What ratios cannot tell you

Picture a company with a perfect set of ratios. Return on capital high and rising, margins fat, cash conversion clean, balance sheet lightly leveraged, valuation fair for the growth. Every number this part has taught you to compute comes out beautifully. And yet that same company could be a fraud whose accounts are invented, a former leader whose customers are quietly defecting to a cheaper rival, a maker of a product a new technology is about to render obsolete, or a two-decade success story resting entirely on a founder who retires next year. The ratios look identical in every one of those cases - and identical to a genuinely excellent business. illustrative

This is the module that draws a boundary around everything the part has built. Ratios are powerful, and we have spent nine modules learning to read them honestly - as questions, against their context, cleaned of cosmetics. But every ratio shares one built-in limit: it is computed from the reported past. So the things that decide a company's future, and that live outside the reported numbers, are invisible to all of them. Four matter most - integrity (are the numbers even real?), competitive position (is the moat widening or eroding?), technology risk (is the whole business about to be disrupted?), and succession (does the quality survive the people who built it?). This closing module is about those four blind spots, and it is the bridge out of the ratios into the parts that examine exactly what the ratios cannot.

The most dangerous mistake a numerate reader can make is to believe a complete ratio analysis is a complete analysis. It is half of one, and the missing half is often where the money is won or lost. What makes this more than a caveat is that all four blind spots fail the ratios for the same underlying reason: a ratio is a backward-looking quantification of numbers a company has already reported. Whatever is not in the reported figures, or not yet in them, no ratio can reach.

Integrity is the deepest, because it undermines the rest. Every number in this part - ROCE, margin, cash conversion, leverage - is calculated from the reported accounts, so if the accounts are false, the ratios are precisely, elegantly false too. A well-built fraud produces a flawless ratio profile, because that is frequently its whole point. No amount of ratio work detects this, since the ratios are downstream of the very numbers in doubt. A ratio can never tell you whether to trust the number it is built from.

Competitive position is invisible because margins lag the moat. A ratio shows that a firm is profitable; it is slow to show that a firm is becoming less defensible. When a cheaper or better rival takes share, the incumbent can protect its margin for years simply by not cutting price - so the margin stays healthy while the moat erodes beneath it, until the day the firm is forced to respond and the margin falls hard and late. The leading signals - share slipping, customers defecting, pricing power weakening - sit outside the accounts, in the market.

Technology risk is invisible because the threat lives in the future. A business about to be made obsolete shows perfect numbers until the obsolescence lands, because a threat that has not yet reduced sales or margins is simply not in the reported figures. Ratios record what has happened; a disruption is, by definition, what has not happened yet. You assess it by looking forward - at the substitute, at what rivals are building - which no backward-looking number can do.

Succession is invisible because ratios record results, not their source. A twenty-year run of high returns built on one founder's judgement looks identical, in ratio form, to one built on a deep, institutionalised organisation - yet one survives the founder's departure and the other may not. The ratios cannot tell you how dependent the record is on individuals, how deep the bench is, or whether the culture outlasts the people who created it.

The limitation is universal, but each blind spot bites hardest where the incentive or the exposure concentrates - and that tells you which non-ratio question to weigh most heavily for a given business.

A composite firm arrives with the most attractive ratio profile in its sector: ROCE in the thirties, twenty years of compounding earnings, clean cash conversion, negligible debt, a valuation fair for the growth. On the numbers alone it is close to a perfect business. Run the four questions the ratios cannot answer, and watch how much of the case they never touched. illustrative

Integrity - are the numbers real? The ratios cannot say, so you check elsewhere: trace twenty years of profit to twenty years of cash, read the auditor's history and any related-party dealings, watch for the tells of the cosmetics module. Suppose this passes cleanly - the forensics earned that reassurance, not the ratios. Competitive position - is the moat holding? Here the ratios actively mislead if trusted alone: the fat margin is a lagging signal, and the market tells you a well-funded rival has been taking share in the core segment for three years, the margin intact only because the incumbent has not yet had to respond on price. That is a leading warning the ROCE cannot show. Technology risk - could the product be displaced? The core offering leans on a standard a newer approach is beginning to erode; the current numbers are pristine, but the roadmap question is live and sits entirely outside the accounts. Succession - who built this? The twenty-year record was built by a founder-CEO now past sixty, with no obvious successor and a thin bench - so the very consistency the ratios celebrate may be the fingerprint of one irreplaceable person.

Put together, the "close to perfect business" on the ratios is, on the four questions the ratios cannot answer, a company with honest accounts (good), an eroding competitive position (concerning), a live technology threat (concerning), and serious founder dependence (concerning). The ratios were necessary - they earned the company a serious look - but they were nowhere near sufficient, and three of the four decisive risks were invisible to every number in this part.

The competitive blind spot is worth seeing as a picture, because it is the one that most often fools a numerate reader: the healthy margin looks like proof of strength right up to the cliff.

This module is about what ratios cannot tell you, so its own honest caveat is the reverse one: the four blind spots are not a checklist that, once ticked, makes you safe. Integrity, competitive position, technology and succession are judgements, not calculations, and each can be got wrong even by someone who knows to ask it. Confirming that profit traces to cash reduces integrity risk but does not eliminate it; judging a moat is a fallible read of a moving picture; assessing a technology threat is genuinely hard and often wrong. Knowing where to look is not the same as seeing clearly once you look.

Nor are these four the only things outside the ratios - regulation, the cycle and plain luck can each overwhelm a single company's numbers. And there is a final, load-bearing limit that runs through the whole part: even a perfect analysis - flawless ratios and sound judgements on all four blind spots - cannot tell you what price the market will pay, or when, or why it so often moves opposite to the fundamentals on results day. A company can be excellent on every dimension and still be a poor investment because it was bought at the wrong price. That gap, between understanding a business and understanding its price, is the one no amount of company analysis closes, and it is what the later part on why the price moves the other way is built to address.

How it comes up. When the numbers are flawless, a thoughtful analyst stops asking about them and starts asking the four things they cannot show: "The financials are excellent, so I want to ask what they don't capture - how is market share trending in your core segment, what's your read on the newer technology some entrants are pushing, and what does succession look like beyond the founder?"

A good answer engages exactly what the ratios miss:

"Right questions. Share in the core segment slipped about two points last year to a newer competitor - we're responding with a lower-priced line, which will cost some margin. On technology, we take the substitute seriously and have a team on it; I won't pretend it's not a risk. On succession, we've brought in a deputy CEO who ran the largest division for five years, and we're deliberately institutionalising decisions that used to sit with the founder. None of that is in the numbers, but it's what we spend our time on."

It concedes share loss, takes the threat seriously, and shows a real plan - rather than pointing back at the financials.

An evasive answer does the opposite:

"Our track record speaks for itself - twenty years of industry-leading returns and rock-solid financials. The same disciplined approach that delivered those results will continue to serve us. We don't see any of these as material concerns given our proven model."

Notice the move: asked about the four things the ratios cannot show, it answers by pointing back at the ratios - "track record," "returns," "financials" - and waves away forward risks by invoking the past. That is precisely the substitution this module warns against. The follow-up nobody asks - "Set the financials aside: walk us through the share trend, the roadmap versus the substitute, and the specific succession bench" - forces the conversation off the numbers and onto the blind spots. A management that answers every question about the future by reciting its past ratios is relying on exactly the limitation this part closes on.

1. Believing a ratio can validate its own inputs. Ratios are computed from the reported numbers, so they cannot tell you whether those numbers are honest. A fraud produces beautiful ratios; integrity is checked by forensics, not by the ratios.

2. Reading a healthy margin as a healthy moat. Margins lag competitive position - a firm losing share can hold its margin for years, then lose it suddenly. The leading signal is market share and pricing power, which sit outside the accounts.

3. Mistaking a great record for a durable institution. Ratios record results, not who produced them, so a founder-dependent record looks identical to an institutionalised one. Both succession risk and the threat of a future technology are invisible to every ratio, because one is about people and the other is about the future.

4. Treating a complete ratio analysis as a complete analysis. The ratios are necessary and never sufficient; several of the decisive risks routinely sit entirely outside them - including the last one, price. A strong ratio set earns a company the next questions; it does not answer them.