Corporate Portfolio Optimization

In annual budget meetings, most companies' capital allocation is a photocopy of the previous year. Every business unit gets a share close to last year's, growers a bit more, shrinkers a bit less. Nobody asks “do we have a reason to keep this business?” because the answer might be uncomfortable.

Corporate portfolio optimization is exactly the habit of asking that question regularly. The company's capital, management attention and best talent are limited. Where you put them matters more than how good you are in any one business.

When you need it

If you have several business units and their returns, growth rates and capital needs diverge clearly, portfolio management has already begun. If the group trades at less than the sum of its parts (the market calls this the “conglomerate discount”), or investors keep asking “why is this business with you?”, the signal is very strong.

Other signs: a business unit has earned below its cost of capital for years yet gets a budget every year, management attention goes to the smallest but noisiest businesses, or the core business's cash covers the gaps of old businesses instead of funding a new growth area.

It is too early when a single core business is still growing and side businesses are not yet meaningful. At that stage portfolio management adds bureaucracy.

What most teams miss

We usually do portfolio management with the question “which business is good, which is bad?” Tools like the BCG matrix and the GE-McKinsey grid make that question systematic but skip the more important one: “Is this business worth more in our hands than in another owner's hands?” This is what Goold, Campbell and Alexander call “parenting advantage.” If the center contributes something to a business unit, such as money, talent, customer access, brand or discipline, keeping it makes sense. If it does not, the unit may be better off with another owner, and selling is value creation.

By this logic even a “bad” business can be kept (if the center adds something to it) and even a “good” business can be sold (if another owner will add more). General Electric's rule under Jack Welch in the 1980s and 90s, “be number one or two in every business, or fix, sell or close it,” is the well-known example of this discipline. In later years GE kept simplifying its portfolio and in 2023-2024 split into three separate companies so each could follow its own capital logic. That story shows that portfolio decisions are not one-off but a continuous job.

Multi-sector groups with over a century of history, like Tata Group, also manage their portfolios with shared values, brand and a long-term ownership mindset. Even there the question is the same: does the common umbrella give every business a real benefit?

A portfolio decision is an honest answer not to “is this business good?” but to “are we the best owner of this business?”

This strategy differs both from the strategy inside individual businesses and from acquisition decisions: the subject here is the capital allocation logic that sits above all the businesses.

The playbook

1. Build a one-page economic table for every business unit

Growth, return on invested capital (ROIC), its spread over the cost of capital, capital intensity and cash generation or consumption. Make shared cost allocation transparent; hidden cross-subsidy is the source of most portfolio problems.

2. Apply the “parenting advantage” test

For each business unit, write the concrete contribution the center provides and the cost it creates. If the center's contribution cannot be proven, either increase the unit's autonomy or evaluate ownership alternatives (sale, separation, joint venture).

3. Classify businesses by role

Grow, maintain-and-harvest, fix, or exit. Each role calls for a different target and management rhythm. If a cash-generating business is managed to a growth target and a new business to a profitability target, both break.

4. Write the capital allocation rules

Which threshold of investment is approved by which committee? How long is the fix period? What percentage of the budget is reallocated each year as part of planning? Having the rules in writing reduces the “whose voice is louder” debate.

5. Build exit discipline

For businesses whose fix period has expired, record the sale or closure decision explicitly. Citing past investment (sunk cost) and fear for employees delay exits. The sale process should also be planned fairly for buyers and employees.

What to measure

• ROIC spread over cost of capital by unit: In which businesses is capital creating value and in which destroying it?
• Capital reallocation rate: How much of annual capital goes to a different place from the year before? If near zero, you are not doing portfolio management.
• Share of capital, and time, held in businesses earning below the cost of capital: Shows the price of indecision.
• Center cost relative to business unit contribution: An indirect measure of whether parenting advantage really works.

Pitfalls

Equal distribution. Spreading capital “fairly” across units leaves the best businesses hungry and the worst fed. Portfolio management is deliberately unequal distribution.

Sunk cost captivity. The logic “we invested so much, we cannot let go now” pours resources into losing businesses. Talk about the investment's future, not its past.

Selling the cash cow too early. Selling a stable business that funds today's new ones, at a short-term attractive price, dries up the financing of the growth plan. Calculate the sale's effect on portfolio structure first.

Start this week

1. Put the ROIC of all your business units and its spread over the cost of capital in a single table; add shared cost allocation transparently.
2. Spend two hours with the management team on the “parenting advantage” question for your three smallest and three largest units: what does the center add here?
3. Compare capital allocation over the past three years and calculate the reallocation rate; present the result to the board.