Reinvest with an allocation ladder

Profit creates options. Capital allocation decides which options survive, which risks remain, and whether the next rupee strengthens the machine or merely enlarges it.

Fictional Kaveri Systems finishes a later year with ₹18 lakh of cash it describes as “available.” That word needs an audit. Some cash may belong to upcoming tax, payroll, customer delivery, debt service, warranty work, or committed purchases. Some may be the minimum buffer that lets the company survive one delayed customer. Only after those claims are resolved does management face an allocation decision.

The first skill in capital allocation is refusing to spend the same rupee twice.

Start from cash, then rebuild the obligations

Free cash flow is commonly presented as:

Free cash flow = operating cash flow − capital expenditure

It is a useful screen, not a universal definition of owner-distributable cash. Operating cash can be lifted by a temporary payable stretch or customer advance. Reported capital expenditure may mix maintenance and growth, while some necessary product or software investment may run through operating expenses. Debt principal, acquisitions, lease payments, restricted cash, statutory obligations, and the next working-capital build can sit outside the simple formula.

For a small operating company, begin with the reconciled cash-flow statement and thirteen-week forecast. Subtract purpose-bound cash, due obligations, credible downside needs, maintenance requirements, and committed growth. The balance is the amount available for a deliberate choice, subject to solvency, contracts, law, and tax review.

The allocation ladder

Use this order as a gate, not as fixed percentages:

1. Keep faith. Fund payroll, tax, promised customer delivery, due vendors, refunds, and contractual obligations. 2. Maintain the earning engine. Replace or repair what must work; fund security, controls, insurance, backups, and compliance proportionate to the risk. 3. Protect survival. Hold enough liquidity for volatility, concentration, and the time needed to act. A buffer’s return appears as choices retained during stress.

4. Remove the proven constraint. Invest where evidence shows one bottleneck limits customer value, capacity, quality, or collection. 5. Scale repeatable economics. Fund channels, inventory, people, or systems only after contribution and cash conversion remain credible at the next scale. 6. Reduce expensive or dangerous claims. Repay debt when its after-tax cost, covenant burden, maturity risk, guarantee, or lost flexibility exceeds the best risk-adjusted alternative. 7. Build capability and options. Develop people, process, data, intellectual property, and small experiments that lower dependency or open a bounded opportunity. 8. Return and diversify capital. Pay lawful owner compensation or distributions without starving the company; reduce the household’s concentration as wealth grows. 9. Increase lifestyle slowly. Make recurring personal costs the last claim, because they are hardest to reverse when business cash weakens.

Several gates can be funded together. The sequence forces each proposal to answer to obligations and ruin before upside.

Measure the return on the next rupee

Historical return on invested capital can describe the existing engine. The allocation question concerns incremental return: what the next unit of capital is expected to produce.

One practical formulation is:

Incremental ROIC = incremental NOPAT ÷ incremental invested capital

NOPAT = operating profit × (1 − normalised tax-rate assumption)

Invested-capital definitions vary. A common operating view uses interest-bearing debt plus equity minus excess non-operating cash; another uses operating assets minus non-interest-bearing operating liabilities. Use average capital where balances move materially, document adjustments, and compare like with like. For a project, include equipment, implementation cost that truly creates the asset, and added working capital. Do not improve the answer by omitting the cash trapped in receivables or inventory.

ROIC must be compared with the capital’s risk-adjusted required return. A private company rarely knows its cost of capital with precision. Use a hurdle range, test sensitivity, and acknowledge illiquidity, concentration, execution risk, and the value of waiting. A forecast percentage does not override a severe downside or irreversible commitment.

A worked allocation memo

Kaveri considers ₹12 lakh of installation equipment plus ₹4 lakh of additional working capital. Management forecasts ₹4.8 lakh of annual incremental operating profit after recurring project costs. With an illustrative 25% normalised tax-rate assumption, expected NOPAT is ₹3.6 lakh and forecast incremental ROIC is 22.5%: ₹3.6 lakh divided by ₹16 lakh.

The downside case produces only ₹1.2 lakh of incremental operating profit. Illustrative NOPAT becomes ₹0.9 lakh and ROIC falls to about 5.6%. The spread between 22.5% and 5.6% is the decision. It says the sales, utilisation, price, and collection assumptions deserve evidence before full commitment.

Kaveri can stage the bet. A ₹3 lakh pilot may test installation time, defects, customer acceptance, technician utilisation, and collection before the remaining capital is released. The pilot does not need the same accounting return as the mature project; its job is to buy decision-grade evidence at a capped loss.

Every proposal should fit one row in a capital-allocation memo:

Compare unlike uses honestly

Some choices have a measurable cash yield. Repaying a loan can offer a relatively knowable saving after fees and tax effects. A capacity investment may have a wide return range. A security control may prevent a low- probability loss rather than create revenue. A cash buffer earns optionality. A distribution reduces founder concentration but removes capital from the business.

Do not force every choice into one false-precision percentage. Put money uses into three buckets:

• • must do: legal, contractual, integrity, and maintenance obligations;
• • return seeking: repeatable growth, pricing, productivity, or acquisition projects;
• • option and protection: buffers, resilience, experiments, and concentration reduction.

Then compare each within its purpose and against the cost of doing nothing.

Allocation failure modes

Capital is commonly destroyed through growth before unit economics work, offices and systems bought for status, permanent hires made for temporary demand, acquisitions without integration capacity, debt-funded distributions, and projects approved from average returns while the incremental return is falling. A profitable company can still destroy value when each additional rupee earns less than its full risk-adjusted cost.

Review past decisions, not only new proposals. At thirty, ninety, and 180 days, compare the memo with actual cash, working capital, margin, utilisation, and customer evidence. Stop weak projects before identity attaches to them. Increase the next tranche only when evidence and loss-bearing capacity both improve.

The founder has two balance sheets: company and household. Reinvestment may be rational while the company has a rare, proven opportunity. Diversification becomes more valuable as the household’s income, equity, guarantees, and reputation all depend on the same enterprise. Decide across both balance sheets without moving restricted or company money casually between them.

· EF pp. 75–82 · FS pp. 20–27, 52–55 · RM pp. 61–65 · TABLE pp. 26–29, 48–50 · GE ch. 27. Mechanisms paraphrased; Kaveri Systems, tax assumptions, projects, and returns are explicitly fictional worked examples, not forecasts or prescriptions.