Playbooks / Behavioural
An investment policy statement
The rule
Nearly every framework on this site ends in a number. This one ends in a page, and it is the only one whose entire value depends on being read at a specific moment: the moment its author would most like to do something else.
The case for it is simple and slightly insulting. Decisions made calmly are better than decisions made at speed under loss, and the person best placed to constrain you in March is you in November, when nothing is happening and you can think clearly. An investment policy statement is that constraint, written down.
Why one page
Because a document nobody rereads does not work, and length is what stops people rereading.
This is not a style preference. The failure mode of a twelve-page policy is not that it is wrong, it is that on the day it matters somebody opens it, sees twelve pages, and closes it. One page can be read standing up, in a bad mood, on a phone. That is the design requirement.
The lines that earn their place
What this money is for, and when it will be spent. Everything downstream follows from the date. Money needed in a year is a different asset class question from money needed in twenty, which is asset allocation by horizon, and the dated portion is where liquidity risk stops being theoretical.
The split, with a band around each number. Not sixty percent equities, but sixty percent with a band of five either side. A number without a band cannot tell you when to act; a number with one can, which is the whole argument in rebalancing bands.
What I will do when it falls twenty percent, and forty. Written in advance, in the first person, as an action. This is the line that most people skip and the only line that gets tested. Note that "nothing" is a perfectly good answer, and a much better one when it was written down beforehand than when it is improvised.
What would make me change the plan. State it now. A plan that can be revised on any Tuesday for any reason is not a plan, and a plan that can never be revised is a trap. The middle is naming the conditions in advance: a change in what the money is for, a change in horizon, a change in circumstances. Not a change in the market, because the market changing is the thing the plan exists to survive.
What I will not buy. Short and specific. Most portfolios are damaged less by the absence of a good idea than by the presence of a bad one that arrived persuasively.
A review date, on the calendar. Not a feeling. A rule consulted only when worried arrives too late to help.
The test for every line
Could somebody else follow it without asking you a question?
Apply that and most first drafts shrink by half. "Stay the course" fails. "Rebalance when any holding is more than five percentage points from its target, checked on 1 January and 1 July" passes. The failed lines are not wrong, they are sentiments, and a sentiment does not produce an action on a bad day.
Include the positions you never chose
The most common gap is not a bad rule, it is an incomplete inventory.
Employer equity is a position. A pension is a position. The home you live in is a position, and so is the currency your salary is paid in. For most people those are the largest holdings they own and the most correlated with each other, and a statement covering only the brokerage account is describing a corner of the portfolio while believing it describes the whole. Concentration limits and position sizing are the arithmetic; the statement is where the answer gets written down.
Why this belongs in a library of arithmetic
Because the arithmetic is not the binding constraint on most outcomes. The gap between what investments return and what investors actually receive is well documented and it is behavioural, not analytical, which is the subject of the behaviour gap. Nobody in that gap lacked access to the right numbers. They had the numbers and acted against them at the wrong moment.
One page, written while nothing is going wrong, is the cheapest available defence against that. It costs an afternoon, it is free, and unlike almost everything else in investing its benefit does not depend on being right about anything.
The arithmetic
Where it breaks
- Writing it during a good month and reading it during a bad one, without ever having tested whether it survives the reading. A rule that has not been rehearsed against a real fall is a guess about your own behaviour, and the evidence on that guess is not flattering.
- Stating intentions rather than triggers. Rebalance periodically and stay the course are sentiments. A band, a date and a named action are rules, and the difference only becomes visible on the day the two would produce different behaviour.
- Leaving out the positions nobody chose. Employer equity, a pension, a home, and the currency a salary is paid in are all positions, usually the largest and most correlated ones held, and a statement that covers only the brokerage account describes a fraction of the actual portfolio.
- Making it long. A document nobody rereads is a document that does not work. One page is not a stylistic preference, it is the constraint that decides whether the thing gets consulted at the moment it exists for.
- Never amending it. A plan that has never changed in five years is either remarkably well built or quietly ignored, and the way to tell is whether the review dates were kept. Amend it deliberately, on a review date, and write down why.
When to use it
Write it on an ordinary afternoon when nothing is happening, and read it on the worst day of the year. If the second half of that sentence never happens, the document was decoration.
Sources
- Bogleheads wiki
- Morningstar, Mind the Gap
- US Securities and Exchange Commission, investor bulletin on fees and expenses
Last reviewed . Educational research, not personal advice. Disclosure standards.