02 · Financial Metrics in Practice
Metrics are not formulas to memorize but a way to "answer business questions with numbers": what makes this company profitable? Can those profits last? Who holds the power in the supply chain? This article covers the DuPont decomposition of ROE and four other metric groups — growth, operating efficiency, solvency, and cash flow — then breaks down three high-ROE business models and exposes the common pitfalls of combining metrics.
1. Profitability: The DuPont Decomposition of ROE
Remember One Core Sentence
ROE (return on equity) = Net profit ÷ Net assets. It measures how much shareholders' money earns per year and is Warren Buffett's most prized metric. Long-term ROE > 15% is excellent; > 20% and sustainable is rare.
The DuPont Decomposition: Where ROE Comes From
ROE = Net profit margin × Total asset turnover × Equity multiplier
= (Net profit ÷ Revenue) × (Revenue ÷ Total assets) × (Total assets ÷ Net assets)| Component | Meaning | Typical Business |
|---|---|---|
| Net profit margin | How much profit each yuan of revenue keeps | High margin, strong brand, pricing power (premium liquor, high-end consumer) |
| Total asset turnover | How much revenue each yuan of assets drives | Thin margins fast turns (retail, supply chain, distribution) |
| Equity multiplier (leverage) | How many times total assets exceed net assets | Amplifying returns with debt (banks, real estate, insurance) |
💡 Key Point: Identifying the Source of ROE Matters More Than ROE Itself
The same 20% ROE can come from three entirely different businesses. Knowing which component drives ROE matters more than the ROE number itself.
What Business Model Each Component Represents
| Model | Traits | Risk |
|---|---|---|
| High-margin driven ("Moutai type") | Brand/patent/monopoly confer pricing power; highest-quality ROE | Margins erode when demand shrinks or competition intensifies |
| High-turnover driven ("retail type") | Low margins, high turnover; lives on management efficiency | Scale is the lifeline; one expansion misstep spells disaster |
| High-leverage driven ("bank type") | Small capital, big bets; ROE amplified by leverage | Backfires hardest when asset quality deteriorates |
Gross/Net Margin Industry Benchmarks
Reasonable ranges for gross margin and net margin are highly industry-dependent; cross-industry comparison is meaningless:
| Industry | Gross Margin Reference | Net Margin Reference | Notes |
|---|---|---|---|
| Premium baijiu | 80%–92% | 40%–55% | Brand premium; costs are a tiny share |
| Pharma (innovative drugs/devices) | 60%–85% | 15%–30% | R&D-driven; high expense ratios |
| Software/SaaS | 60%–80% | 5%–25% | Low replication cost; selling expenses eat profit |
| Consumer electronics manufacturing | 10%–20% | 3%–8% | Contract manufacturing; relies on scale |
| Retail/supermarkets | 20%–30% | 1%–5% | Classic low-margin, high-turnover |
| Banking | — (net interest margin 1.5%–2.5%) | 25%–40% | Profit model is the interest spread, not gross margin |
| Steel/coal | 5%–20% (cyclical swings) | −10%–15% | Strongly cyclical; top-vs-bottom gaps are huge |
💡 Usage: Compare Across With Peers, Over Time With Yourself
Compare horizontally with comparable companies in the same industry, vertically against your own history. A gross margin persistently well above peers with no moat to explain it is a red flag, not a pleasant surprise.
2. Growth: Look at Growth Rates, But Also Their "Quality"
Revenue Growth vs Profit Growth
| Metric | Formula | Meaning |
|---|---|---|
| Revenue growth | (Current revenue ÷ Prior revenue) − 1 | Is the business expanding or contracting |
| Profit growth | (Current net profit ÷ Prior net profit) − 1 | Change in earning power |
| Profit growth − revenue growth | The spread | Reflects operating leverage and expense control |
What Revenue-Profit Divergence Means
| Scenario | Interpretation |
|---|---|
| Revenue +20%, profit +20% | Volume and price both steady — cleanest kind of growth |
| Revenue +20%, profit +40% | Operating leverage releasing or expense ratio falling; confirm sustainability |
| Revenue +20%, profit +5% | Growing revenue without profit: margins eroded, expenses out of control, or consolidation dilution |
| Revenue +20%, profit −10% | Growth bought by burning money; beware the expansion trap |
| Revenue 0%, profit +30% | Profit from cost cuts or non-recurring items — not real growth |
⚠️ Be Extra Wary of "Stagnant Revenue, Surging Profit"
If profit comes from cutting expenses, selling assets, or subsidies rather than a stronger core business, this "optimization" eventually runs out.
3. Operating Efficiency: A Ruler for Supply-Chain Relationships
Three Turnover Ratios
| Metric | Formula | Meaning |
|---|---|---|
| Inventory turnover | Cost of sales ÷ Average inventory | How often inventory sells through; higher is faster |
| Receivables turnover | Revenue ÷ Average accounts receivable | Collection speed; higher means more bargaining power |
| Payables turnover | Cost of sales ÷ Average accounts payable | How long before suppliers get paid; lower means more use of upstream funds |
Cash Conversion Cycle (Working-Capital View)
Cash conversion cycle = Days inventory outstanding + Days sales outstanding − Days payables outstandingThe "squeeze both ends" business: days payable outstanding is very long (delaying payments upstream), receivables are short (downstream pays first), inventory turns fast — the cash conversion cycle goes negative; the company needs no working capital and instead runs on other people's money.
| Typical Case | Traits |
|---|---|
| Appliance chains / supermarkets | Long payment terms upstream, cash sales downstream; negative cycle |
| Construction / labor subcontracting | Fronting money and settling by progress; extremely long cycle |
| Baijiu (distributor model) | Cash before delivery, minimal receivables; naturally high-quality cash flow |
The trend in turnover matters more than the level: a declining receivables turnover = the company is losing bargaining power downstream — an early sign of deteriorating business economics.
💀 Iron Law: Falling Receivables Turnover Is an Early Warning
A declining receivables turnover = the company is losing bargaining power downstream — an early sign of deteriorating business economics. Trends matter more than levels; if gross margin keeps rising while receivables turnover keeps falling, profit is being turned into IOUs and the cash flow is one step from blowing up.
4. Solvency: The Bottom Line of Survival
| Metric | Formula | Meaning |
|---|---|---|
| Interest-bearing debt ratio | (Short-term borrowings + long-term borrowings + bonds payable + lease liabilities) ÷ Total assets | Real debt burden after excluding interest-free payables |
| Interest coverage | (Net profit + interest expense + income tax) ÷ Interest expense | How many times earnings cover interest; below 2x is dangerous |
| Operating cash flow / interest-bearing debt | Operating cash flow ÷ Interest-bearing debt | How much annual internal generation covers the debt |
| Cash-to-short-term-debt ratio | Cash ÷ Interest-bearing debt due within one year | Whether cash on hand can absorb short-term debt |
Common misconception: a low debt-to-assets ratio ≠ safety. Interest-free liabilities (accounts payable) use other people's money; only interest-bearing debt is a true financial burden — so look at the interest-bearing debt ratio first. Real estate and banking are naturally high-leverage; read metrics within their industry context.
⚠️ Counterintuitive: A Low Debt Ratio Is Not Safety
Interest-free liabilities (accounts payable) use other people's money; only interest-bearing debt is a true financial burden. So prioritize the interest-bearing debt ratio over total leverage — a company that looks low-leverage may simply hide its debt in payables and customer advances, and its true burden is anything but light.
5. Cash Flow Metrics
| Metric | Formula | Meaning |
|---|---|---|
| Free cash flow FCF | Operating cash flow − Capex | Money genuinely free after sustaining operations — the ultimate source of enterprise value |
| Cash content | Operating cash flow ÷ Net profit | Quality of profit; consistently > 0.8 is good |
| FCF / net profit | Free cash flow ÷ Net profit | Above 70% long term is the common line for a good business |
| Cash-receipts ratio | Cash received from sales ÷ Revenue | What fraction of revenue is real money |
See cash-flow-analysis.md: full discussion of the three cash-flow patterns, FCF, and DCF.
6. Case Profiles: "Three Types of High-ROE Companies"
Fictional companies illustrate three high-ROE business models:
| Type | Fictional Example | DuPont Breakdown | Traits and Risks |
|---|---|---|---|
| Brand-driven | "Shanhe Niang" baijiu | Net margin 45% × turnover 0.5 × leverage 1.4 ≈ ROE 31% | High gross/net margins, almost no borrowing; risk lies in demand cycles and channel inventory |
| Asset-light | "Yunqiao SaaS" software | Net margin 20% × turnover 1.0 × leverage 1.2 ≈ ROE 24% | 75% gross margin, low replication cost, great cash flow; risk lies in runaway customer-acquisition spending |
| High-leverage | "Jiangpan Bank" | Net margin 28% × turnover 0.04 × leverage 12 ≈ ROE 13% (typical bank level) | Earnings scale built entirely on leverage; risk lies in bad loans detonating at once |
How to judge: compute all three DuPont factors from the statements, see which factor drives ROE, then assess that model's sustainability and fragility. At equal ROE, brand-driven quality > asset-light > high-leverage.
💀 Iron Law: At Equal ROE, Quality Varies Enormously
At equal ROE, brand-driven quality > asset-light > high-leverage. Two companies both at 20% ROE — one via net margin, one via 8x leverage; once the industry cycle turns, the leveraged firm's profits collapse instantly — reading the ROE number without the DuPont breakdown is a classic cause of value traps.
7. Pitfalls in Combining Metrics
Pitfall 1: Judging by ROE Alone
- High ROE may be propped up by leverage: two firms both at 20% ROE — one with 30% net margin, one with 8x leverage — carry completely different risks.
- High ROE may be the result of "slimming down": massive dividends, buybacks, or impairments shrink net assets; a smaller denominator mechanically lifts ROE.
- High ROE may just be a boom year: cyclical stocks post peak ROE at cycle tops — exactly when you should least buy.
- The right way: read DuPont breakdown + a 5-year ROE series + cash content together.
Pitfall 2: Effects of Accounting Policy Changes
Change an accounting policy and the numbers can transform while the fundamentals haven't changed at all:
| Change | Common Direction | Effect |
|---|---|---|
| Depreciation/amortization life adjustment | Lengthening useful lives | Current costs fall, profit inflated |
| Inventory valuation method change | FIFO ↔ weighted average | Profits shift markedly in inflationary times |
| Timing of revenue recognition | Recognizing earlier | Current revenue balloons, later periods pay back |
| Bad-debt provision ratio | Lowering the ratio | Receivable risk gets masked |
| Goodwill impairment-test parameters | Loosening assumptions | Impairments surface late |
💡 Response: Compare Like With Like
Financial statement notes disclose the impact of accounting policy changes; always build year-over-year comparisons on the same basis, and don't be fooled by growth rates manufactured by basis switching.
Pitfall 3: Other Common Misuses
- Comparing gross margins or P/E ratios directly across different industries — meaningless.
- Looking only at year-over-year and never quarter-over-quarter — seasonal industries (baijiu Q4, retail Q4, travel summer) can hide turning points in YoY data.
- Concluding from a single year — cyclical stocks need 5–10 years of history to reveal their patterns.
- Great-looking metrics everywhere except poor cash flow — profit can wear makeup; cash cannot.
⚠️ Risk Warning
⚠️ Risk Warning
All financial metrics are lagging data — they reflect the past, not the future; metrics can be manipulated by accounting techniques, and combining them only improves your odds, it does not remove risk; a high ROE is not a good stock — an excellent company at too high a valuation still loses money. All example companies here are fictional, numbers for teaching only. No combination of metrics can substitute for the fundamental homework of understanding how this company makes money and whether those earnings are real.