Robo-Advisers
A robo-adviser builds a portfolio from a questionnaire and maintains it automatically, charging an annual percentage on top of the underlying fund fees. The portfolios are conventional, so the decision is whether the automation is worth the fee against holding the same funds yourself.
A robo-adviser is a maintenance contract. It is not a different investment philosophy, and judging it as one produces the wrong answer in both directions.
How it works
A questionnaire produces a risk score, the score maps to a model portfolio, and the portfolio is built from low-cost index funds.
It rebalances automatically when weights drift, and directs new contributions toward whatever is underweight — which is the cheapest form of rebalancing and the one people most often skip.
The charge is an annual percentage of assets, on top of the expense ratios of the funds it holds. Both are deducted and neither arrives as an invoice.
A worked example
Take a platform charging 25 basis points holding funds averaging 8 basis points — 33 in total.
Against the same funds held directly at 8 basis points, the difference is 25.
On this site’s arithmetic a 20-basis-point annual drag removes 5.8% of a thirty-year pot and 75
removes 20.2%. A 25-point gap sits near the low end of that range, so the automation costs somewhere
around 7% of a thirty-year result. The figures are in research/series-measurements.json.
Against a traditional adviser at 100 basis points plus funds, the same platform is much cheaper. Both comparisons are valid and they point in opposite directions, which is why naming the alternative is the whole exercise.
What the fee actually buys
Rebalancing that happens whether or not you remember. Drift is the most commonly neglected maintenance task and it changes the portfolio’s risk silently.
A structure with no obvious knobs. On this site’s shared series 95% of bars sit below a prior peak and the longest stretch under water ran 73 bars — a portfolio you cannot easily tinker with is worth something during those.
And in some cases automated tax-loss harvesting, which is genuinely laborious by hand and worth real money in a taxable account at a high marginal rate.
What it does not do
It knows only what the questionnaire asked. It cannot coordinate with holdings elsewhere, does not know about a pension or a property, and cannot advise on anything outside the account.
The questionnaire measures a stated attitude, not behaviour. That is the same limitation described on the risk tolerance page, and automation does not fix it.
And it will not talk you out of anything. The client who sells everything in a bad month can do so in two taps, which is the one job a human adviser might genuinely earn a larger fee for.
Costs
Check what happens if you leave. Some platforms transfer holdings in kind and some liquidate, which in a taxable account turns a change of provider into a tax event.
Proprietary funds are the other lock-in. A portfolio built from widely available index funds moves easily; one built from the platform’s own funds usually has to be sold.
Tax
Automated rebalancing generates disposals you did not time. Better platforms rebalance with contributions first and harvest losses to offset, and the quality of that varies enormously — it is worth asking about specifically rather than assuming.
Comparing two platforms
Three numbers describe any of them and none of them is the headline percentage alone. The platform fee, the weighted average expense ratio of the funds it holds, and the two added together.
That third number is the one to compare, and it is frequently not published as a single figure — which means working it out from the model portfolio’s holdings rather than from the pricing page.
Then ask what happens on exit. Whether holdings transfer in kind or are liquidated is the difference between changing your mind cheaply and paying a tax bill for it, and it is settled by the platform’s policy rather than by anything you do.
And ask what the rebalancing actually does in a taxable account. A platform that rebalances with contributions and harvests losses is doing meaningfully more work than one that simply sells the overweight holding every quarter, and both are described with the same word.
The original data
Of the 24,971 unique videos in research/search-study-corpus.jsonl, zero have a title about
robo-advisers. Financial advisers appear in 5 videos at a median of 516,713 views and index funds in
132 at 69,951. The counts come from site/rank_investing.py, which deduplicates by video id.
Zero videos, against 5 on human advisers at a 516,713 median. The automated version manages very large sums and has no instructional coverage at all, which is the pattern for every subject on this site where nothing visual happens.
The answer to the question on that chart depends on what you would actually do afterwards. If the answer is “hold the same funds and rebalance annually”, the direct version is cheaper and identical. If the answer is “hold the same funds and never rebalance”, the platform was doing something worth paying for — and the honest input is your own record, not your intention.
When it fails
The failure is paying an automation fee for maintenance you would have done anyway. Somebody disciplined enough to research robo-advisers, compare fee schedules and read the methodology is usually also disciplined enough to rebalance twice a year — and the same portfolio is available at a fraction of the cost. The fee is small in any single year and it compounds for decades, which is how a reasonable-looking 25 basis points becomes a measurable share of the result.
The second failure is comparing against the wrong alternative. Against an expensive adviser it looks cheap; against holding the funds it does not.
A third is treating the questionnaire as a measurement. It records an attitude.
A fourth is running one alongside other holdings. It cannot see them.
A fifth is ignoring the exit mechanics. Liquidation on transfer is a tax event.
And a sixth is assuming the tax handling is good. The quality varies and it is worth asking.
Related
Financial advisers is the human version and its fee structure. Expense ratio is the layer underneath the platform fee. And the three-fund portfolio is broadly what these platforms build.
The comparison that matters is against the version of you that does not rebalance. If a portfolio drifts for eight years because nobody got round to it, the automation was cheap. If you would have done the maintenance anyway, the same portfolio is available for a fraction of the fee.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.