Portfolio-ManagementMedium
4 min read · 648 words
What the seat actually does
A portfolio manager decides what a fund owns, in what size, and when that changes. Everything about the seat is shaped by one document that arrives before any of it: the mandate, which says what the fund is for, what it may hold, and what it will be compared against.
Against a benchmark, not owning something is a position. If the index holds four per cent in a company and the fund holds none, the fund has a four per cent short against its own yardstick. That single fact reorders the whole job: the question is never "do I like this" but "how much of this do I hold relative to the thing I am measured against".
- Deciding — which names, which weights, which risks the fund is deliberately running.
- Sizing — the harder half. Being right in one per cent and wrong in six is a bad year.
- Rebalancing — because prices move the weights for you, and doing nothing is itself a drifting decision.
- Explaining — to a client, a board, or a consultant, in a language where "it went down" is not an answer.
A day, and where it goes
- Flows first — money in or out changes the portfolio before any view does, and cash is a position with a drag attached.
- The book against the benchmark — where the fund is over and under, by sector, by factor, by country, and whether any of that was intended.
- Research — internal, external, and the meeting that is supposed to change somebody's mind.
- Trading — usually handed to a dealing desk, because a manager who also trades is two jobs badly.
- The committee — where a position that has gone wrong is defended or cut, which is the part of the job that is actually hard.
What it is measured on
- Return against the benchmark, not return. A fund up eight per cent in an index up twelve had a bad year.
- Tracking error — how far the fund is allowed to stray. It is a budget, and a manager who does not spend it is charging active fees for an index.
- Information ratio — outperformance divided by the tracking error that produced it. See risk measures.
- Consistency over a cycle, because three years is not enough observations to separate skill from a style that happened to be in favour.
What it touches on this site
- What the fund holds — the whole atlas, but especially cash equities, fixed income and ETFs.
- How a portfolio is put together — portfolio practice, diversification and building an allocation.
- What the returns came from — what drives returns.
- What it costs the holder — costs and fees, and reading a factsheet.
How it goes wrong
- Closet indexing. A fund that holds the benchmark with a tilt charges for a decision it is not making, and the tracking error says so before the fee schedule does.
- Style drift. A value fund that quietly bought growth because growth was working has changed what the client bought.
- Liquidity that does not match the promise. Daily dealing over instruments that take weeks to sell is the 2019 shape.
- Confusing a good stock with a good position. Conviction is not a size, and the two are decided by different things.
Concepts to master
- Active risk is the currency. Everything the seat does is spending a tracking-error budget, and where it is spent is the strategy.
- An underweight is a short. The most useful sentence in this job and the one newcomers take longest to feel.
- Attribution before conclusions. Allocation, selection and currency are three different causes, and a good year from the wrong one is not repeatable.
- The mandate outranks the view. A trade the mandate does not permit is not a trade, however right it is.