Why does checking more often make the same portfolio feel riskier?

Why this matters. Nothing about the holdings changes. The observation changes, and the observation is one of the few things here fully under your control.

Watch a portfolio once a quarter and it looks like a line that rises. Watch the same portfolio every ten minutes and it looks like a fight.

The finding

Replicated, and it is the most directly actionable result in this course: Benartzi and Thaler (1995) explained the equity premium puzzle through myopic loss aversion — frequent evaluation makes a risky asset look worse, because at short horizons losses appear more often. Thaler, Tversky, Kahneman and Schwartz (1997) tested it directly: changing how often people saw results changed how much risk they took. Gneezy and Potters (1997) found the same.

Same holdings, same horizon, different observation frequency, different behaviour.

Why frequency does it

At short intervals, movement is mostly noise, and noise is roughly symmetric — which means you see a great many small losses. Because losses register more heavily than equivalent gains — a population finding Portfolio Management teaches in Loss aversion and the disposition effect — a stream of symmetric noise nets out, in experience, as unpleasant.

You are not gathering more information. You are sampling noise more often and paying for each sample.

The uncomfortable implication

Attention is not free and not neutral. Checking is an action with an effect on your decisions, and it is usually the one action nobody counts.

In the data

The same fund, four clocks. Below is the S&P 500 fund over its whole history; switch the bars between Monthly, Weekly and Daily and watch how much more of the line is made of down-steps the finer the clock gets.

Counted on the adjusted close (measured 7 September 2026): of 240 calendar months from September 2006 to August 2026, 67% closed up; of 80 quarters, 75%; of 19 calendar years (2007 to 2025), 84%. Day by day, 55% of the 5,028 day-over-day changes in those 5,029 sessions were up. Nothing about the holding changed between those four numbers. The fraction of red windows you meet is set by how often you look, and at the daily clock you meet a loss almost every other time. That is the whole mechanism, and it is arithmetic on one series.

The artefact

A check counter. For one week, a tally mark each time you look. No judgement, no target, no resolution to look less.

Almost nobody guesses their own number correctly, and the gap between the guess and the tally is the finding.

Try it now

Write down your estimate for yesterday. Then start the tally today, and compare at the end of the week.