What the Acubic Risk Quiz Actually Determines

Key takeaways

• The risk quiz converts your answers into hard constraints on the optimizer, not soft guidance it can override.
• The volatility tolerance answer is the most consequential single input: it sets the per-asset concentration cap applied throughout every rebalance.
• Conservative, moderate, and aggressive risk levels cap each holding at 10%, 15%, and 25% of the portfolio respectively.
• Experience level gates which questions appear: Beginners see fewer inputs and defaults apply for the rest; Advanced users skip the volatility question entirely.
• Asset classes, regions, and sectors define which holdings can enter the portfolio; position sizing and asset count then shape how the optimizer distributes capital across them.

The first screen most new users encounter is the risk quiz. Before Acubic can build a portfolio, it needs to understand your goals, your limits, and your tolerance for loss. But the quiz is not a vague profiling exercise. Each answer controls a specific parameter in the construction process, and the constraints that result are enforced by the optimizer rather than noted as preferences. This article goes through each question in order and explains what it actually changes in the portfolio.

Experience level gates the quiz itself

The first question asks for your investing experience: Beginner, Intermediate, or Advanced. This answer does not directly affect the portfolio. What it does is determine which further questions appear.

A Beginner sees the simplest version of the quiz: volatility tolerance, asset classes, regions, and sectors. Asset count and position sizing are hidden, and the optimizer uses default ranges for those inputs. An Intermediate user sees all of those questions plus asset count, position sizing, and universe filters. An Advanced user sees the full set of inputs with no guided recommendations, and the volatility question is hidden entirely, replaced by a default moderate setting. The methodology page describes the design principle behind this: construction inputs are constraints on the optimizer rather than hints it can set aside. Advanced users are expected to control risk through position sizing and universe rules directly, rather than through a guided tolerance band.

The volatility question: the most consequential answer

For Beginner and Intermediate users, the next question asks how much of your portfolio value you are willing to see fall in a bad period before wanting to act. Four options are presented:

• Low (less than 10%): stable, bond-like behavior
• Medium (10 to 20%): balanced, comparable to a broad equity index
• High (20 to 35%): aggressive growth
• Unconstrained: no predefined limit

This answer maps to a risk profile level: conservative, moderate, or aggressive. That level then sets the per-asset concentration cap the optimizer must respect. Conservative limits each holding to 10% of the total portfolio. Moderate limits each to 15%. Aggressive allows up to 25%. The unconstrained option also maps to the aggressive level for constraint purposes.

These are hard limits. The optimizer cannot exceed them regardless of what the correlation structure or ranking signal would otherwise suggest. If a concentrated position would produce a better risk-adjusted score under the chosen method, the cap still applies. The methodology page is explicit on this: constraints are applied as bounds on the optimizer rather than penalties in the objective. Where the cap cannot be satisfied arithmetically, for example a 10% maximum across only eight holdings cannot sum to 100%, the documentation states that the binding limit is relaxed to the nearest feasible value rather than the run failing.

For small portfolios, the caps adjust with the number of holdings. With n holdings, the per-asset maximum rises to 1.5 divided by n, 2 divided by n, or 3 divided by n for conservative, moderate, and aggressive profiles respectively, when that figure is higher than the base cap. With eight holdings, that works out to 18.75% for conservative, 25% for moderate, and 37.5% for aggressive.

The risk level also shapes the algorithm recommendation shown to Beginner and Intermediate users after the quiz. Conservative profiles are directed toward Minimum Variance or CVaR minimization, methods that prioritize limiting downside. Moderate profiles are directed toward Hierarchical Risk Parity, which spreads risk across correlated groups rather than maximizing raw return. Aggressive profiles are directed toward mean-variance optimization. Advanced users see all three optimizers with no recommendation. The algorithm and the risk level are separate choices: the level sets the bounds, and the algorithm runs within them. For more on how Hierarchical Risk Parity distributes risk across correlated clusters rather than chasing individual rankings, see why Acubic uses hierarchical risk parity.

Asset classes, regions, and sectors: the eligible universe

Three questions define which holdings can enter the portfolio at all. These are filtering inputs: they narrow the candidate universe before screening, ranking, and allocation begin.

Asset classes determine the types of instruments eligible: stocks, ETFs, commodities, bonds, and crypto. Leaving this unset keeps all classes available. Selecting a subset restricts the universe to those types only.

Regions narrow by the company's country of risk across five geographic groups: North America, Latin America, Europe, Asia-Pacific, and Middle East and Africa. An empty selection keeps all regions open.

Sectors narrow by industry classification across eleven options including Technology, Healthcare, Financials, and Energy. A sector selection limits the universe to companies and funds classified within those sectors. An empty selection leaves the full sector space open.

These three together define the pool of candidates that screening then filters by price history and liquidity, that the ranking stage sorts, and that the allocation stage weights. A narrow universe produces a more concentrated shortlist; a wide universe gives the optimizer more material to diversify across. If a ticker you expected in your proposal is absent, the most likely explanation is a universe constraint that removed it before ranking began. The article on how Acubic selects assets for your portfolio covers the screening and ranking stages in detail.

Asset count and position sizing: the shape of the allocation

Intermediate and Advanced users also set two structural constraints: how many holdings the portfolio should target, and the minimum and maximum weight any single holding may carry.

Asset count is entered as a range: Concentrated (5 to 10), Balanced (15 to 25), Diversified (25 to 40), or Ultra-wide (40 to 50). The optimizer stays within this band. The count range also influences the recommended position sizes: a wider portfolio requires smaller per-asset weights to remain feasible, and the quiz adjusts its suggestions as the count grows. Intermediate users can override these suggestions by moving the position-size sliders directly.

Position sizing sets a minimum floor and a maximum ceiling per asset as a percentage of total capital. A minimum weight ensures no position in the output is too small to be meaningful. A maximum weight caps concentration independently of the risk-level cap; the stricter of the two applies. If you set a maximum lower than the risk-level cap, your value wins. If the risk-level cap is lower, it wins instead.

For Beginners, asset count and position sizing are not shown. Default ranges apply: 15 to 25 holdings, a 1% minimum weight, and a 10% maximum weight per position. These defaults produce a reasonably diversified first proposal for a reader who has not yet expressed a preference on those dimensions.

Universe filters: additional company-level rules

The last question, visible only to Intermediate and Advanced users, allows rules that a company must pass before it can be selected. These filters run before screening, ranking, and allocation. A company that fails a filter is excluded from the entire run, regardless of how it would rank otherwise.

The methodology page documents that every ranking method already applies a baseline liquidity floor: an asset must exceed a minimum median daily dollar volume as of the ranking date. Universe filters are additional constraints layered on top of that floor, letting you restrict by financial metrics or other criteria specific to your strategy.

From inputs to hard constraints

The methodology page states the design principle directly: constraints are applied as bounds on the optimizer rather than penalties in the objective. That distinction matters in practice. A penalty approach allows the optimizer to violate a preference when doing so improves the overall score. A bounds approach does not: the constraint is a hard limit that the output must satisfy.

Your risk tolerance cap, your position size floor and ceiling, your asset count range, your asset class and sector selections, your universe filters: none of these are hints. They are structural properties of the run. A proposal that respects them all is the only proposal you will see. The guide to reading an Acubic portfolio proposal explains what this means when you look at the output: why a holding might carry a lower weight than its ranking position suggests, and what an outsized weight on a single asset usually indicates about the constraint structure the optimizer was working within.

What you can change after the first proposal

Nothing is committed at the proposal stage. You can re-run the quiz with different inputs to generate a new proposal from the same pipeline. Changing the risk level tightens or loosens the concentration cap and updates the algorithm recommendation. Changing the universe produces a different candidate pool. Changing position sizing adjusts the bounds the optimizer works within.

Some changes produce results that are surprising at first. Narrowing the universe to a single region and a single sector can cause the optimizer to concentrate more than expected, because the assets remaining after filtering may all move together and the algorithm has limited material to spread risk across. Widening the universe typically produces a less correlated shortlist, and the allocation changes accordingly.

The complete guide to how Acubic works covers each stage from the quiz through to the broker connection and scheduled rebalancing. The AI portfolio builder is where you start a new proposal or adjust an existing one.

Acubic does not promise returns and cannot predict short-term market moves. The quiz sets the constraints; the optimizer works within them; the proposal reflects your stated limits, not a forecast of what those limits will earn.