Investment Optimizer

Your Portfolio

Total Value —

Holdings —

Annual Dividend —

Next Year Forecast —

What would you like to do?

Portfolio Health

Select a portfolio, then open the Discover tab to check ROTS health.

Portfolio

Manage Investments

Investment ID:

Choose Investment:

New Shares:

Transaction Management

Transaction Type:

Ticker Symbol:

Transaction Date:

Currency:

Price per Share (SEK):

Notes (optional):

Simulate Investment

Set your budget, constraints, and optional selling toggle. Then press Simulate Investment to see suggestions.

Investment Amount (SEK):

Max per Stock (% of total): Defaults: min per stock 10%, min stocks 1

Minimum Investment per Stock (SEK):

Maximum Investment per Stock (SEK):

Minimum Number of Stocks:

Maximum Number of Stocks:

Max New Investments:

Include Selling of Existing Investments

Veto Strategy

Enable Veto Strategy

Candidate Model: Model for selecting candidates

Veto Model: Model for veto decisions

Candidate Top %: Top X% to consider as candidates

Veto Bottom %: Reject candidates in bottom Y%

AI Second Opinion

Enable AI Second Opinion

AI Opinion Mode:

Withdrawal Amount (SEK):

AI User Note (optional):

Model Backtesting

Evaluate different model weight sets using historical stock data snapshots. The backtest calculates Information Coefficient (IC), top decile returns, and hit rates for each configured weight set.

Months Ahead (Prediction Horizon): How many months ahead to predict returns (default: 6)

Max Days Difference: Maximum acceptable days between snapshot pairs (default: 30)

Rank Dynamics Analytics

Analyze how stock rankings evolve over time across different models. Re-scores historical snapshots with current weight sets to compute persistence, acceleration, convergence, and regime indicators.

Max Snapshots to Analyze: Number of most recent snapshots to use (default: 50)

Portfolio stocks only Only show stocks in the selected portfolio (rankings still computed against all stocks)

Veto Strategy Backtesting

Test your veto strategy: candidate model selects top-ranked stocks, veto model filters out risky ones. This tests whether the veto improves candidate quality without simulating full portfolio construction.

Candidate Model: Weight set for selecting candidates (default: "old")

Veto Model: Weight set for veto decisions (default: "trained")

Candidate Top %: Top X% of stocks to consider as candidates (default: 20)

Veto Bottom %: Reject candidates in bottom Y% of veto model (default: 50)

Months Ahead: Prediction horizon in months (default: 6)

Max Days Difference: Maximum days between snapshots (default: 30)

Non-overlapping periods (robustness check - reduces periods but ensures independence)

Model Training

Run backtest_optimize.py from the UI, review the resulting weights inline, and apply them to the live trained weight set without restarting the app.

Months ahead: Prediction horizon for the optimizer

Return weight: Exponent on realized return (Sharpe proxy)

⬤ IDLE

Investment Optimizer

Your Portfolio

What would you like to do?

Portfolio Health

Portfolio

Stock Details

Manage Investments

Transaction Management

Simulate Investment

AI Second Opinion

Model Backtesting

Rank Dynamics Analytics

Model Distance Over Time

Veto Strategy Backtesting

Model Training

Feature Correlations & Weights

Stock Screener

User Management

New User