The 93% number to have ready when clients ask about selling at the top

Ben Carlson has published a simple tally of how the S&P 500 has spent its time since 1950: nearly 1,400 new all-time highs, which works out to 7% of all trading days ^[1][2]. The useful half of that arithmetic is the other side of it, because 93% of trading days were spent somewhere below a prior peak ^[3], and that is the base rate any conversation about "taking something off the table at the top" should start from.

The shape of those 93% matters more than the headline. Averaged across every trading day of the past 75 years or so, an investor was sitting in an 11% drawdown ^[4]. On 56% of all trading days the market was 5% or more below its last high ^[5], and more than 40% of the time it was in a double-digit drawdown ^[6]. Subtract one from the other and the mild-discomfort band is thin: at most about 16% of trading days fall between 5% and 10% off the peak ^[1]. Roughly 44% of days sit within 5% of a high ^[2]. The distribution is barbelled. Markets are either near the record or a long way under it, and they do not spend much time in the polite middle where a client might feel like there is time to reposition calmly.

Then there is the measurement problem, which is where the behaviour gets expensive. Up days outnumber down days only 56% to 44% ^[7], and losses register about twice as hard as equivalent gains ^[8]. Score that as Carlson does and each daily glance carries an expected emotional payoff of about minus 0.32 units: 0.56 units of pleasure against 0.88 units of pain ^[4]. He traces the idea to Richard Thaler's myopic loss aversion, which he wrote up in Risk & Reward ^[9]. The nearly 1,400 highs and the 7% share imply something like 20,000 trading days in the sample ^[3], so the observation frequency an investor chooses is doing more work on their experience of returns than the returns are.

Carlson's own rules are unfashionably manual. He keeps a spreadsheet of IRAs, 401ks, a 529 and brokerage accounts, and updates the values once every six months, skipping the exercise entirely during a nasty bear market ^[11]. His stated rule is never to check portfolio value during a bear market unless a change actually needs to be made ^[10]. He flags brokerage accounts as the weak point, since they are easier to check daily than tax-deferred accounts and he knows people who look multiple times a day ^[12]. The blunt version, in his words: checking your investments does not make them go up faster, and monitoring performance more often will not improve your Sharpe ratio ^[13].

The practical consequence is that reporting cadence is a policy decision, and it is one that has to be made while things are pleasant. Carlson notes the market is currently trading at or near all-time highs ^[14], which by his own numbers is the unusual state ^[3]. Set the check-in interval, the drawdown script and the pre-agreed rebalancing triggers now. What to watch is whether the 5%-to-10% band gets used as a reaction window at all, given how few days historically sit there ^[1], and whether brokerage-account behaviour diverges from retirement-account behaviour the way Carlson describes ^[12]. That gap is the tell for who has an investment process and who has a habit.