Your securities. Your assumptions. Explore the allocation.
Choose a list of securities, set a goal and inspect the model weights produced under your assumptions and limits.
See how one limit changes the resultAllocation
Two securities. One model.
- Volatility A
- 20% / year
- Volatility B
- 30% / year
- Correlation
- 0
Fictional securities. No Practical-mode penalties.
From a list to a model
Build. Specify. Inspect.
Work with the securities you choose. Initial coverage is US-listed stocks and ETFs; no connected investment account or account balance is required.
Build your universe
Type or import tickers and company names. Review the resolved securities, then save your list.
Specify the calculation
Choose a goal, history window and allocation limits. Open advanced assumptions when you need more control.
Inspect the allocation
Review model weights totalling 100%, read the assumptions and adjust the inputs for another calculation.
One limit, a different result
What changes when a weight is capped?
Consider two fictional securities with assumed annual volatility of 20% and 30%, and zero correlation. The calculation minimises estimated variance with long-only weights totalling 100%.
Raw minimum variance
The same inputs, one new limit
A different allocation follows from a different constraint. It does not establish which portfolio is suitable for an investor.
This example uses Raw minimum variance, excluding Practical-mode penalties. The first result is rounded from 69.230769% and 30.769231%. These are hypothetical model outputs, not forecasts.
Read beyond the weights
Understand what the model produced.
The result becomes useful when its estimates, limits and data remain visible alongside it.
- Allocation and interpretation
- Inspect ranked weights, zero allocations and the limits that affected the calculation. A security on your list may receive no weight.
- Estimated characteristics
- Read estimated volatility, return and risk-adjusted return with their definitions and estimation basis. These are model estimates, not realised outcomes.
- Reference comparisons
- Compare the result with equal-weight and inverse-volatility references. The app identifies relevant differences in their limits.
- Assumptions and available data
- Check dates, estimation choices and warnings. Review missing-history exclusions before continuing with a reduced list.
- Saved results and exports
- Revisit the latest successful result locally, including offline. Export allocation CSV or the full result JSON with its original calculation context.
Choose the question
Start simply. Keep the assumptions accessible.
The default goal is minimum volatility. Other goals introduce return estimates or targets, so the assumptions deserve equal attention.
Explore the five calculation goals
- Minimise volatility
- Find weights that minimise estimated portfolio volatility under the chosen limits and any displayed penalties.
- Maximise risk-adjusted return
- Seek the highest modelled excess return relative to volatility, using the stated risk-free assumption.
- Target a volatility level
- Seek the highest modelled return within an estimated volatility limit. This limit is not a maximum possible loss.
- Target a return
- Seek lower estimated volatility while meeting an assumed return threshold. The threshold is not a promised return.
- Custom risk/return balance
- Set the strength of the model trade-off between estimated return and variance.
Explore the advanced controls
Review return-estimation and risk-estimation choices, weight bounds, the maximum allocated-security count and diversification preferences. Relevant assumptions appear with the chosen goal.
Practical mode adds its displayed limits and penalties. Raw mode removes automatic penalties. Explicit constraints are not silently loosened to produce a result.
Keep exploring
Understand the mathematics behind the allocation.
Strata Optimiser is planned. Model weights support analysis; they do not provide personalised investment advice or instructions to trade.