About ArthaPilot
Video walkthrough
Tour household setup, a backtest, saved analyses, and monitors.
- Start on the Household page. It holds accounts and tax details that tools reuse.
- This sample household has three accounts with different tax treatments. The overview shows their combined value and drift from the target allocation at the displayed pricing date.
- Your work lives under Household, Workspace, and Monitors. The tools below are grouped by the job they do.
- This tax profile supplies filing status, work state, and income. Its bracket assumptions carry into tax-aware tools.
- Each account keeps its type, owner, and tax lots. The Roth IRA belongs to Jordan Example, a second household member. Tools reuse these saved records.
- Findings highlights which household records or questions need attention.
- This example flags a stale snapshot, tax opportunities, and target drift.
- Each finding links to the tool that can investigate it.
- These tools open with the saved household, so the same accounts and tax context carry into the analysis.
- Portfolio Backtest replays an investment mix through market history. A prepared example compares two portfolios through the financial crisis. The workflow is inputs, run, results, and save.
- Both portfolios share the same dates and rules. Only their allocations differ. One holds sixty percent stocks and forty percent bonds. The other holds forty five percent stocks, twenty five percent bonds, and thirty percent gold.
- Run the comparison across the trading days from 2007 through 2012.
- Both start from ten thousand dollars. The gold mix ends near sixteen thousand three hundred dollars. The stock-and-bond mix ends near twelve thousand nine hundred dollars. Compare the loss along the way before reading that gap.
- Drawdown measures the drop from a previous high. The stock-and-bond mix was down thirty five and a half percent at its worst. The gold mix fell twenty eight point three percent. Its higher ending balance came with a shallower decline in this window.
- Total return includes distributions, with monthly rebalancing and no inflation adjustment. These pretax results use the full historical window, with no holdout.
- Transaction costs and market impact are excluded. Those omissions matter when comparing portfolios that trade differently. The historical window also determines which market events they experience.
- Save this comparison with a descriptive name. Workspace keeps its inputs and its results in one place.
- Share or export from the result bar. A shared analysis replays the saved calculation. Export downloads the series and result figures as a CSV file.
- The tool's tutorial walks through a complete worked example.
- Take a tour highlights the main controls in the tool.
- Chapters and the transcript let you return to a specific explanation.
- The Saved analyses page brings the results from all your tools together.
- This comparison now sits beside three planning runs in the example Workspace. Each can reopen in its tool, and related results can be compared.
- The Library keeps reusable portfolios, strategies, tickers, and households. Their earlier versions remain available. Each saved analysis retains the version it used.
- Monitors follow the household between your visits.
- This household has three: allocation drift, tax-loss opportunities, and wash-sale windows. Each monitor shows what it watches and its notification settings.
- The Strategy Leaderboard provides public research examples. Curated strategies run through the same historical engine. A shared window makes their return and risk figures comparable.
- A strategy row opens its rules and current signal, with an option to copy it into your Library. The signals page collects current states from published strategies.
- Docs collects each tool's reference, methodology, and tutorial for later use.
- Start with a prepared example in Portfolio Backtest. Save the result, change one allocation, and compare the new result.
The story
I built ArthaPilot as a personal research tool for backtesting strategy blends with a tax-aware engine. Friends and family asked to use it, so I sent packaged versions they could run locally. Maintaining those versions became difficult, which led me to build a shared platform.
ArthaPilot is an educational modeling tool, not a trading platform. It does not provide investment, tax, or legal advice. It helps users examine portfolio construction, risk, and long-term compounding with historical data and stated assumptions.
I have worked as a programmer for more than 13 years. ArthaPilot is a long-term passion project.
Research and corrections
ArthaPilot Research is the publication name for studies maintained with this project. The project maintainer is responsible for the software, study methods, published results, and corrections.
Studies state their inputs, data period, sources, and limitations. The verification suite describes the independent numerical checks used by the engines. Reproducible checks support a model; they do not establish that its assumptions fit a particular household.
Report a suspected error through Contact. Include the page, inputs, and expected result. Published corrections belong in the changelog; a study's updated date records a revision to that study.
My costs
Running ArthaPilot requires cloud infrastructure, simulation and backtest compute, market data subscriptions, and development time.
All revenue supports operating costs and continued development.
Contact
Email with feedback, feature requests, or questions about the tool.