"Show me the incentive and I will show you the outcome." Here, Munger's razor is a research procedure. Before you trust a management to run the five-use menu, read what they're paid to do. The documents are public, free, and largely unread.
How do you read a US proxy statement?
United States: the DEF 14A proxy statement (SEC EDGAR, filed annually before the AGM). Go straight to the Compensation Discussion & Analysis (CD&A). It names the metrics behind the annual bonus and the long-term incentive (LTI) plan, their weights, and the targets. Since 2022 the "Pay versus Performance" table, Item 402(v), even forces disclosure of the "Company-Selected Measure", the one metric the board itself says best links pay to performance. Read the metric set as a prediction of allocation behaviour:
| Comp metric | Behavior it purchases | Allocation bias to expect |
|---|---|---|
| EPS / EPS growth | Anything that shrinks the denominator or juices the numerator | Buyback addiction at any price; accretion-driven M&A; underinvestment in long-gestation projects (they hurt EPS now) |
| Revenue / volume growth | Size | Empire-building capex and M&A below WACC (the matrix's bottom rows) |
| EBITDA / adjusted EBITDA | Growth with the capital charge and the D&A of past capex erased | Debt-funded roll-ups (every acquisition is "EBITDA-accretive"); ignore asset intensity |
| Absolute TSR / options-heavy pay | Stock price by any means, on the vesting clock | Leverage, hype, buyback-timed-to-vesting; fine in bull markets, ugly after |
| Relative TSR (vs index/peers) | Outperformance, cycle-adjusted | Better, but still silent on how |
| ROIC / ROCE / economic profit (with a growth gate) | Return on every rupee deployed | Discipline: reluctance to overpay, willingness to shrink, honest buyback math |
| FCF / FCF per share | Cash, not accruals | Good; watch for capex starvation if unbalanced |
The pattern to fear is EPS-heavy comp, plus a stated EPS "roadmap", plus rising buybacks as organic growth stalls. IBM's decade, graded in the scorecard section, had that configuration, and it was visible in the proxy years before the value destruction became undeniable. Respect a return metric paired with a growth metric. Either one alone can be gamed. Pay on ROIC alone and managers harvest and shrink. Pay on growth alone and they build empires.
Advertisement
What else should you look for in a proxy statement?
Collect a few more things from the proxy. Look at actual ownership, skin against salary, meaning the value of shares held ÷ annual pay. Look at option repricing history and at single-trigger change-of-control packages. And look at the peer group used to benchmark pay. A peer set of much larger firms is a pre-registered excuse to overpay for size, and so an incentive to acquire. Two positive live specimens are worth pulling. Home Depot had ROIC in the LTI for years. Texas Instruments publishes an explicit capital-management framework scored on FCF per share growth, and its annual capital-management deck makes it the rare company teaching this material to its own investors.
There's one complication worth knowing. AutoZone compounded per-share value for decades with EPS-flavoured targets because it bought back stock at persistently low multiples with high ROIC. The metric loads the gun, but price paid still decides. Incentives shift your prior, and the allocation record settles it.
Is there an Indian equivalent of a proxy statement?
India: the Board's Report, the corporate-governance report, and the promoter file. India's disclosure stack scatters the same information across several documents. The Board's Report and its annexures carry managerial remuneration and the Section 197(12) ratios: each director's pay ÷ median employee pay, and pay growth against profit growth. It works as a five-second smell test. Remuneration compounding at 20% against flat profits answers question 3 by itself. For the metric structure, read the Nomination & Remuneration Committee policy and the ESOP disclosures (SEBI SBEB). In promoter-run companies it's often thin, because the incentive that counts is the promoter's ~50 to 75% equity. That equity usually aligns on value, but it diverges through specific channels you can check. So your Indian incentive checklist is mostly a related-party and cash-routing audit, the governance thread run with your accounting-quality tools:
- Royalty and brand fees to the parent or promoter. This is a pre-dividend skim that minorities don't share. Track it as a percentage of sales over time, because a rising royalty on flat performance is leakage. SEBI's LODR now pushes material related-party payments, including large royalties, through majority-of-minority approval, so watch how promoters vote-manage these.
- RPT loans, ICDs, guarantees, and purchases or sales with promoter entities (AOC-2, RPT notes). These call on Phase 2's Satyam and IL&FS reflexes. For allocation, the angle is simple. Every rupee routed sideways is a rupee that never reaches the five-use menu.
- Dividend or buyback policy set by group cash needs. You saw this in the dividend and buyback sections. Read the promoter holdco's leverage next to the listco's payout. Pledged-share percentages, from the shareholding-pattern check you already know, tell you when the tail wags.
- Preferential warrants or allotments to promoters priced at regulatory floor prices. In effect it's an option on the company granted cheaply, and it dilutes minorities if it's exercised only when in-the-money. Terms are 25% upfront and an 18-month window (as of mid-2026, check the ICDR mechanics).
- Merger of promoter private entities into the listco at friendly valuations. This is the grand piano of tunnelling. The fairness-opinion and swap-ratio history is the record.
- Managerial remuneration near statutory caps at mediocre ROIC is extraction dressed as pay. Under Section 197 of the Companies Act the cap is ~11% of net profits overall, higher with approvals.
What is the procedure for reading incentives?
None of these is automatically damning. India also produces promoters whose alignment is the asset: skin-in-the-game families who've held themselves to ROIC for decades. What you need is a procedure. Read the incentive documents before the investor deck, predict the allocation bias, then check the prediction against the decade's record. When prediction and record agree, you understand the company. When they disagree, you've found either a governance change or your own error, and both are worth knowing.