01 Try a piece of it
Will this plan get there? Run it 1,500 times and see.
This is a small working version of the simulation at the center of ARGUS. Set a starting amount, a monthly contribution, a stock and bond mix, and a goal. It simulates 1,500 possible futures and shows the range of outcomes.
Illustrative assumptions: stocks average 7% a year with 16% volatility, bonds 3.5% with 6% volatility, rebalanced monthly, before fees, taxes, and inflation. This is a teaching sample and not a forecast or investment advice. The full tool works from a person's actual holdings.
Want the whole household, not just the portfolio? The financial plan on LDB Finance adds pay, taxes, debts, goals and Social Security, and shows what it would take to get to an 85% chance.
02 What the full tool does
Three questions, answered from the actual holdings.
| Question | How ARGUS answers it |
|---|---|
| What do I own? | Holdings are entered lot by lot, grouped by account, so cost basis and gains are tracked at the level taxes are calculated |
| What is really inside it? | An ETF look-through opens each fund and shows the underlying companies, which reveals hidden overlap between funds |
| How might it behave? | A Monte Carlo simulation runs 1,500 paths and reports the range of outcomes and the chance of reaching a goal |
| How would it have behaved? | A historical backtest replays the allocation through market returns from 1928 to 2025, using the long-run return data published by NYU professor Aswath Damodaran |
| How did it hold up in a crisis? | A black swan chart marks major market crises on the timeline so the drawdowns can be seen in context |
| Is the mix efficient? | An efficient frontier shows whether the same return was available with less risk, and a CAPM beta estimate shows sensitivity to the market |
| Can I take it with me? | Full Excel import and export, and print-ready reports formatted like a professional statement |
It also includes an assistant built on the Claude API that can refresh prices and answer questions about the portfolio in plain language, and a light and dark theme.
03 How it is built
Calculations on your device, by design.
ARGUS is a React application whose analysis runs entirely in the browser: the simulations, the backtest and every chart are computed on the user's own device, and holdings are never stored on a server. Two optional features go online, and only when the user presses them: the price refresh, which looks up current prices, and the assistant, which sends the question to an AI model.
Why I Built It
The planning software used in wealth management is powerful and expensive, and a student or a first-time investor never sees it. I wanted to understand how those tools reach their answers by building the calculations myself, and to make something useful that anyone could open for free.
What it taught me. A simulation is only as honest as its assumptions. Most of the work was in deciding what to assume and showing those assumptions to the user, not in the math.