Loading...
Loading...
Loading news...
Select a portfolio to load news for its tickers.
Add tickers to your watchlist to see their news.
Enter ticker symbols and share quantities. The system fetches prices, builds an RMT-cleaned covariance matrix, runs optimisation, and produces a full rebalancing plan.
Add mixed assets with their current dollar allocation. The optimizer finds the best target weights, and shows a rebalancing plan to move from current to optimal.
Enter tickers to optimise. Leave empty to use all tickers from Stocks.xlsx. Applies Marchenko-Pastur covariance cleaning, quality scoring, and sector caps.
Add any asset — stocks, options (call/put, American/European), bonds, futures, commodities, crypto. Type is auto-detected, extra fields appear per type.
| Ref. | Mode | Note | Timestamp | Assets | Value | Status | |
|---|---|---|---|---|---|---|---|
| Loading... | |||||||
| Factor | Category | Mean IC | ICIR | Hit Rate | L/S Spread | Periods |
|---|
| Period | Ann. Return | Volatility | Sharpe | Sortino | Max DD | Calmar |
|---|
| Signal | Category | IC (1M) | IC (3M) | Hit Rate | L/S Spread | T-Stat | p-value | Ann. Spread |
|---|
Loading...
Loading...
Loading...
Enter purchase price (cost basis) per share. Purchase date enables long/short-term classification.
| TICKER | TYPE | SHARES / UNITS | PURCHASE PRICE ($) | PURCHASE DATE |
|---|
MAISNER is currently in closed beta and is provided strictly for testing and evaluation purposes. As beta software, it may contain bugs, errors, inaccuracies, or unexpected behaviour. Features may change, be removed, or become temporarily unavailable without prior notice.
Nothing on MAISNER constitutes investment advice, financial advice, trading advice, or any other type of advice. All analytical outputs — including portfolio weights, optimization results, factor scores, backtest results, stress tests, signals, and risk metrics — are provided for informational and educational purposes only.
Before making any investment decision, consult a qualified and licensed financial advisor. MAISNER is not a licensed financial advisor, broker, or investment manager.
To the maximum extent permitted by applicable law, the developer of MAISNER shall not be liable for any direct, indirect, incidental, special, consequential, or punitive damages arising from:
Market data is sourced from third-party providers (Financial Modeling Prep, Polygon.io, Yahoo Finance). The developer does not guarantee the accuracy, completeness, timeliness, or reliability of this data.
Beta access is granted on a strictly personal and confidential basis. As a beta tester you agree:
These terms may be updated at any time. Continued use of the platform following notification of changes constitutes acceptance of the revised terms.
MAISNER pulls 10 years of monthly price history and fundamentals for every ticker in your database. When you run an analysis, it builds a covariance matrix cleaned with Random Matrix Theory (Marchenko-Pastur filter), then runs a Markowitz Mean-Variance optimisation tilted toward quality factors — ROE, margins, and debt coverage.
The result is not just optimal weights — it's a full rebalancing plan with exact share counts, Monte Carlo projections, sector exposure, and correlation maps.
MAISNER is in closed beta. All outputs are for informational purposes only and do not constitute investment advice. Always verify results independently before making any financial decision.
US tickers work out of the box (AAPL, MSFT, NVDA…). European tickers are auto-resolved: just type the base symbol (e.g. SAP, ASML, ENI) and the system will find the right exchange suffix (.DE, .MI, .L, .PA etc.).
ETFs are fully supported and get a combined 50% sector cap to prevent over-indexing.
Curr % — your current allocation based on market value of shares held.
Opt % — the mathematically optimal allocation for your selected mode.
Δ Shares — exact number of shares to buy (+) or sell (−) to reach optimal weights.
10,000 simulated paths over 10 years based on the portfolio's expected return and volatility. Shows the 5th percentile (pessimistic), median, and 95th percentile (optimistic) outcomes.
Use the median as a rough planning number, not a guarantee.
Shows pairwise return correlations between positions. Low correlation means better diversification. High correlation (>0.8) means two positions move together and you're not getting real diversification benefit.
Standard covariance matrices contain noise — random correlations that look real but aren't. MAISNER applies Random Matrix Theory (Marchenko-Pastur filter) to separate signal from noise before optimization.
When there are too few observations, it blends RMT with Ledoit-Wolf shrinkage to ensure the matrix is stable and invertible.
The optimizer doesn't purely maximize Sharpe. It applies a quality tilt: tickers with higher ROE, stronger margins, and lower debt get a small expected return bonus. This biases the result toward fundamentally strong companies.
No single sector can exceed 40% of the portfolio. ETFs have a combined cap of 50%. This prevents the optimizer from concentrating everything in one hot sector.
For portfolios with fewer than 8 tickers, sector caps are disabled to avoid infeasibility.
Optimized portfolio drawdown significantly smaller than current portfolio. Recovery time shorter. Max loss within acceptable range for your risk tolerance.
Optimized portfolio performs similarly to current during a crisis — means concentration risk is still present. Consider adding uncorrelated assets (bonds, gold, low-vol stocks).
X-axis — Annualized Volatility (risk). Lower is safer.
Y-axis — Annualized Expected Return. Higher is better.
The curve is the efficient frontier — portfolios on it cannot be improved (you can't get more return for the same risk). Portfolios below the curve are suboptimal.
Your current portfolio dot shows where you sit relative to the frontier. The optimizer moves you onto it.
Each ticker has a slider that lets you manually override its weight. As you move sliders, your portfolio dot moves on the frontier in real time — you can see exactly what trade-off you're making.
Use this to understand how much return you sacrifice by constraining a position (e.g. "I want at most 10% in NVDA").
Computes 11 quantitative factors for every ticker and analyzes their predictive power. Shows which factors historically predicted returns best in your universe.
Write a simple strategy in plain text, then run a full walk-forward backtest against 10+ years of price history.
IS/OOS split: results are split into In-Sample (training) and Out-of-Sample (test). A strategy that only works IS is overfit. Good strategies maintain Sharpe OOS.
Tests each factor as a trading signal. Shows hit rate (% of times a top-ranked stock actually outperformed), long/short spread, statistical significance (T-stat), and turnover cost estimate.
The Ensemble section finds optimal weights to combine all signals using OLS regression — giving you a composite score that's better than any single factor alone.
MAISNER uses a beta-adjusted benchmark for fair comparison. If your portfolio has Beta 0.6, comparing against raw SPY is unfair — the benchmark becomes:
This means you're compared against a portfolio with the same market exposure as yours — not against a fully-invested equity benchmark.
MAISNER uses a strict IS/OOS (In-Sample / Out-of-Sample) split. The default is 60/40: 60% of history is used to calibrate the strategy, 40% is held out as a true test.
A strategy that looks great IS but collapses OOS is overfit to historical noise. Good strategies maintain meaningful Sharpe ratios on data they've never "seen".
When options are present, MAISNER computes portfolio-level Greeks — the net Delta, Gamma, Theta, Vega, and Rho of all option positions combined. This tells you the portfolio's aggregate sensitivity to price, time, and volatility moves.
Options max allocation is capped at 30% of total portfolio by default.
For portfolios with options, standard Mean-Variance Optimization is replaced with CVaR optimization using 5,000 Monte Carlo scenarios. This correctly handles the non-linear payoff structure of options — Markowitz cannot.
A broad S&P 500-like shock applied to all equity positions. Enter as a percentage (e.g. −30 for a crash scenario, +15 for a bull run).
Apply additional shocks to specific sectors on top of the market shock. Use + Add sector to add rows. E.g. a banking crisis might apply −50% to Financials while the broad market falls −20%.
Override any individual position with a precise shock. E.g. you expect AAPL to fall −40% in a specific scenario — add a ticker shock to override the sector result.
Enter your positions (ticker, shares, purchase price, purchase date). The system fetches the current market price, calculates your unrealized P&L, and identifies positions with losses that qualify for tax-loss harvesting.
For each candidate, it suggests a substitute security — a highly correlated but legally distinct asset you can buy immediately to maintain market exposure while the wash sale clock runs.
MAISNER flags positions that may trigger wash sale rules and marks them clearly. The tool does not execute trades — always confirm with your tax advisor before acting.
The chart reconstructs historical portfolio performance assuming you held the current weights from the start of the selected period. Each day, it computes the weighted return of all positions and chains them into a cumulative index starting at 100.
This is a static-weight backtest, not rebalanced. The actual past performance of a rebalanced portfolio would differ — use the Backtest module for that.
Auto-refresh: while the panel is open, the chart updates automatically every 5 minutes with the latest prices. Drag to scroll history, scroll to zoom.
X-axis: Underlying price (±30% from current spot)
Y-axis: Days to expiry (1 day → current DTE)
Z-axis / Color: The selected value (P&L, Delta, or Gamma)
Two reference planes are drawn: ■ Gold = strike price, ■ Green = current spot. You can rotate and zoom the 3D chart freely.
Displays current Greeks computed at the live spot price and current DTE using Black-Scholes. If a market price was stored with the option, Implied Volatility is back-solved via Brent's method before computing all Greeks.
Each news item is matched against a library of 23+ geographic regex patterns (company names, index names, central banks, currencies) and a map of 35+ regional ETF tickers. Matches place a glowing dot on the corresponding country or region.
Stories that don't match any pattern are distributed across 24 known financial city locations worldwide — so the globe stays populated even with mixed news feeds.