How to Buy an Index Fund
To buy an index fund, choose which index you want to own, then the account it will sit in, then the cheapest fund that tracks that index properly. Set the contribution to repeat automatically. The fee is the one variable you control, so it decides the choice.
Most guides to this start with which fund to buy, which is the third decision, not the first. Getting the order right removes most of the difficulty, because each choice narrows the next one.
Before you start
A brokerage or retirement account you can already log into. Opening one is a separate task and mixing the two is how this gets abandoned halfway.
An amount you will not need for ten years. Not a target, a floor. Money that has to come back out inside a few years belongs somewhere else entirely.
The fund’s total expense ratio in writing. Not the headline fee — the total, from the fund’s own factsheet. This is the number that decides between two otherwise identical funds.
The steps
1. Choose the index before you look at any fund
A broad domestic index, a global one, or a bond index. This single decision determines what you own. Everything after it is administration.
2. Choose the account before the fund
A tax-advantaged wrapper beats a taxable account holding the identical fund. The wrapper outranks the fund choice, so settle it first.
3. Compare only the total expense ratio
Two funds tracking the same index hold the same companies. The fee is the difference. Sort by it and take the cheapest that clears the next step.
4. Check how closely it actually followed the index
The factsheet reports tracking difference against the benchmark. A fund that is cheap and tracks badly has saved you nothing. Reject anything with a persistent gap.
5. Read the total-return figure, never the price chart
A price chart excludes dividends. For a broad index over decades the income is a large share of the result, so comparing price charts compares the wrong quantity.
6. Set the contribution to repeat automatically
A standing instruction on a fixed date. This converts an ongoing judgement call into a decision you made a single time, which is the entire behavioural mechanism.
7. Then leave it alone for a very long time
No monitoring schedule, no rebalancing between similar funds, no switching for a small fee saving. The plan was finished at step six.
How to tell it worked
Check three things 12 months after you start, and nothing before that.
One: the contribution ran every month without you touching it. 12 out of 12 means the automation holds. Anything less means the mechanism failed, which is a different problem from the market.
Two: you made zero switches. Every switch costs — a round trip on the shared price series is 2% of a median bar’s range — and almost always chases a fee gap smaller than the cost of moving.
Three: your fund’s 12-month return sits within a fraction of a percent of the index it tracks. A wider gap than that is the tracking problem from step four showing up in your own account.
None of these three is about the return. The market decides that. These are the only parts of the outcome that were ever yours.
What the fee actually costs
Compounded over thirty years, the charge alone removes a fixed share of the final pot. On this
site’s arithmetic: 5 basis points costs 1.5%, 20 basis points costs 5.8%, 75 costs 20.2%, and 150
costs 36.5%. The figures are in research/series-measurements.json.
That last number is the argument in full. A fund charging 1.5% a year takes over a third of a thirty-year result, and it does so whether the market rises or falls. Nothing else in this procedure has that kind of leverage on the outcome.
And it is the one input you set yourself. You cannot choose the return. You choose the fee once, at step three, and then live with it for decades.
The original data
Of the 24,971 unique videos in research/search-study-corpus.jsonl, 26 have an instruction-shaped
title mentioning index funds, at a median of 88,014 views across 20 channels, with a maximum of
2,258,643. That is the third-highest median of any instructional subject measured. The counts come
from site/rank_howto.py, which deduplicates by video id before counting.
Twenty-six videos at a median of 88,014, against 562 TradingView tutorials at 2,575. The largest audience in this entire corpus is for the simplest possible procedure, and almost nobody is making it — because it takes seven steps and then tells you to stop.
The answer to the question on that chart is that a year is not a measurement. On the shared series, 95% of bars sit below a prior peak, the median drawdown is 1.36%, and the longest stretch below a peak runs 73 bars. Being underwater is the ordinary condition of holding anything. Selling converts an ordinary state into a permanent one — which is precisely why step six removed the decision from you.
When it fails
In a flat decade the charge is the only certain event. A tracker returns the index minus its costs, so when the index goes nowhere the fee is the entire measurable outcome — and a 150 basis point fund has taken a real amount of money for delivering a flat line. It is the strongest argument for step three and the one nobody makes until it has already happened to them.
The second failure is reacting to an opening gap. A sharp fall feels like information and is almost never relevant to a holding period measured in decades.
A third is buying the fund before choosing the account. The tax treatment outranks the fund, and unwinding it later means selling and repurchasing.
A fourth is comparing price charts across funds. Excluding dividends compares the wrong number, and the error grows with time.
A fifth is switching to save a few basis points. The trading cost usually exceeds the saving, and the saving was never the reason you were down.
And a sixth is checking it constantly. The procedure was designed to be finished. Monitoring reintroduces exactly the decision step six removed.
Related
Index funds explains what a tracker holds and why the index rules matter more than any manager. ETF investing covers the exchange-listed version and how it differs at the point of purchase. And tracking error is the measurement behind step four, and why a cheap fund that tracks badly is not cheap.
The part I got wrong early was treating this as a stock-picking problem with a shortcut attached. It is not. Once you have chosen the index, every fund tracking it owns the same companies in the same proportions, and the only things separating them are the fee and how tightly they track. That reframing turned a research project into a two-minute comparison.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.