Why does the recent past feel like the future?

Two biases share one mechanism: a number or an experience gets into your head first, and everything afterwards is measured against it. One anchors you in time, the other anchors you to a price.

Recency: the last three years become the next thirty

Human memory weights the recent heavily. After a strong run, risk feels tamed and expected returns feel high; after a drawdown, the same asset feels permanently dangerous. Since allocation decisions get made in exactly these emotional states, recency reliably produces the worst possible sequence — more risk near the top, less risk near the bottom.

Watch what that does. Two investors, €100,000 each, in a market that falls 30% and then rises 30%.

• The chaser has just enjoyed three good years and holds 80% equities. The fall takes equities from €80,000 to €56,000 — the portfolio is at €76,000. Shaken, they cut to 40% equities. The recovery lifts equities from €30,400 to about €39,500. Final value: about €85,000.
• The steady investor holds a written 60/40 policy. The fall takes the portfolio to €82,000, at which point the rule says rebalance back to 60% equities — buying while it is uncomfortable. The recovery lifts equities to about €64,000. Final value: about €97,000.

Same market, same two moves, a gap of about €12,000 — around 12% of the starting capital. Note also that the steady investor carried less risk into the fall and more into the recovery. That is the whole trick, and it was performed by a document, not by nerve.

Anchoring: the number that arrived first

An anchor is any reference number your brain treats as meaningful because it saw it early. In markets, three anchors do most of the damage:

• Your purchase price. Already covered, still the champion.
• The 52-week high. A stock at 40 whose high was 100 feels "60% off." But regaining 100 requires +150%. Meanwhile a stock at 40 that has never traded above 45 feels expensive. Same price, same business prospects, opposite feelings — produced by a number that contains no information about future cash flows.
• A round number or a published target. €100, or an analyst's €62 target, becomes the frame within which all subsequent reasoning happens.

Anchoring is stubborn because it survives being pointed out. Knowing an anchor is arbitrary does not stop it working; experiments show people adjust away from anchors they know are random, and still adjust too little.

The defence

Anchors lose their grip when you generate your own number first. Estimate what an asset is worth — or what weight it deserves in your portfolio — before looking at where it has traded. Then look. If the two disagree, you at least have a real disagreement rather than a slow drift toward whatever number was on the screen.

For recency, the defence is the same one the numbers above showed: a written allocation policy that pre-commits the response, made before there is anything to respond to.

Try it now

1. Go through your watchlist and, for each name, write down the first number that comes to mind. Recognise how many of them are purchase prices or 52-week highs — neither of which is a fact about the business.
2. Here is the last twelve months of one fund. Write your impression of it in a single sentence before you read on.

Now the same fund across five years. If your sentence changed, you have just watched recency operate on yourself.

3. Write your target allocation on paper and date it. That dated line is what a future frightened or excited version of you will have to argue with.