best businesses because those businesses aren't complaining. Silence is not a signal to stop investing.

Cash Cows — High spread, low reinvestment returns

Mature, profitable, but with declining incremental returns. The correct posture is to harvest: run them for cash, invest only maintenance capital, and route the free cash flowfree cash flowFree Cash Flow is the cash a company generates from operations after funding the capital expenditures needed to maintain and grow its asset base.Voir la définition complète → to compounders or to shareholders. This is where the hardest discipline lives, and we'll return to it.

Value Traps — Low spread, low reinvestment returns

Below cost of capital, no path to improvement. These are candidates for fix, sell, or exit. The GE Power lesson lives here: a business can be large, iconic, and central to your identity while systematically destroying value. Size is not a defense.

Turnarounds — Low current spread, high potential reinvestment returns

Currently underwater but with a credible thesis for clearing the hurdle. Fund these *conditionally*, with milestone-based tranches, not open-ended commitments.

The Two Diagnostics That Matter Most

Before you allocate a single dollar, run two tests:

The self-funding test. Which units generate more cash than they consume, and which are net absorbers? A healthy portfolio has cash generators funding cash consumers *by design*, not by accident. If your cash cows are quietly reinvesting their own output instead of remitting it upward, your internal capital market is broken—the cash never reaches the allocator.

The "would we buy it today" test. For each unit, ask: at its current invested capital, would we acquire this business at book value today? If not, you are holding it out of inertia. This reframes retention as an active decision. Every quarter you keep a value-destroying unit, you are re-buying it.

The Hardest Discipline: Starving the Mature to Fund the Growth

Here is where portfolio strategy stops being an analytical exercise and becomes an act of organizational courage.

The mature units are run by your most senior, most credible executives. They deliver the earnings that make your guidance. They employ the most people. Their leaders sat next to you at the last three offsites. When you propose cutting their capital allocation to fund a growth bet run by someone half their tenure, you are not making a spreadsheet decision—you are picking a fight with your own power structure.

And the numbers seem to defend the mature unit. It has *lower risk*, *proven cash flows*, and *established returns*. But those are averages, and averages are the enemy. The mature unit's *incremental* investments—the tenth distribution center, the incremental capacity in a saturated market—earn far less than the compounder's next dollar. The reallocation is correct precisely when it feels most uncomfortable.

Amazon made this its founding discipline: for years, retail cash funded AWS when AWS was an unproven bet, and later AWS's profits funded logistics and devices. The internal capital market moved capital *toward* the high-incremental-return frontier, not toward the incumbent that generated the cash. Netflix did the reverse of GE—it deliberately starved its profitable DVD business to fund streaming, cannibalizing its own cash cow before the market forced the issue.

How to Execute the Starve—On Monday Morning

The framework is clean; the execution is a knife fight. Here is the practical playbook:

1. Separate maintenance capital from growth capital in every budget. Mature units are entitled to maintenance capital—the spend required to hold current position. They are *not* entitled to growth capital. Make this distinction explicit in the budget architecture so that "we need to keep investing" cannot smuggle low-return growth spend through under the banner of necessity.

2. Make hurdle rates non-negotiable and public. When each unit knows its own risk-adjusted hurdle and knows that projects below it will not be funded, the debate shifts from politics to numbers. The CFO's leverage is that the hurdle rate is *the rule*, not this quarter's opinion.

3. Tie incentives to economic profit, not growth or absolute earnings. If a division head is compensated on revenue or reported profit, they will hoard capital regardless of returns. Compensate on economic profit (return above the divisional cost of capital), and a mature-unit manager will *voluntarily* return capital they can't deploy above the hurdle—because holding underperforming capital now hurts their scorecard.

4. Sunset commitments. Growth funding for turnarounds and new ventures should be tranched against milestones. Mature-unit maintenance capital should be zero-based annually, not rolled forward. The default should be "justify the spend," not "continue last year."

5. Force the exit conversation. Once a year, run the "would we buy it today" test at the board level for every unit. Naming a value trap out loud, in front of the board, is what breaks the inertia. Silence protects the incumbent.

The failure mode is always the same: the CFO knows the analysis, agrees with the reallocation, and then loses the will in the room when the senior division head pushes back with tenure and relationships. Portfolio discipline is 20% analysis and 80% the willingness to be unpopular with powerful people who are, on the numbers, wrong.

Managing the Portfolio Through the Cycle

Allocation is not a once-a-year event tied to the budget calendar. The best CFOs run a continuous reallocation process and treat the portfolio as dynamic.

Reallocate actively. McKinsey's long-running research on capital reallocation found that firms that meaningfully shift capital across businesses year to year deliver substantially higher total shareholder returns over the long run than "sticky" allocators whose divisional budgets barely move. Most companies allocate this year's budget as last year's budget plus or minus a few percent. That inertia is the single largest destroyer of value in multi-business firms—and it is entirely self-inflicted.

Vary posture with the cycle. In a downturn, the instinct is across-the-board cuts—the "peanut butter" approach that spreads pain evenly. This is precisely wrong. Downturns are when the *relative* attractiveness of businesses diverges most, and when starving the wrong unit does the most permanent damage. Protect the compounders, accelerate the exits, and use the disruption to acquire assets that strengthen the growth frontier.

Watch for the diversification discount. If the market values your sum-of-the-parts below your enterprise value, investors are telling you the portfolio destroys value through cross-subsidization or complexity. That discount is a direct verdict on your internal capital market. The answer may be divestiture, spin-off, or split—and a CFO who understands portfolio strategy sees the break-up option not as failure but as the ultimate act of allocation.

Key Takeaways

• Use divisional hurdle rates, never a single corporate WACC. Applying one blended rate systematically overfunds risky units and starves safe ones. Build unit-specific costs of capital using the pure-play method.
• Allocate on incremental returns, not historical averages. A unit with strong average ROIC can be destroying value at the margin. The only question that matters is what the *next* dollar earns against that unit's risk-adjusted hurdle.
• Separate maintenance capital from growth capital in the budget architecture. Mature units are entitled to the former, never automatically to the latter. This one structural change defeats most low-return "we must keep investing" arguments.
• Compensate division heads on economic profit above their cost of capital. This makes managers voluntarily return capital they cannot deploy above the hurdle, aligning the internal capital market without a fight.
• Run the "would we buy it today" test annually at board level. Retention is an active decision; naming value traps out loud breaks the inertia and cross-subsidization that quietly destroy multi-business firms.