An Information Diet: What to Read and How Often

Scope, and this article will not answer its own title as posed. It recommends no frequency, publishes no source list, and does not tell anyone how often to look at their holdings. The reason is given at the end and is not caution: the right frequency depends on what a person does with what they see, which this portal has no way of knowing. What the article does supply is the arithmetic of observation frequency, which is computable and is almost never stated. Figures use the Pillar 22 parameter set — a 9.0% expected return with 16.0% volatility — and are illustrative teaching values, not forecasts.

What the observation schedule does

Taking the canonical parameters and asking a single question — what is the chance that a given observation shows a loss?

The one-year figure of 28.7% matches the calculation in Risk Tolerance and Risk Capacity Are Different Things, which is the same arithmetic asked for a different purpose. (Each row is a normal approximation with mean 9.0% × t and standard deviation 16.0% × √t; the loss probability is the normal tail below zero.)

Worked example — the same year, seen through different schedules. A daily observer takes about 252 readings a year, of which roughly 122 show a loss. An annual observer takes one, which has about a 29% chance of showing a loss. The portfolio is identical in both cases. The holding period is identical. The return is identical. What differs is that one person experiences roughly 122 pieces of bad news a year and the other experiences about a third of one. The mechanism is that short horizons have almost no expected return to overcome their variability — over a single day the expected gain is 0.04% against a standard deviation of 1.01%, so the outcome is very nearly a coin toss — while over a decade the expected gain of 90.00% dominates a standard deviation of 50.60%. The return is a property of the holding. The experience is a property of the observation schedule. Figures are illustrative teaching values and are not forecasts; every one changes if the assumed parameters change.

Why that interacts badly with what Pillar 32 has already established

Prospect Theory establishes that losses are weighted more heavily than equivalent gains — with the direction robust and the magnitude smaller and more context-dependent than the textbook figure suggests.

Combine that with the table above and the consequence is arithmetic rather than psychological. Frequent observation does not merely produce more information. It produces disproportionately more of the specific kind of information that carries extra weight — and it does so without altering the underlying return in any way. This is the mechanism studied under the name myopic loss aversion, and the direction of the finding is that more frequent evaluation is associated with less willingness to hold the variable asset — reported here as the direction of an effect, not a prescription.

Two further findings from this pillar compound it. Reading Financial Media Critically established that a daily reader receives a daily volume rather than a daily signal. And Why Forecasts Fail established that information without scored feedback does not improve calibration — so additional observation is not self-correcting: it supplies more to react to without supplying any means of learning whether the reactions were right.

The part that is genuinely undetermined

None of the above says what frequency is appropriate, and the reason is specific rather than evasive.

The cost of observation depends entirely on what follows it. A person who looks daily and acts on nothing bears no cost from looking — the arithmetic above describes their experience, not their outcome. A person who looks annually and acts on every observation may transact more consequentially than the first. Observation frequency and action frequency are different variables, and only the second appears in the cost arithmetic in Overtrading.

There are also reasons to observe that have nothing to do with trading — verifying that records are correct, noticing an unauthorised transaction, checking that a contribution arrived. An article that recommended looking less would be recommending against those too.