Anchoring, Framing and Mental Accounting

Three of the sneakiest biases in this whole course don’t feel like biases at all. They feel like reasonable. A number you saw once quietly sets your sense of “fair.” The very same fact makes you cautious or reckless depending on how someone phrased it. And a dollar of winnings somehow burns a hole in your pocket that a dollar of salary never would — even though every dollar is identical. None of this is stupidity. It’s the predictable wiring your mind uses to make fast decisions, and the market is happy to bill you for it. Let’s expose all three.

Anchoring — an irrelevant number drags your estimate

Imagine being asked to guess how tall the tallest tree on Earth is. If I first ask “is it taller than 1,200 feet?”, your guess lands high. If I first ask “is it taller than 180 feet?”, your guess lands low — same tree, same you, two different answers. That first number is an anchor: a starting value your mind grabs and then adjusts away from — but never adjusts enough. You drop anchor near the number you saw and drift only a short distance from it.

Anchoring is the tendency to rely too heavily on the first piece of numerical information you encounter when making an estimate, even when that number is irrelevant or arbitrary. The mechanism is “anchor and adjust”: System 1 (the fast, automatic mind you met in Two Minds) latches onto a reference number, and System 2 adjusts from it — but adjustment is effortful and lazy, so it stops too soon. The anchor wins.

The classic evidence is Tversky and Kahneman’s 1974 wheel-of-fortune study. They spun a wheel rigged to stop on either 10 or 65, asked people whether the percentage of African countries in the United Nations was higher or lower than that number, then asked for an actual estimate. People who saw 10 guessed about 25%; people who saw 65 guessed about 45%. The wheel was visibly random and carried no information whatsoever — yet it moved the answer by 20 percentage points. That’s how strong this is.

Where anchors hide in your portfolio

In investing, the anchors aren’t random wheels — they’re specific, seductive numbers that feel meaningful but mostly aren’t:

• Your purchase price. What you paid is the single biggest anchor in investing. The market does not know or care what you paid; the stock’s future depends on the business from here, not on your cost. Yet “I won’t sell below what I paid” is a sentence almost every investor has thought.
• The 52-week high. A stock that hit $200 and now trades at $120 feels cheap — “it was $200!” — even if $120 is richly valued. The high is an anchor, not a target.
• Broker price targets. “Analyst sets $90 target” plants a $90 anchor in thousands of heads. Targets are frequently revised and often wrong, but the round, confident number sticks.
• Round numbers. Indices “struggling at 5,000,” a stock that “needs to break $100” — the human love of round numbers turns arbitrary thresholds into psychological walls.

Worked example — “I won’t sell below what I paid”

You bought a stock at $80. It’s now $55. The business has deteriorated — guidance cut, a key product flopped — and on today’s facts you’d value it at $50. You catch yourself thinking: “I’ll hold until it gets back to $80, then sell.”

That sentence is pure anchoring. Notice what it does and doesn’t depend on:

The market has no memory of your $80. The stock doesn’t owe you a recovery. The right question — which you’ll meet again in Sunk Costs — is “knowing only what I know now, would I buy this at $55?” If the answer is no, the $80 anchor is the only thing keeping you in. The anchor isn’t information; it’s a leash.

When anchoring matters most

Anchoring bites hardest exactly when you have no firm independent value of your own — a new stock, an IPO, an unfamiliar asset — because an empty mind grabs the nearest number. It’s also weaponised against you: a seller’s first “asking price,” a fund’s headline “since-inception” figure, a negotiation’s opening offer. The defence is to form your own estimate of value before you look at the suggested number, and to ask “would this number move me if I’d never seen the anchor?” If not, drop it.

Framing — the same fact, dressed two ways

Ground beef that’s “90% lean” sounds like health food. The identical beef labelled “10% fat” sounds like a heart attack. Nothing about the meat changed — only the costume the fact is wearing. That’s framing: the way information is presented — as a gain or a loss, a survival rate or a death rate, a glass half full or half empty — changes the decision people make, even when the underlying facts are mathematically identical.

The landmark demonstration is Tversky and Kahneman’s 1981 “Asian disease problem.” People were told a disease threatened 600 lives and asked to choose between two programmes. One group saw a gain frame: Program A “saves 200 people,” Program B “1/3 chance to save all 600, 2/3 chance to save no one.” Most chose the sure save (A) — risk-averse. A second group saw the mathematically identical options in a loss frame: Program A means “400 people die,” Program B is “1/3 chance nobody dies, 2/3 chance 600 die.” Now most chose the gamble (B) — risk-seeking. Same outcomes, opposite choices, just because one was framed as lives saved and the other as lives lost.

This is not a random quirk — it’s Prospect Theory (Kahneman & Tversky, 1979) in action, the theory you met two lessons ago. Its core says we judge outcomes as gains or losses against a reference point, and we’re risk-averse over gains but risk-seeking over losses. Framing simply chooses which side of the reference point a fact lands on — and that flips your appetite for risk.

The same portfolio, framed two ways

Watch how reframing a single fact about a fund changes how it feels:

Every row can describe the exact same fund in the same year. The fund’s actual prospects are unchanged across all four. Only the frame moved — and notice the loss frames quietly smuggle in an anchor (the 2021 peak, the index), tying these two biases together.

Worked example — a fee, framed two ways

A fund charges 1% a year. Framed as “just 1%,” it sounds trivial — who quibbles over 1%? Framed honestly over a 30-year horizon, that 1% annual fee can eat roughly a quarter of your final wealth through lost compounding. Same fee, two frames: the small-percentage frame makes it disappear; the cumulative-dollars frame makes it scream. Industries that profit from your inattention will always pick the friendlier frame for you. Your job is to re-frame it yourself — flip every gain frame to its loss frame and vice versa, and see if the decision survives.

Mental accounting — money is fungible, but we bucket it anyway

Here’s an iron rule of finance: money is fungible. A dollar is a dollar is a dollar; it has no label, no memory, no history. The $50 you found on the street spends exactly like the $50 you earned at work. And yet — you’d blow the found $50 on something frivolous and carefully save the earned $50. That gap is mental accounting: the human habit of sorting money into separate mental “buckets” and treating each bucket by different rules, as if the dollars weren’t interchangeable.

Richard Thaler named and documented this. We keep a “serious money” account (retirement, the mortgage) we’d never gamble with, and a “play money” account (a tax refund, a bonus, lottery winnings) we’ll cheerfully risk. We treat dividends as spendable “income” while guarding the principal as untouchable — even though selling $100 of shares and receiving a $100 dividend leave you in the identical financial position. The buckets feel real. They aren’t.

The house-money effect

The most dangerous bucket is house money — gains you’ve already made. After a win, people gamble more freely with the winnings than they ever would with their original stake, reasoning “it’s not really my money, it’s the casino’s house money.” The term comes straight from the gambling table, and it’s exactly the wrong instinct: those winnings are 100% yours, and a dollar lost from them hurts your net worth precisely as much as a dollar lost from your salary.

Worked example — $5,000 bonus vs $5,000 salary

You receive a $5,000 year-end bonus. It feels like a windfall — “house money” — so you punt it on a volatile meme stock you’d never normally touch.

Now imagine instead someone hands you $5,000 in salary and says “go put this in a meme stock.” Suddenly it feels reckless. You’d hesitate.

But look at your actual finances:

The two columns are financially identical. The bonus isn’t a special species of money that’s allowed to be gambled — it spends, compounds, and disappears exactly like every other dollar. The “house money” label is doing all the work, and it’s leading you to take a risk you’d reject if the same money wore a different label. That’s the error: not the bucketing itself, but letting an arbitrary label set your risk appetite.

The pitfall in reverse — mental accounting can be useful

Here’s the twist that most “behavioural bias listicles” miss: mental accounting is not always a mistake. The same bucketing instinct that makes you gamble bonuses can also make you behave better. A ring-fenced emergency fund you mentally label “do not touch” keeps you from raiding it for a holiday. A budget that splits money into “rent / groceries / fun” stops you overspending. A “retirement account I never look at” survives market crashes precisely because the label keeps your hands off it.

So the bias is two-faced. The error is when an arbitrary label drives an irrational risk choice — gambling house money, refusing to touch principal that should be spent, treating a windfall as license to be reckless. The feature is when a deliberate label enforces self-control — separating money so you don’t make a worse decision later. Same mechanism; the question is whether the label is helping your future self or fooling your present one.

Putting it together

A mixed recap pulling from all three biases:

Key Takeaways