AI glossary · Beginner · AI platforms & solutions · Finance & insurance
Robo-advisor
A robo-advisor is an online investment advisory and management service that uses a questionnaire to identify an investor's goals and risk tolerance, after which an algorithm automatically handles asset allocation and rebalancing (readjusting weights).
About this term
- Level
- Beginner · Core ideas for first-time readers
- Category
- Industry AI
- Fields
- AI platforms & solutions
- Industries
- Finance & insurance
- Added
- 2026-10-10
How it works
First, a questionnaire asking about age, investment horizon, income, how much loss you can tolerate and so on is used to classify the investor's profile. Once the profile is set, predesigned rules or models decide in what proportions to hold assets such as stocks, bonds and cash. Low-cost products such as ETFs (listed funds that bundle many securities) are typically used. When market movements cause the weights to drift from the initial targets, the system automatically buys and sells to restore them; this is called rebalancing. Investors check returns and changes in composition through reports in the app.
How a robo-advisor works, step by step
- 1QuestionnaireGoals, time horizon, risk tolerance
- 2Profile classificationConservative, aggressive, etc.
- 3Asset allocationSet the stock/bond/cash mix
- 4RebalancingAutomatically restores drifted weights
- 5ReportingCheck returns and holdings in the app
Why it matters
It lowered the barrier to investment advice by making asset management, once provided by human advisors, available at lower cost and with smaller minimum amounts. Another commonly cited advantage is that it keeps allocations to set rules without being swayed by emotion. Recently, it has been expanding to include LLM-based conversational advice that explains things in response to investors' questions, and to cover pension and tax planning as well.
Things to know
Because it recommends financial products, it is subject to investment advisory regulations in many countries, and firms must check whether products recommended fit the investor's profile (suitability) and whether the algorithm works as designed. A few questionnaire items cannot fully capture an investor's real situation, and when the same algorithm is applied to many users, similar trades can happen all at once during sudden market moves. Adding LLM-based advice also brings the risk of incorrect explanations due to hallucination, so the scope of answers is often limited and human review is put in place. The algorithm itself also falls under a financial institution's model risk management.
Example
In a brokerage or fintech app, when you answer questions such as "What is your target time frame?" or "What would you do if your investment fell 20%?", the app assigns a profile such as 'conservative' or 'aggressive' and proposes an ETF-focused portfolio. Afterwards, if the stock weighting rises or falls significantly, the app rebalances automatically and shows the results in a monthly report. Services that automatically manage pension accounts use the same approach.