Knowing when to do nothing

The instinct to do something when circumstances are uncomfortable is well documented across domains and is particularly costly in investing.

Action bias

The general finding.

People prefer action to inaction, particularly under pressure and particularly when an outcome feels controllable.

Studies across various domains find people acting where the evidence favours waiting, and feeling worse about bad outcomes following inaction than following action.

Which means the default is to act, and the effort required to not act is greater than the effort to act.

Why it is costly here

Specifically.

Each transaction has a cost.

Each decision can be wrong.

Research on retail trading consistently finds that more frequent activity produces worse net outcomes.

And the moments when action feels most necessary — during declines, after a strong run, following alarming news — are precisely the moments when action is most damaging.

The situations where doing nothing is correct

A list worth having.

A market decline of any magnitude, where the plan already anticipates it.

Alarming news, which is already reflected in prices.

A forecast of any kind.

Another asset performing better than yours.

A holding underperforming over a period shorter than the strategy's horizon.

An annual review finding drift within the rebalancing threshold.

A recommendation from someone with no knowledge of your objectives.

And the general feeling that you ought to be doing something.

The situations requiring action

Also worth listing.

A change in objectives, horizon or circumstances.

Allocation drift beyond the rebalancing threshold.

A material increase in costs, or a materially cheaper equivalent becoming available.

A fund changing its mandate or announcing closure.

An annual allowance about to expire unused.

Reaching a planned de-risking milestone.

And correcting an error, such as a contribution not being invested.

Which is a short list, and everything not on it is generally noise.

The written plan as permission

Its most useful function.

A plan stating what would and would not cause a change converts the question from what should I do into what did I decide.

Which provides explicit permission not to act, and which is considerably easier than resisting the impulse without one.

And which means the discomfort of inaction is attached to a prior decision rather than to a current failure to respond.

The delay technique

Practical.

Impose a mandatory waiting period before any change: a week for anything routine, a month for anything substantial.

Write down the proposed change and the reasoning at the point the impulse arises.

Revisit it after the period.

A substantial proportion of proposed changes are abandoned, which is the point, and the record of abandoned changes is itself informative.

Making inaction feel like action

A useful reframe.

Reading your own plan is an action.

Recording the decision not to act, with reasoning, is an action.

Checking the allocation against the threshold and confirming no rebalancing is required is an action.

Which satisfies the need to have done something without doing anything to the portfolio.

And which produces a record that makes the next occasion easier.

The productive alternatives

Where the energy can go.

Increasing the contribution rate, which affects the outcome more than any portfolio change.

Reducing costs, which is certain rather than speculative.

Reviewing fixed household costs.

Making or updating a will and powers of attorney.

Checking beneficiary nominations.

Producing the summary document for a partner or executor.

And doing something entirely unrelated to money, which is generally the best use of the impulse.

The hardest case

Worth acknowledging.

During a severe decline, with a coherent explanation available and other people acting, doing nothing feels like negligence rather than discipline.

Which is why the decision must be made in advance, in writing, with the reasoning recorded.

The version of you reading it during a crisis will not construct that reasoning independently.

And the record that it was written beforehand is what gives it authority.

The summary

Which is the whole argument.

For a long-horizon investor with a diversified portfolio and automated contributions, the correct action in almost all circumstances is none.

Which is unsatisfying, provides no sense of control, and is what the evidence supports.

General information only, not investment advice. Investments can fall in value and past performance does not indicate future returns. Consult a regulated financial adviser.