Conservative Guidance Is a Habit, Not a Discipline

A listed company in Jakarta beats its guided quarterly EPS by two cents. The stock rises five percent. Three quarters later the same company misses guided EPS by one cent. The stock falls twelve percent.

Everyone in the IR function is appropriately confused about what just happened. The two-cent beat was a conservative-guidance success: we under-promised and over-delivered, the discipline is working. The one-cent miss was a conservative-guidance failure: we guided conservatively, we still missed, the market is unfair. The IR team will explain the second event by pointing at the macro, the sector, the unforeseen one-time charge, the FX move. Nobody on the team will quite say what actually happened, which is that conservative guidance, consistently applied, gradually conditions analysts to model the conservatism into their estimates, until the guided number stops being the operative number and the consensus number is what trades. After several quarters of the conservatism becoming visible, the cushion is no longer a cushion. It is the floor, and the actual report is being judged against an estimate that already removed it.

This is what the IR practitioner literature calls conservative guidance, consistently applied, and it is the most-practised operational principle in modern IR. It is also, when you read across the broader literature, the principle most thoroughly disputed from outside the IR-practitioner tradition. Lawrence Cunningham disputes it from the Berkshire record. Baruch Lev disputes it on empirical-evidence grounds. Aswath Damodaran disputes it on valuation grounds. Meir Statman disputes it on behavioural-finance grounds. The four disputes overlap but do not duplicate each other. Each is independent enough to deserve engagement. Together they make a problem the IR consensus has been carrying around for forty years without anyone quite naming it.

I want to walk through what the consensus teaches, why it is more popular than it is defended, what the four counter-disciplines argue, and what to do if you take any of them seriously. (We previously walked through this same ground more accessibly in our Perspectives piece Failed to Exceed Expectations; this essay is the cross-disciplinary version of that argument.) This is the second essay in the Divergence series, mapping places where the IR practitioner tradition diverges from the broader literature on what listed companies should communicate. The first, The Investor Base You Have Is the One You Chose, covered targeting. This one covers guidance: the operational default that follows from the targeting decision and that, like the targeting decision, most companies have practised one way for a long time without examining whether the practice still holds up.

The principle, as Bragg states it

Steven Bragg’s Running a Public Company names the principle directly. Conservative guidance, consistently applied, is the key to the long-term happiness of an analyst. The framing is twenty years old in print and several decades older in practice. Wild earnings claims produce gyrating prices that cost analysts their coverage. Consistent conservatism produces predictable prints that analysts can model with confidence. Predictable models attract additional analysts who know their estimates will not embarrass them.

The rest of the practitioner canon corroborates it. The association bodies of knowledge teach it, the operational guidebooks build on it, and Palizza’s introductory text presents it as the starting point new IROs should learn first. The principle is taught across decades and jurisdictions with a unanimity the field manages on very few other questions.

The principle has a secondary defence in Bragg himself. He also names the aggressive guidance trap: CEOs who issue better-than-expected guidance in response to short-seller pressure paint the company into a performance corner that eventually guarantees a miss, providing nearly guaranteed profits for the short sellers the aggressive guidance was supposed to repel. Conservative guidance, in Bragg’s framing, is not just easier for analysts. It is the only effective long-term defence against the volatility short sellers feed on. The argument is internally consistent. It is also, when you put it next to the cross-disciplinary critique, embarrassingly incomplete.

Why conservative earnings guidance gets gamed

Here is the part the IR consensus does not engage with carefully. Conservative guidance only works as long as analysts treat the guided number as a true forecast. The moment the conservatism becomes visible, typically after three or four quarters of consistent two-to-three-cent beats, analysts adjust. They begin modelling the conservatism explicitly. The published consensus migrates above the guided number. The guided number becomes the floor of expectations rather than the centre of them.

Operationally, this means the company is now being judged against an estimate that has the conservatism removed. Beating the guided number by two cents and the consensus number by zero produces no stock-price reward, because the beat was already priced. Beating the guided number by two cents and missing the consensus number by one cent produces a stock-price punishment, because the miss is real to the market even though the company did better than its own published guidance. The conservatism has, in this state, become a tax. The company is paying the cost of cautious guidance without receiving the credibility benefit, because the audience has noticed.

A different mechanism produces the same result over a longer horizon. CFOs who hit guidance consistently get rewarded, internally by their boards and externally by analyst coverage and pricing stability. The reward conditions the behaviour. The CFO learns to set guidance with a comfortable cushion, knowing the cushion ensures the beat. The cushion grows over time, partly because the CFO becomes more comfortable with the practice and partly because the consequences of an occasional miss get worse than the reward of a beat. The guided number drifts further below the company’s actual best estimate. Analysts notice. The consensus migrates upward. The cushion stops working.

There is no version of this in which conservative guidance, consistently applied, remains stable. Either the audience notices and prices the conservatism in, or the company progressively widens the cushion and degrades the informational content of the guidance until it stops being information at all and becomes a target the CFO is committed to clearing.

The evidence, fully read

This is where it gets uncomfortable for the IR practitioner tradition, because the empirical evidence the tradition usually cites in defence of guidance does not support what it is taken to support.

Baruch Lev, Robert Houston, and Phillip Tucker studied 222 companies that publicly stopped quarterly guidance against 680 matched companies that continued. The number that travels with the study — a mean five percent stock-price drop on cessation announcement, from Chen, Matsumoto, and Rajgopal’s companion work — gets cited routinely. IR practitioners use it to argue that cessation is costly and guidance is therefore necessary.

What the IR practitioners do not usually cite is Lev’s own framing of why the price drop occurs. “Guidance stoppers clammed up simply because they sat on bad news.” The five percent drop is not evidence that guidance is valuable. It is evidence that cessation, in the population Lev studied, was a signal of approaching trouble that the market correctly read. The companies that ceased guidance were largely companies whose deteriorating fundamentals were about to make the guidance embarrassing. The price drop reflected what was happening underneath, not the loss of the guidance practice as such.

Lev does not stop there. The same body of work documents a 2003 CFO survey finding that eighty percent of CFOs would cut discretionary spending on research and development, advertising, and maintenance specifically to meet an earnings target. Fifty-five percent would delay a new project, accepting a small loss of value to make the print. The myopia, as Lev frames it, is a real cost the guidance regime imposes on operational decisions. The companies that practise conservative guidance most rigorously are the same companies most likely to cut long-term value-creating activities to clear short-term guided targets. The cost does not appear on any income statement. It accrues across years.

This is not a circular argument. It is the same body of evidence the IR practitioner tradition selectively cites in favour of guidance, fully read. The complete read does not support the practice. It provides serious grounds for scepticism.

The Berkshire posture

Lawrence Cunningham’s Quality Shareholders documents the Berkshire alternative directly. Buffett’s posture across the Berkshire shareholder letters has been consistent across fifty-plus years: communicate without forecasting. The shareholder letter discusses the business at length, what happened, why, what management thinks of the result, what management is doing differently. It does not contain quarterly EPS guidance, revenue guidance, or operational metrics with point-estimate predictions. The Berkshire annual meeting works the same way. Buffett and Munger answer detailed questions about the business. They do not provide forward numerical commitments.

The Berkshire shareholder base has reacted as Cunningham would predict. The holders are concentrated, long-duration, narrative-aligned. They are buying the business, not the quarterly print. The absence of guidance does not produce price volatility because the holders are not pricing on quarterly print proximity to guided numbers. They are pricing on the business’s intrinsic value progress, which they assess through the annual letter and meeting rather than through quarterly guidance.

This is not a model that scales to every company. Berkshire is a unique organisation with a unique investor base, and the posture is partly enabled by Buffett’s personal communication skill. The underlying claim Cunningham extracts from the example is not Berkshire-specific. The shareholders a company attracts are determined by how the company communicates. Companies that practise quarterly guidance attract shareholders who want quarterly guidance. Companies that do not practise it attract shareholders who do not need it. The clientele effect makes the guidance posture a self-selecting mechanism for shareholder-base composition. The companies that practise conservative guidance most rigorously end up with shareholder bases that demand guidance, which is the cycle the IR practitioner tradition treats as evidence that guidance is necessary.

Damodaran and the valuation-discipline angle

Aswath Damodaran’s critique of guidance is independent of Buffett’s and Cunningham’s. Narrative and Numbers treats intrinsic value as a function of a story-anchored valuation model with multi-year horizons. Quarterly guidance distorts the conversation between company and analyst toward short-cycle metrics that do not drive intrinsic value. The growth rate over the next decade matters. The EPS print for the next quarter does not. The capital reinvestment trajectory matters. Whether the quarterly result clears the guided number by two cents or misses by one cent does not.

What guidance does in Damodaran’s framework is move the analyst conversation onto terrain where the answers do not affect the valuation model in any meaningful way. The analyst spends time updating estimates for the quarter ahead. The valuation model spends time waiting for information about the long-cycle drivers it actually needs. The company that operates a meticulous quarterly-guidance routine ends up with an analyst community that knows the company’s quarterly print to the cent and does not know the company’s long-cycle reinvestment trajectory. The pricing reflects what the analysts attended to, which is the quarterly mechanics. The intrinsic value, in Damodaran’s framing, ends up underpriced because the conversation has been about the wrong things for years. If your last earnings call produced a dozen questions about the quarter and none about the reinvestment runway, you have watched the mechanism at work.

Statman and the behavioural-finance angle

Meir Statman’s What Investors Really Want adds a fourth angle. Investors care about non-utilitarian motivations, things like regret aversion, fairness, identity, and status, that the rational-investor model in textbook finance does not capture. Conservative guidance is a tool calibrated for the rational-investor model: provide accurate forecasts and the rational investor updates accordingly. The behavioural reality is that investors do not update purely on the print-versus-forecast comparison. They update partly on fairness perceptions (did the company hedge unfairly?), partly on identity considerations (is this the kind of company I want to be associated with?), and partly on regret-aversion mechanics (will I regret holding this through the next miss?).

Conservative guidance, Statman would argue, is uniquely vulnerable to all three. The fairness perception breaks the moment the conservatism becomes visible. The identity association breaks the moment the company starts looking like every other company that beats by two cents in good quarters and misses by one cent in bad ones. The regret-aversion mechanics break the moment the miss arrives, because investors have been conditioned to expect the cushion, and the cushion’s absence feels like a betrayal of an implicit promise the guidance regime has been making. The behavioural cost of guidance is not captured by any rational-investor model and is precisely the kind of cost the IR practitioner tradition is structurally not measuring.

After several quarters of the conservatism becoming visible, the cushion is no longer a cushion.

Three earnings-guidance alternatives that don’t game themselves

The IR practitioner tradition is not entirely uniform on this. Palizza, in his introductory text, lists three alternatives to providing specific point-estimate EPS or revenue guidance.

Each of the three alternatives is operationally workable. Each asks more of the communication function than point-estimate quarterly guidance does. None has the cushion-gaming pathology. None is widely practised in Asian capital markets, where the regulatory floor (OJK in Indonesia, MAS in Singapore, IDX listing rules) treats forward-looking statements as voluntary and most companies still default to point-estimate guidance because that is what their peers do and what their analysts have asked for.

On having practised this, too

The pattern in this essay matches the pattern in the previous one in the series, and the confession needs to be made again. I have advised on guidance policies that were point-estimate-and-conservative by design. The policies produced results my clients liked: predictable beats, stable analyst coverage, manageable conference-call dynamics. The policies were also, by the cross-disciplinary critique, doing exactly the gaming that the critique describes. The cushion was visible to the analysts before it was visible to my clients. The eventual misses were larger than the eventual beats. The shareholder bases that the conservative-guidance regime cultivated were exactly the shareholder bases that responded badly when the misses arrived.

I do not think the IR practitioner tradition is wrong about every operational element of guidance practice. Some of the mechanical advice (about message discipline, about CEO-CFO consistency, about avoiding aggressive forward claims) is correct on its own terms. The framework that organises the mechanical advice has been incomplete.

The error I want to call out specifically is that the IR practitioner tradition has been arguing for conservative guidance as a discipline when the cross-disciplinary literature is fairly clear that it is, in practice, a habit. A discipline is something that produces reliable results when applied consistently. A habit is something that produces uneven results that the people running it stop noticing. The four bodies of work on the other side of the question (Cunningham, Lev, Damodaran, Statman) suggest we have been confusing one for the other for a long time.

The Asian-market angle

Most Asian capital markets give companies meaningful latitude on guidance. OJK in Indonesia treats forward-looking statements as voluntary. MAS-IDX in Singapore does the same. HKEx and the Indian SEBI framework similarly. The companies that provide point-estimate quarterly guidance in these markets are doing so because the IR practitioner tradition tells them to, not because the regulator requires it. The regional latitude creates an opening the US-market context does not have to the same extent.

The opening is operational. Companies with controlling-shareholder structures (family or state) and concentrated long-duration registers, which is most Asian listed companies of any size, already have the kind of shareholder base that Cunningham’s alternative is designed to work for. The shareholders are buying the business, not the quarterly print. Switching from conservative point-estimate guidance to Palizza’s ranges-or-direction alternative, or to the Berkshire-style no-forecasting posture, would fit the existing register better than the practice most regional companies currently run. The misalignment is not regulatory. It is habit-driven, and the habit was imported from a US-market context that does not match the regional structural conditions.

Most companies will keep the habit

Most listed companies will continue to provide conservative point-estimate quarterly guidance. The institutional inertia is significant. The analyst community is structured around it. The board reporting cycle expects it. The competitor set provides it. Changing the practice requires accepting a year or two of awkward analyst conversations and the possibility that some sell-side coverage will drop.

The companies that make the change will produce different shareholder bases and different stock-price dynamics over the next few strategy cycles. Their pricing will reflect their actual business performance more directly. Their internal capital-allocation decisions will not be distorted by the eighty-percent-cut-R&D-to-make-the-number pathology Lev documents. Whether that change is worth making depends partly on the kind of investor base the company is willing to attract and partly on what the company believes its stock price is actually for, which is the question the next essay in the Divergence series, Stock Price Is the Outcome, Not the Objective, takes up directly.

Four disciplines that do not normally talk to each other have spent forty years arriving at the same verdict. The IR practitioner tradition has kept practising as if none of it existed. The companies willing to read all four bodies of evidence and act on what they say will be a small set, in Asian markets and globally. Whether IR Advantage’s clients are in that set is the question this essay is, ultimately, asking.