Institutional-grade tools. Without the institutional price tag.
Bloomberg charges $24,000 a year. StocksBro v2 gives you convex portfolio optimization,
AI-powered equity research, Monte Carlo simulations, and risk analytics — completely free.
Built for the 58 million self-directed investors who deserve better than a pie chart and a prayer.
The gap between Wall Street and everyone else is absurd.
Professional investors build portfolios with quantitative optimization, stress-testing, and AI-driven research.
Everyone else gets brokerage pie charts and gut feelings. That disparity is not a feature — it's a failure of the market.
Bloomberg Terminal
$24,000 / year
Portfolio construction, risk analytics, real-time data, equity research — the gold standard. Reserved for institutions.
FactSet / Refinitiv
$12,000 – $22,000 / year
Quantitative analytics and multi-asset research. Priced exclusively for hedge funds and RIAs.
Portfolio Visualizer
$30 / month
Backtesting only. No optimization engine. No AI research. No risk decomposition. Interface from 2012.
Your Brokerage App
Free (in exchange for your order flow)
Designed for execution, not analysis. A pie chart, some news headlines, and a "buy" button. That's it.
No product currently combines MPT portfolio optimization, AI-generated equity research, and full risk analytics at a consumer price point.
That's the gap StocksBro v2 fills — for 25 to 35 million active self-directed investors in the US alone.
Capabilities
Six analytical panels. One terminal.
Everything an institutional portfolio manager uses — shrunk down to a browser tab. Convex optimization with CVXPY, historical backtesting, Monte Carlo forward projection, and AI-powered research — all solving in seconds.
fn+F1
Portfolio Optimizer
Exact convex optimization via CVXPY/CLARABEL. Not a toy approximation — the same math hedge funds use to allocate billions.
Max Sharpe, Min Volatility, Quadratic Utility strategies
Mean Historical, CAPM, and Black-Litterman return models
Interactive efficient frontier with Capital Market Line
fn+F2
Backtesting Engine
See how your optimized allocation would have actually performed. No cherry-picking. No survivorship bias. Just the equity curve, naked.
Benchmark overlays (SPY, QQQ, IWM, BND)
Drawdown analysis with max-drawdown markers
Monthly OHLC candlestick view of portfolio performance
Monte Carlo simulation with P10 to P90 fan charts
fn+F3
P&L and Rebalancing
Track holdings, identify drift, and generate exact trade instructions to bring your portfolio back to target weights.
Per-position P&L with weighted return contribution
Individual asset analytics (Sharpe, Sortino, Max DD)
Drift visualisation with BUY/SELL/HOLD signals
Dollar-denominated trade log with cost basis tracking
fn+F4
Risk Intelligence
Understand where your risk actually lives. Decompose it, stress-test it, and see what a 2008-style crash would do to your book.
Value at Risk and CVaR tail risk metrics
Marginal risk attribution per position
Correlation heatmap with period switching
Historical stress scenarios (GFC, COVID, dot-com)
fn+F5
Market Guide
Cut through the jargon. An interactive guide that explains the concepts behind every metric, strategy, and risk measure — so you make informed decisions, not guesses.
Plain-language explainers for Sharpe, VaR, and Black-Litterman
Strategy selection guidance based on your goals
Risk tolerance framework with real-world examples
Glossary of institutional finance terminology
fn+F6
AI Research
Type a ticker, get an institutional-grade research report. A multi-perspective AI analyst team debates bull, bear, and base cases — then a committee delivers a verdict.
Three-analyst adversarial debate with data citations
Chief strategist critique of all positions
Committee verdict: Attractive, Neutral, or Avoid
Real-time fundamentals from Finnhub + FRED macro context
AI-Powered Analysis
Your personal investment committee. On demand.
Type a ticker. Get an institutional-style research report powered by a multi-perspective AI analyst team — bull case, bear case, chief strategist critique, and a committee verdict. Combining real-time data from Finnhub, macro context from FRED, and adversarial analysis from Claude. In seconds, not weeks.
Bull vs Bear Debate
Three AI analysts argue the bull case, bear case, and base case for every security — citing specific fundamentals, not vibes. Then a chief strategist tears apart the weaknesses in all three.
Committee Verdict
An investment committee chair weighs all arguments and delivers a final verdict: Attractive, Neutral, or Avoid — with confidence level, key risk, and 12-month directional view.
ETF Overlap Detection
Holding NVDA + SPY + QQQ? The optimizer detects overlapping exposure across ETFs and individual stocks, flags concentration risk, and factors it into the AI's portfolio allocation.
Macro-Aware Regime Detection
Seven FRED indicators — fed funds rate, yield curve, VIX, credit spreads, inflation — fed into every analysis. The AI classifies the current regime and adjusts its views accordingly.
6
Analytical Panels
3
Optimization Strategies
5,000
Monte Carlo Paths
$0
Price
How It Works
Three steps. Zero guesswork.
The terminal is designed to feel immediate. Add tickers, choose your parameters, and let the solver do what solvers do best — find the optimal answer to a well-defined problem.
1
Add your tickers
Search any stock, ETF, or cryptocurrency by symbol or name. The autocomplete covers every instrument Yahoo Finance tracks — US equities, ASX-listed stocks, crypto pairs, global ETFs. Build your universe in seconds.
2
Configure your strategy
Pick an optimization strategy and returns model. Set weight constraints, lookback period, base currency, and risk-free rate. Using Black-Litterman? Enter your forward views with confidence levels and let the Bayesian engine blend them with market priors.
3
Run and explore
The solver returns optimal weights, efficient frontier, backtest, risk decomposition, Monte Carlo projections, rebalancing trades, and stress scenarios — all in one pass. Switch between panels to explore every angle of your portfolio.
Important Legal Disclaimer
StocksBro v2 is an educational and informational tool only. Nothing on this platform constitutes financial advice, a recommendation, or a solicitation to buy, sell, or hold any security, cryptocurrency, or financial instrument.
All outputs — including optimized portfolio weights, backtests, Monte Carlo simulations, AI-generated research reports, and risk metrics — are derived from historical data and mathematical models. Past performance does not guarantee future results. Markets are inherently unpredictable, models have structural limitations, and all investing carries risk of partial or total capital loss.
The AI research reports are generated by a large language model and may contain inaccuracies, outdated information, or errors of interpretation. They should never be treated as a substitute for professional due diligence or licensed financial advice.
You are solely responsible for your own investment decisions. The creator of StocksBro accepts no liability whatsoever for any losses, damages, or consequences — direct or indirect — arising from the use of this tool or reliance on its outputs. Always consult a qualified financial advisor before acting on any information presented here.
By using StocksBro v2, you acknowledge that you understand these limitations and agree to use the platform entirely at your own risk.
●
●
◆ OPTIMISING PORTFOLIO
→ Fetching price history from Yahoo Finance
→ Estimating covariance via Ledoit-Wolf shrinkage
→ Solving convex programme with CVXPY / CLARABEL
First request may take 30–60 s while the API server warms up on Render free tier.
fn+F1
Portfolio Optimizer
Institutional-grade portfolio construction powered by convex optimization and Ledoit-Wolf covariance shrinkage. The same mathematical framework used by hedge funds and asset managers — with an AI investment committee built in.
Expected Return
18.4%
Volatility
14.2%
Sharpe Ratio
1.47
Max Drawdown
-11.3%
MAX SHARPEMEAN HISTORICAL5YUSDRFR 4.5%
EFFICIENT FRONTIERSAMPLE DATA
Every dot is a possible portfolio built from your assets. The curve traces the highest return achievable at each level of risk. Your optimised portfolio sits at the tangency point on the Capital Market Line.
OPTIMAL ALLOCATIONSAMPLE DATA
Recommended weight for each position as solved by the optimizer. Longer bars indicate a larger allocation to that asset.
AAPL
28.1%
MSFT
22.4%
GOOGL
18.0%
AMZN
15.2%
NVDA
10.1%
META
6.2%
POSITION BREAKDOWNSAMPLE DATA
Per-ticker weight and its proportional contribution to the portfolio's expected annual return.
Ticker
Weight %
$ Allocation
Wtd. Return %
AAPL
28.1%
$28,100
5.16%
MSFT
22.4%
$22,400
4.48%
GOOGL
18.0%
$18,000
3.24%
AMZN
15.2%
$15,200
2.89%
NVDA
10.1%
$10,100
1.82%
META
6.2%
$6,200
0.81%
3 Optimization Strategies
Max Sharpe Ratio (highest risk-adjusted return), Min Volatility (smoothest ride), or Max Quadratic Utility (custom risk aversion). Each solves a different convex problem via CVXPY/CLARABEL.
4 Return Models
Mean Historical, CAPM (market-adjusted), Black-Litterman (your forward views + confidence), or AI Analyst — where Claude acts as an investment committee generating views from live fundamentals and macro data.
AI Investment Committee
Select the AI Analyst model and Claude generates bull/bear/base cases per ticker using live Finnhub fundamentals and FRED macro data. Views feed directly into Black-Litterman, with regime detection and ETF overlap alerts.
Current Holdings Input
Enter your existing dollar positions per ticker. The optimizer calculates exact BUY/SELL trade amounts to move from where you are to the target allocation — not just target weights, but actionable trades.
Add 2+ tickers in the sidebar, choose your strategy and return model, then hit RUN to optimize.
EXPECTED ANNUAL RETURN
ANNUAL VOLATILITY
SHARPE RATIO
MAX DRAWDOWN
EFFICIENT FRONTIERiEvery dot is a possible portfolio built from your assets. The curve traces the highest return achievable at each level of risk. Your optimised portfolio sits at the tangency point on the Capital Market Line — the highest achievable Sharpe Ratio.
Every dot is a possible portfolio built from your assets. The curve traces the highest return achievable at each level of risk. Your optimised portfolio sits at the tangency point on the Capital Market Line — the highest achievable Sharpe Ratio.
OPTIMAL ALLOCATIONiRecommended weight for each position as solved by the optimizer. Longer bars indicate a larger allocation to that asset.
Recommended weight for each position as solved by the optimizer. Longer bars indicate a larger allocation to that asset.
WHY THIS ALLOCATION
EXPOSURE OVERLAP DETECTED
INVESTMENT COMMITTEE
AI VIEWS PER ASSET
Ticker
Exp. Return
Confidence
Reasoning
POSITION BREAKDOWNiPer-ticker weight and its proportional contribution to the portfolio's expected annual return. Wtd. Return = weight × asset expected return.
Per-ticker weight and its proportional contribution to the portfolio's expected annual return. Wtd. Return = weight × asset expected return.
Ticker
Weight %
$ Allocation
Wtd. Return %
●
●
◆ OPTIMISING PORTFOLIO
→ Fetching price history from Yahoo Finance
→ Estimating covariance via Ledoit-Wolf shrinkage
→ Solving convex programme with CVXPY / CLARABEL
First request may take 30–60 s while the API server warms up on Render free tier.
fn+F2
Backtesting Engine
See how your optimized portfolio would have actually performed over real market history. No cherry-picking, no survivorship bias — raw equity curves against institutional benchmarks, with Monte Carlo forward projection.
Total Return
+127%
Benchmark (SPY)
+89%
Alpha
+38%
Max Drawdown
-11.3%
PORTFOLIO PERFORMANCE — INDEXED TO 100SAMPLE DATA
Portfolio cumulative return plotted against SPY. Switch benchmarks live to compare across bull runs, corrections, and crashes.
PORTFOLIO DRAWDOWN FROM PEAKSAMPLE DATA
How far the portfolio fell from its rolling all-time high at each point in time. Deeper troughs represent larger historical peak-to-trough losses.
MONTHLY OHLC — PORTFOLIO VALUESAMPLE DATA
Candlestick chart of the backtested equity curve's monthly open/high/low/close. The benchmark overlay lets you compare your portfolio's range against a market index.
Set slippage in basis points. A net-of-cost equity curve appears alongside the gross curve, showing exactly how rebalancing friction erodes returns over time.
Export Everything
Download any chart as PNG or any dataset as CSV. Equity curve, drawdown, OHLC, and Monte Carlo data — all exportable for your own analysis or reporting.
Run an optimization first — backtest results populate automatically from your optimized allocation.
TOTAL RETURN
BENCHMARK (SPY)
ALPHA vs SPY
MAX DRAWDOWN
PORTFOLIO PERFORMANCE - INDEXED TO 100
BENCHMARK
PORTFOLIO DRAWDOWN FROM PEAKiHow far the portfolio fell from its rolling all-time high at each point in time. Deeper troughs represent larger historical peak-to-trough losses.
How far the portfolio fell from its rolling all-time high at each point in time. Deeper troughs represent larger historical peak-to-trough losses.
MONTHLY OHLC — PORTFOLIO VALUEiCandlestick chart of the backtested equity curve's monthly open/high/low/close. The benchmark overlay lets you compare your portfolio's range against a market index.
Candlestick chart of the backtested equity curve's monthly open/high/low/close. The benchmark overlay lets you compare your portfolio's range against a market index.
HORIZON
SIMULATIONS
MONTE CARLO FORWARD SIMULATIONiBootstrap simulation projecting future return paths from historical daily returns. P10 = pessimistic (bottom 10%), P50 = median, P90 = optimistic (top 10%). Select a horizon and simulation count above to explore scenarios.
Bootstrap simulation projecting future return paths from historical daily returns. P10 = pessimistic (bottom 10% of paths), P50 = median outcome, P90 = optimistic (top 10%). Select a horizon and simulation count above to explore different scenarios.
●
●
◆ OPTIMISING PORTFOLIO
→ Fetching price history from Yahoo Finance
→ Estimating covariance via Ledoit-Wolf shrinkage
→ Solving convex programme with CVXPY / CLARABEL
First request may take 30–60 s while the API server warms up on Render free tier.
fn+F3
P&L & Rebalancing
Four sub-tabs covering holdings, per-asset analytics, drift detection, and a manual trade log. The operational layer that turns an optimized portfolio into actionable trades with dollar-denominated instructions.
Total Portfolio Value
$127,400
Total P&L
+$27,400
Total Return
+27.4%
HOLDINGSANALYTICSREBALANCINGTRADES
HOLDINGSSAMPLE DATA
Per-position P&L table showing model value, dollar gain/loss, and percentage return for each ticker in your optimised allocation.
Ticker
Weight %
Model Value
P&L ($)
P&L (%)
Return
AAPL
28.1%
$35,797
+$7,697
+27.4%
+27.4%
MSFT
22.4%
$28,538
+$6,138
+27.4%
+32.1%
GOOGL
18.0%
$22,932
+$4,932
+27.4%
+18.6%
AMZN
15.2%
$19,365
-$635
-3.2%
-3.2%
NVDA
10.1%
$12,867
+$6,867
+114.5%
+114.5%
META
6.2%
$7,901
+$2,401
+43.6%
+43.6%
TOTAL
100%
$127,400
+$27,400
+27.4%
SHARPE RATIO COMPARISONSAMPLE DATA
Risk-adjusted return ranked across positions. Higher Sharpe = better return per unit of risk. The green dashed line marks Sharpe = 1.0.
REBALANCING DRIFTSAMPLE DATA
Bidirectional drift bars show how far each position has drifted from its target weight. BUY/SELL signals trigger when drift exceeds your configured threshold.
AAPL
+2.4%
MSFT
-1.6%
GOOGL
-3.1%
AMZN
+1.0%
NVDA
+3.6%
META
-2.1%
Configurable Drift Threshold
Set your rebalance trigger from 0% to 10% in the sidebar. Only positions that drift beyond your threshold generate a BUY or SELL signal.
Dollar-Mode Trades
When current holdings are entered, rebalancing outputs switch from percentages to exact dollar amounts — TRADE $2,340 of AAPL, not just "+2.3%".
Run an optimization first — P&L and rebalancing data populate from your optimized weights.
TOTAL PORTFOLIO VALUE
TOTAL P&L
TOTAL RETURN
HOLDINGSiPer-position P&L table showing model value, dollar gain/loss, and percentage return for each ticker in your optimised allocation.
Per-position P&L table showing model value, dollar gain/loss, and percentage return for each ticker in your optimised allocation.
Ticker
Weight %
Model Value
P&L ($)
P&L (%)
Return
INDIVIDUAL ASSET ANALYSISiPerformance of each asset indexed to 100 at the start of the lookback period. Use this to compare relative momentum and trend across your holdings.
Performance of each asset indexed to 100 at the start of the lookback period. Use this to compare relative momentum and trend across your holdings.
SHARPE RATIO COMPARISONiRisk-adjusted return (excess return ÷ volatility) ranked across positions. Higher Sharpe = better return per unit of risk. The green dashed line marks Sharpe = 1 (a common benchmark for a quality position).
Risk-adjusted return (excess return ÷ volatility) ranked across positions. Higher Sharpe = better return per unit of risk. The green dashed line marks Sharpe = 1 (a common benchmark for a quality position).
REBALANCING DRIFT — 1-YEAR LOOKBACKiCompares current holding weights to your optimal target weights over the past year. Bars drifting away from centre signal a position that may need trimming or topping up.
Compares current holding weights to your optimal target weights over the past year. Bars drifting away from centre signal a position that may need trimming or topping up.
REBALANCING ACTIONSiBUY / SELL / HOLD recommendations derived from the drift analysis. Dollar amounts indicate how much to trade per position to return to your target allocation.
BUY / SELL / HOLD recommendations derived from the drift analysis. Dollar amounts indicate how much to trade per position to return to your target allocation.
TRADE LOGiEnter your actual buy trades to track real P&L against current market prices. Click a row to edit — date format YYYY-MM-DD, quantity in shares, price in your base currency.
Enter your actual buy trades to track real P&L against current market prices. Click a row to edit — date format YYYY-MM-DD, quantity in shares, price in your base currency.
No trades recorded. Click a row below to add your first trade.
DATE
TICKER
QTY
BUY PRICE
LIVE P&L SUMMARYiAggregate mark-to-market P&L across all trades entered in the Trade Log above, using the latest fetched prices from the API.
Aggregate mark-to-market P&L across all trades entered in the Trade Log above, using the latest fetched prices from the API.
TICKER
CURRENT PRICE
COST BASIS
CURRENT VALUE
P&L ($)
P&L (%)
●
●
◆ OPTIMISING PORTFOLIO
→ Fetching price history from Yahoo Finance
→ Estimating covariance via Ledoit-Wolf shrinkage
→ Solving convex programme with CVXPY / CLARABEL
First request may take 30–60 s while the API server warms up on Render free tier.
fn+F4
Risk Intelligence
Understand where your risk actually lives. Decompose it by position, stress-test against historical crashes, quantify tail risk with VaR/CVaR, and visualize cross-asset correlations — the same analytics institutional risk desks use daily.
Portfolio Volatility
14.2%
Diversification Ratio
1.23
Largest Single-Asset Risk
NVDA (37%)
STRESS TEST SCENARIOSSAMPLE DATA
Estimated portfolio impact during historical market stress events using portfolio beta vs S&P 500. Results are indicative — factor exposures may vary.
GFC 2008
-38.2%
S&P fell -56.8%
COVID 2020
-22.4%
S&P fell -33.9%
Dot-Com 2000
-41.7%
S&P fell -49.1%
Rate Hikes 2022
-16.8%
S&P fell -25.4%
RISK ATTRIBUTION — MARGINAL CONTRIBUTION TO VOLATILITYSAMPLE DATA
Each asset's percentage share of total portfolio variance. High contributors carry concentrated, undiversified risk.
NVDA
37%
AAPL
22%
MSFT
18%
GOOGL
12%
AMZN
7%
META
4%
CORRELATION MATRIXSAMPLE DATA
Pairwise return correlations between your assets. Values near 0 indicate low correlation (good diversification). Values above 0.8 signal concentrated risk.
FULL3Y1Y6M
AAPL
MSFT
GOOGL
AMZN
NVDA
META
AAPL
1.00
0.72
0.65
0.58
0.61
0.54
MSFT
0.72
1.00
0.70
0.63
0.67
0.59
GOOGL
0.65
0.70
1.00
0.68
0.55
0.62
AMZN
0.58
0.63
0.68
1.00
0.52
0.57
NVDA
0.61
0.67
0.55
0.52
1.00
0.48
META
0.54
0.59
0.62
0.57
0.48
1.00
VaR & CVaR
Parametric Value at Risk at 95% and 99% confidence, plus Conditional VaR (expected shortfall). Know your worst-case annual loss at institutional confidence levels.
Period Switching
Switch correlation windows between 6M, 1Y, 3Y, or Full history to spot regime-dependent breakdowns that only appear in stressed markets.
Run an optimization first — risk analytics populate automatically from your optimized portfolio.
PORTFOLIO VOLATILITY (ANNUAL)
DIVERSIFICATION RATIO
LARGEST SINGLE-ASSET RISK
STRESS TEST SCENARIOSiEstimated portfolio impact during historical market stress events using portfolio beta vs S&P 500. Results are indicative — factor exposures may vary.
Estimated portfolio impact during historical market stress events using portfolio beta vs S&P 500. Results are indicative — factor exposures may vary.
RISK ATTRIBUTION - MARGINAL CONTRIBUTION TO VOLATILITYiEach asset's percentage share of total portfolio variance. High contributors carry concentrated, undiversified risk — consider trimming positions that dominate this chart.
Each asset's percentage share of total portfolio variance. High contributors carry concentrated, undiversified risk — consider trimming positions that dominate this chart.
CORRELATION MATRIXiPairwise return correlations between your assets. Values near 0 indicate low correlation (good diversification). Values above 0.8 signal concentrated risk — those assets tend to move together.
Pairwise return correlations between your assets. Values near 0 indicate low correlation (good diversification). Values above 0.8 signal concentrated risk — those assets tend to move together.
WHAT IS THIS TOOL?
StocksBro v2 is a portfolio calculator. Give it a list of stocks or funds and it works out the best possible way to split your money across those assets — based on their historical prices and a branch of finance called Modern Portfolio Theory (MPT).
Think of it like making a smoothie: you don't throw in equal amounts of everything. Some ingredients complement each other; others clash. This tool does the same for investments — it finds the blend that gives you the most return for the least risk, or whichever goal you choose.
What you need: a list of ticker symbols (e.g., AAPL = Apple, SPY = S&P 500 fund). Choose a strategy, press ▶ RUN OPTIMIZATION, and the tool fetches real price data, runs the maths, and shows you the optimal allocation across all four panels.
fn+F2BACKTESTHistorical equity curve + drawdown + Monte Carlo
fn+F3P&LHoldings + rebalancing + trade log
fn+F4RISKRisk attribution + correlation matrix
fn+F5GUIDEThis reference panel
fn+F6RESEARCHAI equity research — bull/bear debate, committee verdict
HOW TO USE
01
CONFIGURE
Open the sidebar (▶ toggle on the left edge). Add tickers manually, use the Ticker Lookup search, or pick a Quick Load preset (MAG 7, S&P TOP 10, FAANG+, DIVIDEND, ETF MIX, CRYPTO). Set your strategy, returns model, lookback, risk-free rate, weight constraints, portfolio size, and transaction cost estimate.
02
OPTIMIZE
Press ▶ RUN OPTIMIZATION. The engine fetches adjusted-close price history, estimates the covariance matrix via Ledoit-Wolf shrinkage, and solves the convex optimization problem. If using AI Analyst mode, the engine also detects ETF overlap, gathers live fundamentals and macro data, runs an AI investment committee (bull/bear/critique), and feeds the resulting views into a Black-Litterman model.
03
ANALYZE
Use the tab bar (or keyboard shortcuts) to explore your results — optimal weights and AI explanation in OPTIMIZER, historical performance and Monte Carlo in BACKTEST, position P&L and rebalancing in P&L, risk metrics and stress tests in RISK, and individual stock research in RESEARCH. Use the ↓ export buttons to download any section as CSV.
▶ PLATFORM CAPABILITIES
OPTIMIZER
PORTFOLIO OPTIMIZATION
Three MPT strategies — Max Sharpe Ratio, Min Volatility, and Max Quadratic Utility — solved with CVXPY/CLARABEL. Configurable per-position weight bounds and Ledoit-Wolf covariance estimation.
OPTIMIZER
EFFICIENT FRONTIER
Visualize the full risk/return opportunity set. Your allocation is plotted on the curve with the Capital Market Line overlay showing the optimal Sharpe tangency portfolio.
OPTIMIZER
AI ANALYST (CLAUDE)
AI-powered Investment Committee. Claude analyses live Finnhub fundamentals and FRED macro data per ticker, generates forward-looking return views with confidence scores, and feeds them into Black-Litterman. Includes regime assessment, risk flags, and a natural-language explanation of why the portfolio is allocated the way it is.
BACKTEST
HISTORICAL BACKTESTING
Equity curve indexed to 100 with selectable benchmark overlays (SPY, QQQ, IWM, BND). Cumulative return, active alpha, max drawdown, and monthly OHLC candlestick chart. Net-of-cost equity curve when transaction costs are configured.
BACKTEST
MONTE CARLO SIMULATION
Bootstrap up to 5,000 simulated return paths from historical daily returns. View P10 / P25 / P50 / P75 / P90 outcome bands across 1–10 year horizons.
RISK
RISK ANALYTICS
VaR (95%), CVaR / Expected Shortfall, marginal contribution to volatility per asset, diversification ratio, and correlation heatmap with period selector (6M, 1Y, 3Y, Full).
RISK
STRESS TESTING
Estimated portfolio impact during GFC 2008, COVID 2020, and Rate Rise 2022 stress events. Scaled by portfolio beta vs S&P 500.
P&L
REBALANCING GUIDANCE
Drift analysis compares current vs target weights. BUY / SELL / HOLD recommendations with per-position dollar sizing. Enter current holdings for accurate trade sizing. Full export to CSV.
RESEARCH
AI EQUITY RESEARCH
Enter any US ticker to generate a multi-perspective AI research report — bull case, bear case, base case, and a chief strategist critique. An investment committee then delivers a final verdict (ATTRACTIVE / NEUTRAL / AVOID) with confidence level and key risk.
A stock (also called a share or equity) is a tiny slice of ownership in a company. When you buy one share of Apple (AAPL), you own a microscopic piece of Apple Inc. If Apple becomes more valuable, your share is worth more. If it does poorly, it's worth less.
What is a ticker?
Tickers are the short codes used to identify each stock on an exchange:
AAPL = Apple
MSFT = Microsoft
GOOGL = Alphabet (Google's parent company)
SPY = ETF tracking the S&P 500 (top 500 US companies)
BTC-USD = Bitcoin priced in US dollars
What is an ETF?
An ETF (Exchange-Traded Fund) is a basket of many stocks bundled together. When you buy SPY, you're effectively buying tiny pieces of all 500 companies in the S&P 500. ETFs let you diversify instantly without picking individual companies.
What is a portfolio?
A portfolio is your total collection of investments. Instead of putting all your money in one stock, you spread it across several — this is diversification. The core idea: if one investment tanks, the others can cushion the blow. This tool asks: given these assets, what's the smartest way to split my money?
SUPPORTED EXCHANGES
United States (default)
US stocks and ETFs use their ticker as-is — no suffix needed: AAPL, MSFT, SPY, BTC-USD.
ASX — Australian Securities Exchange
ASX-listed stocks require the .AX suffix appended to the ticker:
Company
Enter as
Commonwealth Bank
CBA.AX
BHP Group
BHP.AX
NAB
NAB.AX
CSL Limited
CSL.AX
Wesfarmers
WES.AX
Other major exchanges
Exchange
Suffix
Example
London Stock Exchange
.L
HSBA.L
Toronto Stock Exchange
.TO
RY.TO
Frankfurt
.DE
SAP.DE
Tokyo
.T
7203.T (Toyota)
You can mix US, ASX, and other international stocks in one portfolio. However, be aware of the currency limitation below — mixing exchanges without setting a base currency can distort the optimization.
MULTI-CURRENCY
All prices fetched from Yahoo Finance are in their native currency (usually USD). When you select a Base Currency other than USD in the sidebar, the tool:
Fetches the daily USD→[currency] exchange rate (e.g., USDEUR=X)
Multiplies all asset prices by that rate
Runs the entire optimization in your chosen currency
This removes FX distortion from global portfolios — a 10% stock gain means less if USD weakened 5% against your home currency during the same period.
Example — Australian investor: if you hold CBA.AX (priced in AUD) alongside AAPL (priced in USD), selecting AUD as base currency ensures both are evaluated in AUD so the optimizer correctly accounts for the FX impact on your US holdings.
REBALANCING & DRIFT
What is portfolio drift?
When the optimizer sets your target weights — say 40% AAPL, 35% MSFT, 25% NVDA — those weights are correct on day one. But as prices move, positions drift. If NVDA rallies 20%, it becomes a larger share of your portfolio than intended. Left unchecked, drift gradually turns your carefully optimised allocation into something unrecognisable.
What is rebalancing?
Rebalancing is the act of selling positions that have grown above their target weight and buying positions that have fallen below it — restoring the portfolio to its intended allocation. This is how you maintain the risk profile the optimizer calculated.
What is the Rebalance Threshold?
The Rebalance Threshold (set in the sidebar) defines the minimum drift required before the tool recommends action. For each position, the tool compares its current weight to its target weight. If the difference exceeds the threshold, the position is flagged BUY or SELL. If it falls within the threshold, it shows HOLD.
Without a threshold, you'd rebalance every day — incurring constant transaction costs and tax events for negligible benefit. The threshold creates a deadband: a zone of acceptable drift where you leave the portfolio alone and only act when misalignment is meaningful.
Typical settings: 1–2% — tight, suits large portfolios where transaction costs are small. 3–5% — standard for most retail and HNW investors. 5%+ — loose, implies annual or semi-annual rebalancing only.
The right setting depends on your portfolio size, transaction costs, and tax situation. Every rebalance may be a taxable event — the threshold helps you avoid unnecessary trades.
RISK VS RETURN
The single most important idea in all of investing: higher potential returns always come with higher risk. There is no free lunch. The question is always whether the extra risk is worth the extra expected return.
Return
Return is how much your investment grew (or shrank) over a period. If you invested $100 and it's now $112, your return is +12%. This tool shows Annual Return — what you'd expect to gain per year on average, based on 1–10 years of historical data. This is an estimate, not a guarantee.
Volatility (Risk)
Volatility measures how wildly a price bounces around. A low-volatility stock (e.g., a utility company) moves slowly and predictably. A high-volatility stock (e.g., a startup or crypto) can swing 10% in a single day. Mathematically, volatility is the standard deviation of daily returns, scaled to one year.
Annual Volatility
σ = std(daily_returns) × √252
There are ~252 trading days per year. A 20% annual volatility means returns in any given year typically fall within ±20% of the expected value — but can be further in rare years.
MODERN PORTFOLIO THEORY
MPT was invented by Harry Markowitz in 1952 (Nobel Prize). The core insight: combining assets that don't move in lockstep reduces your overall risk without necessarily reducing your return.
Think of two ice cream shops: one sells hot drinks, one cold. They do terribly in opposite weather. But if you own both, your combined business stays steady year-round. That's diversification. MPT turns this intuition into maths using each asset's expected return, volatility, and how pairs move together (correlation).
The Efficient Frontier
The Efficient Frontier is the curve in the OPTIMIZER tab showing every optimal portfolio possible from your assets. Every point on the curve is an allocation you can't improve without accepting more risk. The left end = minimum volatility; moving right = higher return but higher risk. Any portfolio below the curve is inefficient — you're taking the same risk for less return.
Capital Market Line
The Capital Market Line is the straight line drawn from the risk-free rate tangentially to the Efficient Frontier. Where it touches the curve is the maximum Sharpe portfolio — the single best risk-adjusted return available from your assets.
Measures how much return you get per unit of risk. Sharpe above 1.0 = decent; above 2.0 = excellent; below 0.5 = poor. Max Sharpe finds the allocation that maximizes this ratio — the best bang for your risk buck.
The Risk-Free Rate (default 4.5%) is what you'd earn risk-free — e.g., parking money in government bonds. It's the baseline: any investment should ideally beat it, or why take the risk?
Min Volatility
Ignores returns entirely and finds the portfolio with the lowest possible price swings. Useful if you want a calm, steady ride and don't need to maximize growth. Often results in more diversified allocations than Max Sharpe.
Max Quadratic Utility
Balances return and risk directly: Utility = Expected Return − (½ × risk_aversion × Variance). A middle ground between Max Sharpe and Min Volatility — it explicitly penalizes variance while still chasing return.
RETURNS MODELS
Mean Historical Return
The simplest approach: average the daily returns over the lookback period and annualize. Assumes the past is a reasonable guide to the future. Works well for stable, mature assets; can mislead for assets with unusual recent performance.
Mean Historical Return
μ = ((1 + mean_daily_return)^252) − 1
We compound the average daily return up to an annual figure using geometric compounding (~252 trading days per year).
CAPM
Capital Asset Pricing Model
μ_i = R_f + β_i × (R_market − R_f)
Estimates each asset's expected return based on how much it moves with the overall market (β = beta). β > 1 amplifies market moves; β < 1 is more stable. R_market is estimated from SPY historical returns.
Black-Litterman
A Goldman Sachs model (1990) that starts from market-implied returns — what the market collectively believes each stock should return, based on current market capitalisations. It uses Bayesian statistics to produce more stable, diversified estimates. Mean historical returns can produce extreme, concentrated allocations because a stock that happened to do well recently looks very attractive; BL corrects for this. The tool fetches market caps from Yahoo Finance to compute the prior.
When you select Black-Litterman, the Investor Views panel appears in the sidebar. You can enter your own expected annual return and confidence per ticker. Tickers left blank inherit only the market-implied prior. Confidence (0–100%) controls how much your view overrides the market consensus.
AI Analyst
The most advanced returns model. AI Analyst uses Claude AI as an automated Investment Committee. It gathers live fundamental data (PE, PB, EPS, margins, analyst consensus) from Finnhub and macro indicators (Fed Funds rate, yield curve, breakeven inflation, credit spreads, VIX) from FRED. Claude synthesizes this into forward-looking expected return views with confidence scores per ticker, which are then fed into the Black-Litterman model as machine-generated investor views.
Three additional controls appear when AI Analyst is selected: Risk Tolerance (1–5 scale from Conservative to Aggressive), Market Outlook (Bullish / Neutral / Bearish / Let AI Decide), and an optional Investment Thesis text box where you can guide the AI's reasoning. See the AI ANALYST tab for full details.
RISK METRICS
Sharpe Ratio
Sharpe = (Ann. Return − Risk-Free Rate) / Ann. Volatility
Return per unit of total risk. Uses all volatility — both up and down days count against you.
Like Sharpe, but only penalizes downward volatility. Going up fast doesn't count as risk — only going down does. A higher Sortino than Sharpe means the asset's volatility is mostly to the upside, which is good.
Max Drawdown
Max DD = min((Portfolio Value − Peak Value) / Peak Value)
The largest peak-to-trough decline. If your portfolio hit $100K then fell to $65K before recovering, the max drawdown is −35%. This measures the worst-case loss you'd have experienced if you held throughout. Lower (less negative) is better.
The correlation matrix in the RISK tab shows all pairwise correlations. Low values = good diversification. Values above 0.8 signal concentrated risk.
COVARIANCE & LEDOIT-WOLF
The covariance matrix captures how every pair of assets moves together. It's the engine of MPT — without it, we can't compute portfolio volatility or solve the optimization.
The problem: with many assets and limited data, the raw sample covariance matrix tends to be noisy and unreliable. Small historical quirks get amplified, producing extreme, concentrated allocations.
Ledoit-Wolf shrinkage is a mathematical technique that "shrinks" the noisy empirical covariance matrix toward a more stable target. The result is a cleaner, more reliable estimate of true co-movement — leading to better, more robust portfolios. Think of it like applying noise reduction to an audio recording before analyzing it.
This is why the optimizer uses Ledoit-Wolf instead of the raw sample covariance — it produces meaningfully better results, especially when you have fewer data points or a large number of assets relative to the lookback window.
VALUE AT RISK (VaR) & EXPECTED SHORTFALL (CVaR)
Value at Risk (VaR 95%)
VaR_95 = −(μ − 1.6449 × σ)
The maximum loss you'd expect in 95% of years (annual, parametric). A VaR of 15% means there's a 5% chance you could lose more than 15% in a single year.
Conditional VaR / Expected Shortfall (CVaR 95%)
CVaR_95 = −(μ − 2.0628 × σ)
The average loss in the worst 5% of scenarios. Always worse than VaR — it answers: when things go bad, how bad on average? Regulators and institutional investors focus on CVaR because VaR alone doesn't tell you how deep the tail goes.
STRESS TEST SCENARIOS
The RISK tab shows estimated portfolio impact during three historical crises: GFC 2008 (Oct 2007–Mar 2009, market −56%), COVID 2020 (Feb–Mar 2020, −35%), and Rate Rise 2022 (Jan–Oct 2022, −25%). Each scenario is scaled by your portfolio's beta vs the S&P 500 — if your portfolio has a beta of 1.2, the estimated impact is 1.2× the market drop.
These are approximations based on linear beta scaling. Actual drawdowns during crises depend on factor exposures, sector concentration, and liquidity — the numbers give you a directional sense of tail risk, not a precise forecast.
DIVERSIFICATION RATIO
The Diversification Ratio is the weighted average of individual asset volatilities divided by the portfolio's total volatility. A ratio above 1.0 means the portfolio is less volatile than its parts — the higher the ratio, the more diversification benefit you're receiving. A ratio of 1.0 means no benefit at all (perfectly correlated assets).
CORRELATION PERIOD SELECTOR
The correlation matrix in the RISK tab supports period selection: 6M, 1Y, 3Y, and FULL. Shorter periods highlight recent regime shifts — correlations often spike during market stress. If two assets show low full-period correlation but high 6M correlation, the diversification benefit may be weakening.
MONTE CARLO SIMULATION
Monte Carlo simulation explores the future by running thousands of possible scenarios. Here's how it works for a portfolio:
We take all the daily returns your portfolio actually experienced historically.
We randomly resample those returns thousands of times to simulate possible futures.
Each simulation traces a different path your portfolio could take over the chosen horizon.
We summarize the range of outcomes using percentiles.
Percentile Bands — BACKTEST tab
P10 · P50 · P90
P90 (optimistic): 90% of simulations ended below this — only 1-in-10 paths beat it. P50 (median): half of all runs were better, half were worse — the most likely single outcome. P10 (pessimistic): 10% of simulations ended below this — a cautious planning floor.
The shaded cone widens as you look further into the future — uncertainty compounds over time. Use the SIMULATIONS buttons to run 500, 1,000, or 5,000 paths; more simulations = smoother bands.
Important caveat: Monte Carlo assumes future daily returns are drawn from the same distribution as the past. Real markets can behave very differently — this is a probabilistic estimate, not a crystal ball. Use it to understand the range of plausible outcomes, not to predict a specific one.
REBALANCING DRIFT
When you set up a portfolio, you target certain weights — e.g., 40% Apple, 30% Microsoft, 30% Google. Over time, as prices move differently, those weights drift. If Apple doubles and the others stay flat, Apple now represents ~57% of your portfolio, not 40%.
Why does this matter?
Your portfolio is no longer optimized the way you intended.
You may have unintentionally become overexposed to your best performer.
Rebalancing — selling what grew, buying what lagged — restores the original risk profile.
The Rebalancing Drift chart (P&L tab → REBALANCING sub-tab) shows each position's current weight vs. target. The Rebalancing Actions table gives you exact BUY/SELL/HOLD instructions with dollar amounts.
Drift Signal
Drift = Current Weight − Target Weight
+ve drift: position has grown above target → candidate for trimming (SELL signal). −ve drift: position has shrunk below target → candidate for topping up (BUY signal). HOLD: drift is within tolerance — no action needed.
AI ANALYST — CLAUDE-POWERED INVESTMENT COMMITTEE
The AI Analyst returns model transforms StocksBro v2 from a backward-looking calculator into a forward-looking advisor. It replicates an institutional Investment Committee: multiple AI analysts debate the bull and bear cases, a chief strategist critiques all arguments, and the committee delivers a final positioning — with ETF overlap detection to catch hidden concentration risk.
Select AI Analyst from the Returns Model dropdown in the sidebar. Three new controls appear: Risk Tolerance, Market Outlook, and Investment Thesis.
HOW IT WORKS
01
DATA GATHERING
For each ticker in your portfolio, the engine fetches live fundamental data from Finnhub: PE ratio, PB ratio, EV/EBITDA, EPS, revenue growth, gross and net margins, ROE, debt-to-equity, beta, 52-week range, and analyst consensus (buy/hold/sell counts).
02
MACRO SNAPSHOT
Seven macro indicators are fetched from FRED (Federal Reserve Economic Data): Fed Funds Rate, Yield Curve (10Y−2Y and 10Y−3M spreads), 5Y Breakeven Inflation, High-Yield Credit Spread, Investment-Grade Credit Spread, and VIX (volatility index).
03
AI INVESTMENT COMMITTEE
Claude AI receives all fundamentals, macro data, your investor context (risk tolerance, outlook, thesis), and any ETF overlap analysis. It operates as a multi-perspective investment committee: bull and bear analysts argue their cases, a chief strategist critiques all positions, and the committee delivers a regime assessment, per-asset expected return views with confidence scores, risk flags (including overlap warnings), and allocation tilt guidance.
04
BLACK-LITTERMAN INTEGRATION
The AI's views are converted into absolute return views and fed into the Black-Litterman model with Idzorek confidence weighting. BL blends these views with market-implied prior returns (from market capitalizations) to produce posterior expected returns, which the optimizer then uses to compute weights.
05
EXPLANATION
A second Claude call generates a natural-language explanation of why the portfolio is allocated the way it is, including why specific stocks were excluded (0% weight) and how any ETF overlap was handled. This appears in the Why This Allocation section below the allocation bars in the OPTIMIZER tab.
SIDEBAR CONTROLS
Risk Tolerance (1–5)
Adjusts the AI's aggressiveness. 1 (Conservative) biases toward lower return estimates and lower confidence. 5 (Aggressive) increases conviction on growth-oriented views. Default is 3 (Moderate).
Market Outlook
Bullish — AI biases expected returns upward. Neutral — AI uses data as-is, no directional bias. Bearish — AI biases expected returns downward. Let AI Decide (default) — AI determines its own outlook from the macro data.
Investment Thesis
Optional free text (up to 500 characters) to guide the AI. Examples: "Overweight tech on AI tailwinds, avoid rate-sensitive sectors" or "Focus on dividend sustainability and cash flow". When provided, the AI factors your thesis into its view formation.
RESULTS — WHY THIS ALLOCATION
When AI Analyst is used, several new sections appear in the OPTIMIZER tab after the allocation bars:
Regime Badge
A color-coded label showing the AI's macro regime assessment: EXPANSION (green), LATE CYCLE (amber), CONTRACTION (red), or RECOVERY (blue).
Explanation
A 4–6 sentence plain-prose explanation of the allocation, referencing specific data points (forward PE, analyst consensus, yield curve, etc.). Includes reasoning for why excluded (0% weight) stocks were left out, and how any ETF overlap was factored in.
Risk Flags
Short warning chips highlighting portfolio-level risks identified by the AI — e.g., sector concentration, inverted yield curve, elevated VIX. Also includes ETF overlap warnings when near-duplicate ETFs or overexposed individual stocks are detected.
Exposure Overlap Detected
When your portfolio contains both individual stocks and ETFs that hold those stocks (e.g., NVDA + SPY + QQQ), this section shows the overlapping exposure. It also flags ETF pairs with high holding overlap (e.g., SPY and VOO are near-duplicates). The AI factors this into its confidence levels and allocation views.
Investment Committee
A brief summary of the AI committee's multi-perspective reasoning: the bull case (why upside exists), bear case (what could go wrong), a critique (what both sides are missing), and the committee decision (how the final positioning was determined). This shows that the AI is thinking adversarially, not just presenting one view.
AI Views Per Asset
A table showing the AI's view for each ticker: expected annual return, confidence level (0–100%), and one-sentence reasoning. These are the exact inputs that were fed into the Black-Litterman model.
Important: AI views are blended with market-implied priors via Black-Litterman. A stock with a seemingly high AI return (e.g., 9%) may still receive 0% weight if its market-implied prior was even higher (e.g., 14%) — in that case, the AI is actually bearish relative to the market. Confidence controls how strongly the AI view overrides the prior.
RESEARCH TAB — COMMITTEE ANALYSIS
The Research tab (F6) uses a two-pass AI committee approach for individual ticker analysis:
Pass 1: Three analysts present the bull case, bear case, and base case — each citing specific fundamentals. A chief strategist then critiques all three arguments, identifying what each is ignoring or overstating.
Pass 2: An investment committee chair reviews the full debate and delivers a final verdict (ATTRACTIVE / NEUTRAL / AVOID) with a confidence level (HIGH / MODERATE / LOW), which case they sided with, the key risk the winning case underweights, and a 12-month directional view.
COST
Each AI Analyst optimization costs approximately $0.002 in Claude API usage (two calls: committee views + explanation). Each Research report costs approximately $0.0004 (two calls: analyst debate + committee verdict). All other data sources (Finnhub, FRED, Yahoo Finance) are free.
●
●
TICKER
Enter any US equity ticker and press ANALYSE to generate an AI research report.
fn+F6
AI Research
Your personal investment committee — on demand. Type any ticker (US equities, ASX, crypto) and receive an institutional-grade research report. A two-pass adversarial AI system debates bull, bear, and base cases using live fundamentals and macro context, then delivers a committee verdict.
AAPLApple Inc.
NASDAQTechnologyUSD
Market Cap
$3.4T
P/E Ratio
32.8
EPS (TTM)
$6.42
Revenue Growth
+8.1%
BULL CASESAMPLE DATA
Apple's services revenue continues to compound at 12%+ annually, creating a high-margin annuity stream that the market undervalues relative to pure-play SaaS peers. The installed base of 2.2B active devices provides an unassailable distribution moat for new product launches. iPhone cycle upgrades driven by Apple Intelligence AI features should accelerate ASP growth through FY25.
BEAR CASESAMPLE DATA
At 32.8x forward earnings, Apple is priced for perfection with limited margin of safety. China revenue (18% of total) faces sustained macro headwinds and Huawei competitive pressure. Hardware growth has stagnated at low-single-digits, and the Vision Pro has yet to demonstrate mass-market traction. Regulatory risk from EU DMA and US antitrust probes could force App Store fee reductions.
FINANCIAL HEALTHSAMPLE DATA
Gross Margin
46.2%
Net Margin
25.3%
Debt/Equity
1.87
FCF Yield
3.1%
Revenue
$385B
EBITDA
$130B
VALUATION ANALYSISSAMPLE DATA
P/B Ratio
48.2
EV/EBITDA
26.1
PEG Ratio
2.94
52W Range
$164–$237
Beta
1.24
Div. Yield
0.44%
CHIEF STRATEGIST CRITIQUESAMPLE DATA
The bull case overstates Apple Intelligence as a near-term catalyst — AI feature adoption historically follows a slow curve and won't materially change upgrade cycles for 2+ quarters. The bear case underestimates Services pricing power and ignores that China revenue has already troughed. Both cases neglect the $110B annual buyback program's mechanical support to EPS growth regardless of top-line trajectory.
COMMITTEE VERDICTSAMPLE DATA
ATTRACTIVE
Confidence:MODERATE
Sided with:Bull Case
Analyst Consensus:24 Buy / 8 Hold / 2 Sell
Key Risk:China revenue deterioration beyond current consensus expectations
12M View:Moderate upside (+10-15%) driven by Services re-rating and buyback support
RECENT NEWSSAMPLE DATA
2 hours agoApple Intelligence rollout expands to 18 new countries
1 day agoServices revenue hits $24.2B quarterly record, up 14% YoY
3 days agoEU fines Apple €1.8B over App Store anti-steering rules
2-Pass AI Committee
Pass 1: Three analysts argue bull/bear/base cases. A chief strategist critiques all arguments. Pass 2: Committee chair delivers final verdict with confidence level and 12-month view.
Global Coverage
US equities (AAPL), ASX stocks (CBA.AX), crypto (BTC-USD), and any Yahoo Finance-compatible symbol. Reports cached 30 days with fresh/cached indicators.
Type a ticker in the search bar above and press ANALYSE to generate your report.