This is a personal project I built to help people make better investment decisions. The idea was to combine AI with portfolio analysis tools, so users can build portfolios, check risk, and get suggestions without needing a finance background.
I wanted to make something that works for both beginners and more experienced investors. Beginners can follow a guided survey to set up their portfolio, while advanced users can manually pick stocks and ETFs.
Portfolio Construction
Users can build a portfolio in a few ways:
- Templates: I included some well-known portfolio templates like “Golden Butterfly” and “Bogleheads Three-Fund” so users can start from a proven allocation and adjust from there.
- Goal survey: There’s an interactive survey that asks about risk tolerance, expected returns, and investment horizon. Based on the answers, the system recommends an asset allocation.
- Manual entry: For users who already know what they want, they can just add specific stocks or ETFs directly.
Risk Assessment
The risk dashboard shows a few things that I think are useful for understanding portfolio risk:
- VaR (Value at Risk): Shows the 95% confidence interval on potential losses. This gives a rough idea of how much the portfolio could lose in a bad scenario.
- Drawdown analysis: Shows the worst drops and how long it took to recover, compared against benchmarks like SPY.
- Stress testing: Users can see how the portfolio would perform under different economic conditions.
Market Insights
- Correlation analysis: A color-coded matrix that shows how different assets in the portfolio move together. This helps users see if they’re actually diversified or just holding assets that go up and down at the same time.
- AI chatbot: I added a chatbot that can answer questions about the portfolio, like explaining what a drawdown period means or why a certain asset is volatile.
- Global data: The platform pulls real-time stock and ETF data so users are not looking at stale numbers.
Performance Dashboard
- Key ratios: CAGR, total return, volatility, Sharpe ratio. These are the standard metrics that most investors look at.
- Growth chart: A chart comparing portfolio growth against a benchmark over time, so users can see if they’re doing better or worse.
- Risk-adjusted returns: Side-by-side comparison showing where the portfolio is strong and where it could improve.
AI Insights
- Summary section: The AI reads the portfolio data and highlights areas that might need attention, like underperforming holdings or sectors that are too concentrated.
- Optimization suggestions: Things like rebalancing triggers, tax-loss harvesting opportunities, or weight adjustments. These are suggestions, not automatic actions.
Settings and Accessibility
- Guided flow for beginners: The step-by-step survey makes it less overwhelming for people who don’t have investing experience.
- Advanced options: Manual stock entry, transaction cost simulation, and more control for experienced users.
- Mobile responsive: The design works on phones and tablets.
Pre-Built Portfolio Templates
There are 10+ pre-configured portfolios like “Total Stock Market” and “Simple Money” that users can pick and modify. I added these because most beginners don’t know where to start, and having a template to work from is easier than building from scratch.
What I’d Work on Next
- Better prediction models: Using multi-factor forecasting to give longer-term views.
- Sentiment analysis: Pulling in social media data for short-term trading signals.
- Mobile app: A native app so the experience is smoother on phones.
The platform is live at stock.twinight.co if you want to try it.