Recency and narrative

Here is a scene you will meet again and again. A company reports one excellent quarter, and within a sentence the reader has decided: this is a turnaround. A different company reports one weak quarter, and just as fast: this business is decaying. In both cases a single, fresh data point has been promoted, almost instantly, into a settled story about the future.

Notice how little work that promotion took. No one asked what usually happens after one strong quarter in this kind of business, or what else — a festive season, a round of discounts, a one-off order — could have produced the same number. The story simply appeared, fully formed and satisfying, the moment the data landed. That speed is the subject of this module.

Two instincts are working together here, and they are close cousins. One over-weights the latest, brightest information; the other wraps that information into a clean tale. Apart, each is manageable. Together, they can turn a fortnight of price action or a single quarter's result into something that feels like a permanent truth with a compelling reason attached. This module is about slowing that hand-off — the one between a fresh number and a firm conclusion — before it reaches for the buy button.

The market pours fresh information over you constantly — prices tick every second, results arrive every quarter, headlines every hour. And the mind has a standing rule about fresh information: it feels more alive, more real, more decisive than the older pattern sitting quietly behind it. This is recency bias — the habit of giving the newest and most vivid data far more weight than the long, dull record it belongs to. The last three months feel like the truth; the previous nine years feel like history.

On the plains, this was sound. The rustle you heard just now mattered more than the calm of the last hour, because the danger was current and the old calm could not eat you. Freshness was a fair proxy for relevance. In a market, freshness and relevance come apart: the newest number is often the noisiest, and the boring long record is frequently the more reliable guide to what comes next.

Recency supplies the raw material; the second instinct does the assembly. The mind cannot bear a heap of disconnected facts — it wants a cause, a shape, a point. So it reaches for the narrative fallacy: the pull to stitch messy, partial facts into a clean, satisfying story before the facts have earned it. One good quarter becomes "premiumisation"; a rising sector becomes "India's decade"; a run of green becomes "a structural rerating." The story feels like understanding. Very often it is only the feeling of understanding — which is a different, and far more expensive, thing.

Put the two together and the danger sharpens. Recency hands narrative a bright, recent slice of reality; narrative gives that slice a plot and a reason; and the reader ends up holding a story that feels both current and explained — the two qualities most likely to make a person act. The defence, which the rest of this module builds toward, is almost embarrassingly plain: before accepting the story, ask what usually happens in this class of situation. That is base rates before stories — the outside view arriving before the vivid one closes the case.

Trace the machinery step by step and it loses much of its grip. It runs in four quiet moves.

A recent point gets emotional weight. The latest quarter is easy to picture and easy to remember; the ten-year record takes effort to retrieve. The mind spends attention like money, and the vivid, recent thing is cheap to hold. So it looms larger than its true importance — not because you decided it should, but because it was simply closer to hand.

The point gets extended into a line. Having over-weighted the recent stretch, the mind draws it forward. Two great quarters become "and it will keep compounding"; a hot sector's last two years become "the runway is a decade." This is extrapolation — assuming the recent slope simply continues. It is the most natural move in the world and, in markets, one of the most dangerous, because trends in prices and margins are unusually prone to bending back.

The line gets a story. A bare projection is uncomfortable, so narrative dresses it: a theme, a name, a reason. "EV supercycle." "China plus one." "Structural decline." The cleaner and more repeatable the phrase, the more carefully you should inspect what it leaves out — a good slogan is engineered to travel, not to be true.

The base rate goes missing. At no point did anyone ask the outside-view question: across many companies that posted one strong quarter, or many sectors that ran hot for two years, how often did the story hold? That number — the base rate, the starting odds for the whole group — is exactly what the vivid story crowds out. And the base rate for hot trends is unkind: most revert. Mean reversion — the tendency of stretched prices, margins and growth rates to drift back toward their long-run average — is the quiet force the extrapolated line ignores.

Set the two worlds side by side and the shelf's one inversion appears again: the instinct that read the world well on the plains misreads it at the screen.

Watch the machinery run on an ordinary case, then take it apart yourself. illustrative

A retailer reports a strong festive quarter — sales up sharply — after six flat quarters in a row. The story writes itself in the reader's head almost before the numbers finish loading: the turnaround has begun. The chart obliges: over the last three months the line only points up.

Now look at what the story quietly skipped. The store additions were modest, so the growth did not come from more shops. The discounting was heavy, so a chunk of the sales was bought, not won. And six flat quarters sit directly behind the one bright one — the base rate the narrative stepped over. Recency lifted the festive quarter above its weight; extrapolation drew it forward into "a turnaround"; narrative gave it a name. Three moves, one confident conclusion, and not one durability question answered.

The sound read is neither "it's saved" nor "it's nothing." It is quieter: the quarter is evidence, not a biography. A single strong period earns study, not a story — and the test is repetition. Does the growth hold across the next two or three quarters without leaning on discounts? Does margin survive when the sale ends? Does the cash actually arrive? Those answers, spread across time, are what turn one bright quarter into a real turn — or expose it as a festive blip dressed up as a destiny.

Now play with the instinct directly. The toy below shows a single illustrative stock — one price history — across three windows. Start on the last three months and read the story it tells. Then widen the window twice and watch the same surge change meaning: first into a fall being un-crashed, then into a flat round trip. Try the "extrapolate" button on the short view to see the straight line the recent slope invites, and how far it drifts from a record that keeps reverting.

The lesson the toy makes on its own is the module's whole method: read across three timeframes, not one. The latest quarter, the trailing few years, and the through-cycle record often tell three different stories about the same company — and only the reader who asks for all three can tell a trend from a bounce.

Reading across timeframes and reaching for the base rate protects you from a great deal. It also cannot do everything, and pretending otherwise turns a good habit into a new blind spot.

It does not mean the recent data is wrong. Sometimes the latest quarter genuinely is the turn, and the long flat record genuinely is about to break. A base rate is the starting odds for the group, not a verdict on the single case in front of you — and a real change in the facts is allowed to move it. The skill is to name which new fact is strong enough to override the base rate, not to worship the base rate as if nothing ever changes.

It does not license dismissing every story as "just narrative." A bias label swung at genuinely important, genuinely new evidence is no longer a defence — it is confirmation bias wearing the costume of self-awareness. "That's just recency" can hide from a real signal as easily as recency can inflate a fake one.

And waiting for total certainty has its own cost. By the time a turn is proven across many quarters, the price has usually moved to reflect it. The point of the three-timeframe read is never permanent delay; it is proportionate confidence — believing a one-quarter story about as much as one quarter deserves, and no more.

The same handful of moves catch beginner after beginner. Named once, each is far easier to spot in yourself in the heated moment.

1. Naming a trend after one data point. One quarter is "a turnaround"; one bad month is "the story is over." A single point has no direction — direction needs at least a line, and a line needs time.

2. Buying the theme at the price the theme is already in. By the time "EV supercycle" or "India's decade" is on every screen, the optimism is priced. A true story and a fully-priced story are often the same story.

3. Extrapolating two great years forever. The recent slope feels like a law of nature. In markets, stretched growth rates and margins are unusually prone to bending back toward the average — the reversion the straight line pretends away.

4. Letting the latest window stand in for the whole record. The last three months feel like the truth because they are vivid and near. Widen to three and ten years and the same surge can turn into a bounce, or a round trip.

5. Waving "base rate" at real news. The outside view is a tool for humility, not a reflex for dismissal. When a genuinely new fact appears, the job is to weigh it against the base rate — not to use the base rate to avoid looking.