How to Think Like a Capital Allocator

A rigorous guide to cash flow, fragility, incentives, and the discipline that separates durable systems from failing ones.

Most companies do not fail because they ran out of ideas. They fail because they kept funding the wrong ones. Every naira a business spends is a capital-allocation decision. Most managers do not know they are making one.

That sounds harsh. It is also precise. When conditions are favourable—when credit is easy, revenue is growing, and refinancing is available—weak capital allocation stays hidden for years behind headline profit. The income statement flatters the operator. The cash flow statement, read honestly, condemns the allocator. The difference between the two is where most of the real destruction of enterprise value takes place, and it is almost never discussed in those terms.

This post is a framework for reading that difference. It is aimed at analysts, executives, and serious students of business who want to move beyond surface-level financial commentary toward the kind of structural thinking that actually explains why systems succeed or collapse.

Executive Summary

To think like a capital allocator is to judge every decision by its effect on long-term intrinsic value, cash generation, and resilience. It means moving beyond headline profit and asking how cash is produced, what must be reinvested to sustain it, and which fixed obligations or external bottlenecks can break the loop. The framework I use has two components: an allocator’s lens — five diagnostic questions applied to any business — and a fragility scanner, a short test that separates businesses with structural durability from those whose health is conditional on conditions remaining favourable.

The Score That Matters

When I read a set of accounts, I am not looking first for revenue growth or reported profit. I am looking for the movement of cash: where it originated, what claims exist on it, and what quality of decision management has made with the residual.

An operator can improve throughput, launch products, and build market share. An allocator decides where the next unit of capital goes: maintenance capex, growth capex, inventory, receivables, debt reduction, distributions, buybacks, acquisitions, or liquidity buffers. These are not equivalent. They produce different future cash flows, different obligations, and different levels of fragility under stress.

The governing objective should not be size. It should be long-term per-share intrinsic value. Berkshire Hathaway’s owner principles are explicit on this: acquisitions matter only if they raise per-share intrinsic value, and intrinsic value is best understood as the discounted cash that can be taken out of a business over its remaining life.[footnoteRef:1] The current capital-allocation framework reinforces the same logic: deploy only where risk-adjusted returns can grow intrinsic value per share over long horizons, and retain earnings only when each dollar kept is likely to create more than one dollar of market value.[footnoteRef:2]

This distinction matters because accounting profit can flatter structurally weak businesses. Aswath Damodaran’s treatment of cash flow remains one of the clearest correctives: moving from operating earnings to free cash flow requires subtracting both capital expenditure and changes in non-cash working capital. When non-cash working capital rises, cash flow falls — because that increase is a real investment, not an accounting reclassification. Receivables, inventories, and short-term operating assets are not administrative details. They are claims on capital.[footnoteRef:3]

The question I care about is therefore never simply, “Is this company profitable?” It is: “Can this system generate cash, redeploy that cash at attractive rates, and survive a bad year without structural breakdown?”

What the Consensus Gets Wrong

The standard narrative of corporate success runs roughly as follows: grow revenue, expand margins, reinvest in the business, and the rest will follow. This is not wrong as far as it goes. The problem is how much it leaves out.

Most financial commentary focuses on average returns — what the company has earned on its existing asset base — rather than incremental returns — what the next unit of deployed capital will actually earn. These two figures can diverge dramatically. A great legacy asset base with pricing power and established market position can sustain attractive average returns even while management is systematically destroying value at the margin through poor reinvestment decisions.[footnoteRef:4] The headline looks healthy. The allocation is failing.

A second gap in conventional analysis is the treatment of growth itself. Growth absorbs capital. If a business must reinvest most of its earnings simply to maintain its existing position — replacing ageing assets, defending market share, funding working-capital expansion — then reported profit is largely illusory. The cash is not available. It has already been spent keeping the engine running. This is common in capital-intensive, inflation-sensitive industries, and it is almost never flagged in investor presentations, which tend to celebrate revenue growth without disclosing reinvestment intensity.

The third and most consequential gap is the failure to think about funding structure. Companies that borrow short to fund long assets, borrow in foreign currency while earning in naira, or rely on continual refinancing to remain solvent are not simply managing a balance sheet. They are making a fragility choice. That choice is invisible in calm markets and catastrophic when conditions turn.

The Allocator’s Lens

The allocator’s lens is a discipline for forcing honest analysis. Instead of asking whether a company is growing, I ask five questions.

Where does cash actually come from? Not from revenue in the abstract, but from a real economic engine: pricing power, cost advantage, repeat demand, fast collections, low reinvestment intensity, or some durable combination of them. If the cash engine cannot be described specifically, the analysis is not yet complete.
What must remain true for that cash to continue? Here I look for hidden dependencies: subsidy, imported inputs, foreign-exchange access, regulatory protection, political goodwill, monopoly position, or infrastructure the company does not own or control. Every dependency is a point of fragility.
What is the reinvestment opportunity set? A good business is not merely one that earned strong returns on yesterday’s capital. It is one that can redeploy fresh capital at similarly attractive rates. The difference between average returns and incremental returns is decisive, and it is only revealed by looking at what the business is actually doing with each new naira of retained earnings.[footnoteRef:5]
How is the system funded? I want funding tenor to match asset life, currency exposure to match revenue currency, and fixed obligations to remain serviceable under a severe stress scenario. Mismatch in any of these dimensions is one of the fastest routes from profitability to insolvency.
What can break first? Every system has a failure path. It may be receivables, fuel supply, regulation, input costs, debt covenants, customer concentration, or a public asset outside management’s control. The allocator’s job is to identify that path before the market does — not to predict when it will fail, but to know whether the upside is worth the downside if it does.

Growth is useful only if it earns attractive returns on the capital it absorbs. Scale is attractive only if it strengthens economics rather than merely enlarging complexity. Market leadership is worthless if the balance sheet has become too brittle to defend it.

This is why I call it a lens rather than a checklist. It forces me to see a business not as a static income statement but as a living cash-flow system with dependencies, delays, buffers, and obligations. The quality of management appears less in its rhetoric than in the architecture of that system. Berkshire’s current shareholder letter expresses this with unusual clarity: allocate to opportunities that are understood, durable, integrity-led, and capable of compounding over time, while actively managing the risks that could impair reputation, financial strength, or long-term opportunity.[footnoteRef:6]

The Fragility Scanner

The allocator’s lens tells me where to look. The fragility scanner tells me how quickly weakness can become crisis. It is a five-minute discipline I apply before I become committed to any narrative.

Obligation load: How much of incoming cash is already committed to interest, leases, payroll, maintenance, or regulated payments? If obligations are high, there is no room for error.
Cash conversion: Are reported profits turning into operating cash, or disappearing into receivables, inventories, and capitalised optimism? Divergence here is one of the earliest warning signs.
Funding mismatch: Is the company borrowing short to fund long, borrowing in dollars to earn in naira, or relying on refinancing to survive? Each mismatch is a structural bet on conditions remaining benign.
Concentration risk: How dependent is the model on one customer, one regulator, one corridor, one supplier, or one critical infrastructure node? Concentration amplifies fragility without raising reported risk.
External control: How much of the company’s economics sits in decisions made by actors outside management’s control? Tariffs, import licences, grid access, and government subsidy are examples. Management may be excellent; the system may still be unfundable.

If fragility is high, valuation becomes secondary. A fragile business can appear cheap for years and still destroy capital. This is why I prefer to think of risk not as volatility — a statistical property of prices — but as the probability that the cash engine seizes up before capital is recovered. Howard Marks frames it similarly: the critical error is not accepting risk, but accepting risk without compensation.[footnoteRef:7] Cheapness without durability is often just deferred disappointment.

What the Cases Show

Dangote Cement: When Internal Funding Creates Compounding

In H1 2025, Dangote Cement’s Nigerian operations grew revenue by 45.5% to ₦1.44 trillion, EBITDA by 82.4% to ₦845.4 billion, and EBITDA margin to 58.6%. At group level the business generated ₦954.3 billion of cash from operations before working-capital movements, ₦874.3 billion after, spent ₦166.3 billion on capital expenditure, and — crucially — still reduced net debt to ₦2.01 trillion.[footnoteRef:8]

Nigeria’s Power Sector: When the Cash Chain Breaks

The opposite logic appears in the power sector. NERC’s Q4 2025 data shows that DisCos billed ₦795.06 billion and collected ₦630.93 billion — a collection efficiency of 79.36%. Billing efficiency was 82.03%. Aggregate ATC&C losses reached 34.90%, against an allowed efficient loss target of 20.54%, translating into ₦139.19 billion of revenue loss in a single quarter.[footnoteRef:9]

The deeper point is structural, not operational. The World Bank estimates that tariff shortfalls owed to power distribution companies had accumulated to ₦2.1 trillion by end-2024, while the annual sector deficit runs at approximately US$1 billion. High losses, low collections, and tariffs systematically set below cost-recovery produce a system in which the DisCos are too weak to finance the investments that would improve service — meaning the structural condition that creates the problem also prevents the solution.[footnoteRef:10]

The failure was not a surprise. The structure made it inevitable. When pricing is political and costs are real, the system can only survive on subsidy — and subsidy is a dependency, not a business model.

Berkshire Hathaway: The Discipline of Refusal

The global comparator I return to most often is Berkshire Hathaway — not because it is the largest or the most complex, but because its capital-allocation discipline is conceptually clear and therefore impossible to fake. Retain earnings only when retention is likely to create more than one dollar of market value for each dollar kept. Repurchase shares only when they are below intrinsic value. Judge every opportunity by its capacity to grow intrinsic value per share over long horizons.[footnoteRef:11][footnoteRef:12]

The brilliance is not in the complexity. It is in the refusal. Refusal to worship size. Refusal to treat cash as if it must always be spent. Refusal to confuse activity with value creation. For anyone studying capital allocation seriously, William Thorndike’s documentation of similar discipline across eight unconventional CEOs confirms that this is not an idiosyncrasy of one firm — it is the common pattern of the best allocators across very different industries.[footnoteRef:13]

Rules I Use

Start with cash, not earnings. Earnings are an opinion shaped by accounting conventions. Cash is a tougher, harder-to-manipulate test of economic reality.
Judge incremental returns, not average returns. A great legacy asset base can disguise poor capital allocation at the margin for years.
Match funding to asset life and currency. Long-duration assets funded by short, unstable, or misaligned liabilities invite fragility that does not show until conditions tighten.
Treat working capital as real capital. Inventory and receivables are not administrative details. They are claims on cash, and they grow with revenue whether or not management notices.
Separate resilience capex from vanity capex. Some investment protects the cash engine. Some merely enlarges the photograph.
Build slack before stress arrives. Liquidity buffers, conservative leverage, and operational redundancy look inefficient in calm markets. They are the only reason survival is possible in stressed ones.
Kill bad projects early. The sunk-cost fallacy is one of the most expensive habits in management. Past investment is irrelevant to future decision quality. What matters is the marginal return from here.

These rules are deliberately simple. They are operational versions of two deeper ideas: first, that owners are ultimately paid in cash rather than accounting theatre; and second, that retained capital is justified only when it compounds intrinsic value rather than merely enlarges the firm. Damodaran formalises the first.[footnoteRef:14] Berkshire states the second explicitly.[footnoteRef:15] Between them lies most of what a serious allocator needs to remember.

When a system cannot convert activity into cash, scale makes it more fragile, not more durable.