willBB Markets Terminal

A Bloomberg-style markets terminal built ground-up on free data sources. Self-contained, ~36 source files, no SaaS dependencies, ~160 KB First Load JS. Designed to demonstrate that a real-time quant terminal with a Pine-DSL backtester, Carhart 4-factor regression, and TradingView-grade charts can be built without spending a dollar on data.

Standalone version of the willBB app from w1zz7/willos-98-portfolio. Same code, no Win98 desktop chrome — just the terminal, fullscreen.

What it does

Four top tabs, each a real workspace:

1. Markets

TradingView widget chart for the focused symbol (real-time, regulated data feed via TradingView)
Live ticker tape at the top: S&P 500, Nasdaq, Dow, Russell 2K, VIX, WTI, Brent, Gold, Silver, EUR/USD, USD/JPY, Bitcoin, Ethereum
144-symbol watchlist with prices polling every 15 seconds
Click any row → focus that symbol → all other tabs follow

2. Equity Research (14 sub-tabs per ticker)

Sub-tab	Source	What's there
Profile	Yahoo `/v10/quoteSummary`	Sector, industry, employees, business summary
Technicals	embedded TradingView	TV's full studies stack (RSI, MACD, Bollinger, etc.)
Statistics	Yahoo modules	Market cap, P/E, EPS, ROE, beta, 52-week range, short interest
Income / Balance / Cash Flow	Yahoo modules	4 years of fundamental statements with TTM column
Analysts	Yahoo modules	Recommendation distribution, price targets, upgrades/downgrades
Earnings	Yahoo modules	EPS history with surprise % vs estimate
Holders	Yahoo modules	Top mutual fund + institutional holders
Smart Money	Alpha Vantage	Insider transactions (last 50, color-coded buy/sell) + institutional holdings (top 25 by % O/S) side-by-side
Transcript	Alpha Vantage	Earnings call transcript with paragraph-level LLM sentiment scoring + speaker stats (top 6 by paragraph count)
Dividends	Yahoo modules	Yield, payout, ex-date, history
Options	Yahoo `/v7/options`	Full chain with greeks (when available), put-call ratios
News + Sentiment	Alpha Vantage	LLM-tagged ticker sentiment per article + topic chips + aggregate gauge

3. Discovery (3 sub-tabs)

Screeners — Yahoo's day_gainers / day_losers / most_actives predefined screeners as 3 columns
Macro — yield-curve hero (3M / 2Y / 5Y / 7Y / 10Y / 30Y) plus 7 sparklines: Fed Funds Rate, 10Y Treasury, 2Y Treasury, CPI, Inflation YoY, Unemployment, Real GDP. Inversion warning baked in.
Calendars — Earnings calendar (next 3 months) + IPO calendar (next 3 months). Click a row to focus that ticker.

4. Research (the quant terminal — 4 panes)

Studies

Symbol header with live ticking price + day-over-day %, source badge
Hand-rolled SVG candlestick chart (no lightweight-charts dep): TradingView-style polish — right-axis price scale, hollow up-candles, faint ticker watermark, pulsing live-tip dot, dashed live-price line, hover crosshair with floating OHLC tooltip showing all overlay values at the hovered bar
12 indicator toggles: SMA, EMA, RSI, MACD, ATR, Bollinger Bands, Yang-Zhang vol, Garman-Klass vol, Parkinson vol, close-to-close vol, log returns, rolling beta to ^GSPC
Quant readouts panel (right rail): Sharpe (1y), Annualized σ, RV YZ, RV GK, β to SPX, Sortino, Information Ratio, Hurst (R/S), ADF stat, ACF(1), with significance bands
ACF + PACF correlogram bar charts with ±1.96/√N significance bands

Cross-Section

32-name liquid US large-cap universe
4 cross-sectional alpha signals: 12-1 momentum, 1-month reversal, low volatility, drawdown recovery
Decile-sort + IC + IC decay + L/S equity curve
12-asset 60-day return correlation matrix
Watchlist heatmap with sector rotation

Alpha Lab

Pine v5-flavored DSL editor with 6 built-in presets: MA Crossover (50/200), RSI mean-reversion, BB breakout, momentum, carry, two-factor combo
DSL compiles to JS via new Function(...) with helpers plot/long/short/exit and full indicator library injected as ind.* — write strategy code, hit RUN BACKTEST, see signals overlaid on the chart
Walk-forward CV (configurable train/test windows that step forward)
Transaction-cost model: fixed commission + bid-ask half-spread + sqrt market impact (Almgren-Chriss-style) + borrow cost annualized to per-bar
Results: equity curve, drawdown, Sharpe, Sortino, win rate, profit factor, expectancy, max drawdown, Calmar
Paper Blotter: every signal generates a paper trade, mark-to-market live

PnL Attribution

Rolling 60-day Sharpe / Sortino / Information Ratio
Realized vol: Yang-Zhang vs close-to-close
Drawdown chart with peak/trough markers
Returns histogram + QQ plot vs normal (visualize tail fatness)
Carhart 4-factor regression of strategy returns on (Mkt-RF / SMB / HML / MOM) reconstructed from ETF spreads:
- SMB ≈ IWM − SPY (small-minus-big)
- HML ≈ IUSV − IUSG (high-minus-low book-to-market)
- MOM ≈ MTUM − SPY (momentum)
Multivariate OLS via Gauss-Jordan matrix inversion (no math library) with t-stats, R², adjusted R²
Variance decomp: % systematic vs % idiosyncratic
Rolling factor exposures (60d windows)
VaR / Expected Shortfall (95%, 99%) via historical simulation + Cornish-Fisher
Position sizer: account equity × risk % / stop distance → shares + leverage

Data architecture

A 4-tier provider failover behind every chart and quote, with each response tagged source so the UI can show a colored badge:

Yahoo Finance v8  →  CoinGecko (crypto)  →  Alpha Vantage  →  Stooq  →  Synthetic
       |                  |                     |                |          |
  rate-limited       real-time spot         daily compact      CSV daily  regime-switching
  retry x4 with      via /simple/price      free 25/day        last-known GBM placeholder
  rotating UA +      (BTC/ETH/SOL/DOGE)     EOD quote +        cached     so panels render
  cookie refresh                            news + macro                  even if all fail

Source badge in the header: LIVE (green) · DELAYED (amber) · EOD (purple) · CACHED (gray) · SYNTHETIC (red).

Yahoo Finance v8 (primary)

4-attempt retry with exponential backoff (350 → 3500 ms)
Rotating User-Agent pool (Chrome 120/121, Win Chrome, macOS Safari 17.2)
Cookie auto-refresh from fc.yahoo.com (force-refresh on attempt ≥ 2)
Alternates between query1.finance.yahoo.com and query2.finance.yahoo.com
Per-symbol module-scope success cache (90s for charts, 5s/30s split for quotes)

Alpha Vantage (free 25/day budget)

Multiplexed /api/markets/alpha?fn=... route covers 17 endpoints
Rolling 24-hour budget guard: caps at 22 calls (3 in reserve)
5/min throttle: ≥13s spacing between calls (under the 5/min ceiling with margin)
Per-endpoint cache TTLs from 5 min (live spot prices) to 30 days (earnings transcripts)
Symbol translation: ^GSPC → SPY, ^IXIC → QQQ, etc. (free tier doesn't ship indices)

Real-time chart

useLiveQuote(symbol, intervalMs=5000) hook polls /api/markets/quotes every 5 seconds
Splices live tick into the latest candle's close (h/l auto-extend if the print exceeds the daily high/low)
All downstream computations re-derive: closes, log returns, BB, RSI, MACD, realized vol, Sharpe, ACF/PACF
Pauses when document.hidden; per-symbol AbortController on unmount/symbol-change
Server-side cache TTL is split: 5s for ≤24-symbol calls (live), 30s for batch (watchlist)

Stress-tested

20-symbol concurrent burst → 12 Yahoo + 8 Stooq, zero degraded
Rapid symbol switch (4 tickers in 3s) → all in-flight aborted, only the final settles
12-poll long session at 5s → no memory leaks, polls keep firing
Tab backgrounded → polls pause, resume on focus

Quant primitives

components/apps/willbb/quantdesk/indicators.ts:

// Realized volatility (annualized)
export function realized_vol_cc(closes: number[], n: number): number[]    // close-to-close
export function realized_vol_pk(bars: Bar[], n: number): number[]         // Parkinson (HL)
export function realized_vol_gk(bars: Bar[], n: number): number[]         // Garman-Klass
export function realized_vol_yz(bars: Bar[], n: number): number[]         // Yang-Zhang (full OHLC + overnight)

// Time-series properties
export function log_ret(closes: number[]): (number | null)[]
export function cum_log_ret(closes: number[]): (number | null)[]
export function acf(returns: (number | null)[], maxLag: number): number[]
export function pacf(returns: (number | null)[], maxLag: number): number[]  // Durbin-Levinson
export function hurst(closes: number[]): number                              // R/S analysis
export function adf_stat(closes: number[]): number                           // augmented Dickey-Fuller

// Risk-adjusted return
export function rolling_beta(assetRet: number[], mktRet: number[], n: number): number[]
export function sortino(returns: (number | null)[], n: number): (number | null)[]
export function information_ratio(activeRet: number[], n: number): (number | null)[]

// Cross-section utilities
export function rank(values: number[]): number[]                  // 1-based ranks
export function zscore(values: number[]): number[]
export function winsorize(values: number[], pct: number): number[]
export function pct_change(values: number[]): (number | null)[]
export function rolling_corr(x: number[], y: number[], n: number): number[]

Plus 12 retail classics (SMA, EMA, RSI, MACD, ATR, Bollinger, Stochastic, OBV, VWAP, ADX, Donchian, Ichimoku) and helpers (crossover, crossunder, highest, lowest).

Backtest engine

components/apps/willbb/quantdesk/backtest.ts:

runBacktest(bars, signals, params) → BacktestResult
  - Per-bar evaluation
  - Long / short / exit signals
  - Per-trade P&L with: commission + bid-ask half-spread + sqrt market impact + borrow cost
  - Stats: total return, Sharpe, Sortino, win rate, profit factor, expectancy,
    max drawdown, Calmar, exposure %, avg holding period, turnover

runWalkForward(bars, signalFn, trainBars, testBars, stepBars) → WalkForwardResult
  - Train/test windows step forward through history
  - Out-of-sample only — no look-ahead
  - Returns per-window stats + concatenated OOS equity curve

Factor regression (factorRegression.ts):

runFactorRegression(strategyReturns, factors) → FactorRegressionResult
  // OLS with Gauss-Jordan matrix inversion
  // Returns: alpha, betas, t-stats, R², adjusted R², residuals, systematic %

buildFactorReturns({ SPY, IWM, IUSV, IUSG, MTUM }) → CarhartFactors
  // Mkt-RF ≈ SPY (subtract risk-free if available)
  // SMB ≈ IWM - SPY
  // HML ≈ IUSV - IUSG
  // MOM ≈ MTUM - SPY

API routes

Route	Returns	Cache
`GET /api/markets/chart?symbol=AAPL&range=1y&interval=1d`	OHLCV with 4-tier failover	90s success / 0 synthetic
`GET /api/markets/quotes?symbols=AAPL,MSFT`	Last-price snapshot	5s ≤24 sym / 30s batch
`GET /api/markets/equity?module=profile&symbol=AAPL`	Yahoo modules (profile, statistics, income, balance, cashflow, analysts, earnings, holders, dividends, options, news)	per-module
`GET /api/markets/alpha?fn=NEWS_SENTIMENT&tickers=AAPL`	Multiplexed AV endpoint (17 functions)	5min - 30day per fn

Diagnostic: GET /api/markets/alpha?fn=BUDGET returns {keyPresent, callsLast24h, budget, canSpend} without spending a call.

Local development

npm install
npm run dev

Open http://localhost:3000. The terminal renders fullscreen.

Script	What it does
`npm run dev`	Next.js dev server (port 3000)
`npm run build`	Production build
`npm run start`	Run the production build locally
`npm run typecheck`	`tsc --noEmit` (0 errors expected)

Required environment variables

The app works fully without any env vars — Yahoo + CoinGecko + Stooq + synthetic fallback covers charts/quotes/equity-research/screeners. However, the following panels are powered by Alpha Vantage and render empty without a key:

News + Sentiment
Smart Money (Insider transactions, Institutional holdings)
Macro (Yield curve, CPI, Fed Funds, GDP, Inflation, Unemployment)
Earnings + IPO calendars
Symbol Search autocomplete
Earnings Transcripts

Env var	Required?	How to get
`ALPHA_VANTAGE_API_KEY`	Optional (recommended)	https://www.alphavantage.co/support/#api-key (free, 25 calls/day)

Local development:

# .env.local
ALPHA_VANTAGE_API_KEY=your_free_key

Netlify production: Site → Site settings → Environment variables → add ALPHA_VANTAGE_API_KEY. Then Deploys → Trigger deploy → Clear cache and deploy site so the function bundle picks up the new value.

Deployment

Netlify (recommended — `netlify.toml` is included)

netlify.toml pins Node 20 LTS and sets cache headers on /api/markets/*. Netlify's built-in Next.js Runtime handles SSR pages, the API routes (deployed as on-demand Netlify Functions), and ISR cache headers automatically — no plugin block needed.

npx netlify-cli login
npx netlify-cli init       # link this folder to a new or existing site
npx netlify-cli deploy --prod

Or via the Netlify dashboard:

Add new site → Import from Git → choose this repo
Build settings auto-detect from netlify.toml (build cmd npm run build, publish .next)
Optional: add ALPHA_VANTAGE_API_KEY under Site settings → Environment variables to unlock News+Sentiment, Smart Money, Macro, Calendars, Symbol Search, Transcripts. Without it, those panels show "unavailable" and everything else still works.
Deploy. Subsequent pushes to main auto-deploy.

Vercel

npx vercel --prod

No vercel.json needed — Next.js conventions are auto-detected. Add ALPHA_VANTAGE_API_KEY in the Vercel dashboard's env vars.

Other Node hosts

Standard Next.js — Railway, Render, Fly all work without changes. Set ALPHA_VANTAGE_API_KEY in the host's env config.

Project structure

willbb-markets/
├── app/
│   ├── api/markets/
│   │   ├── chart/         # 4-tier OHLCV failover
│   │   ├── quotes/        # split-TTL quote cache
│   │   ├── equity/        # Yahoo modules
│   │   └── alpha/         # multiplexed Alpha Vantage
│   ├── layout.tsx         # root layout (no Win98 chrome)
│   ├── page.tsx           # mounts <WillBBTerminal /> fullscreen
│   └── globals.css        # tailwind + willbb-livepulse keyframes
│
├── components/apps/willbb/
│   ├── OpenBB.tsx         # top-level: tabs (Markets/Equity/Discovery/Research) + ticker strip
│   ├── BootScreen.tsx     # Bloomberg-style cold-boot animation
│   ├── EquityResearch.tsx # 14-sub-tab equity research panel
│   ├── EquityAlphaViews.tsx # News+S, Smart Money, Transcript (AV-powered)
│   ├── Discovery.tsx      # Screeners + Macro + Calendars tabs
│   ├── MacroPanel.tsx     # yield curve hero + 7 macro sparklines
│   ├── CalendarsPanel.tsx # earnings + IPO
│   ├── PriceChart.tsx     # legacy small line chart
│   ├── TechnicalsView.tsx # TradingView embedded widget config
│   ├── TradingViewChart.tsx
│   ├── SymbolSearch.tsx   # AV-powered autocomplete
│   ├── SourceBadge.tsx    # LIVE/DELAYED/EOD/CACHED/SYNTHETIC pill
│   ├── symbols.ts         # INDEX_STRIP + WATCHLIST_ORDER
│   └── quantdesk/
│       ├── QuantDesk.tsx        # 4-sub-tab Research panel
│       ├── Cockpit.tsx          # Studies pane
│       ├── Scanner.tsx          # Cross-Section pane
│       ├── StrategyLab.tsx      # Alpha Lab pane (DSL editor)
│       ├── RiskDashboard.tsx    # PnL Attribution pane
│       ├── QuantChart.tsx       # hand-rolled SVG candlestick chart
│       ├── PaperBlotter.tsx     # paper-trade ledger
│       ├── indicators.ts        # 18 quant primitives
│       ├── backtest.ts          # backtest engine + walk-forward CV
│       ├── factorRegression.ts  # OLS + Carhart 4-factor
│       └── presets.ts           # 6 built-in DSL strategies
│
├── lib/
│   ├── alphavantage.ts    # 17 AV endpoint fetchers + budget guard
│   ├── stooq.ts           # CSV historical + last-quote
│   ├── useLiveQuote.ts    # 5s polling hook
│   ├── marketsFallback.ts # SEED_QUOTES dictionary
│   ├── equityFallback.ts  # equity research seed data
│   ├── equityModuleFallback.ts
│   └── wm/types.ts        # minimal WindowState stub
│
└── data/
    └── trades.ts          # 267 paper-trade records (for the Paper Blotter)

Bundle size

First Load JS: 160 KB — comparable to a static landing page
All AV/Stooq/Yahoo retry logic is server-side only (zero client-side AV)
Hand-rolled SVG candlestick chart instead of lightweight-charts (would have added ~50 KB)
No state management library beyond React hooks (no Redux/Zustand)

License

MIT. Use this code freely.

The Paper Blotter records in data/trades.ts are real personal trade history — the data isn't licensed for redistribution but the schema/format is. Replace with your own.

Credits

Designed + engineered by Will Zhang as part of the WillOS 98 Portfolio.

Data: Yahoo Finance v8, CoinGecko, Alpha Vantage, Stooq, TradingView (Markets-tab embed).

Built with Next.js 15, React 19, TypeScript 5, Tailwind CSS 4.

Name		Name	Last commit message	Last commit date
Latest commit History 9 Commits
.claude		.claude
app		app
components/apps/willbb		components/apps/willbb
data		data
lib		lib
public		public
.env.example		.env.example
.gitignore		.gitignore
LICENSE		LICENSE
README.md		README.md
netlify.toml		netlify.toml
next.config.ts		next.config.ts
package-lock.json		package-lock.json
package.json		package.json
postcss.config.mjs		postcss.config.mjs
tsconfig.json		tsconfig.json

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Repository files navigation

willBB Markets Terminal

What it does

1. Markets

2. Equity Research (14 sub-tabs per ticker)

3. Discovery (3 sub-tabs)

4. Research (the quant terminal — 4 panes)

Studies

Cross-Section

Alpha Lab

PnL Attribution

Data architecture

Yahoo Finance v8 (primary)

Alpha Vantage (free 25/day budget)

Real-time chart

Stress-tested

Quant primitives

Backtest engine

API routes

Local development

Required environment variables

Deployment

Netlify (recommended — `netlify.toml` is included)

Vercel

Other Node hosts

Project structure

Bundle size

License

Credits

About

Uh oh!

Releases

Packages

Uh oh!

Contributors

Uh oh!

Languages

Folders and files

Latest commit

History

Repository files navigation

willBB Markets Terminal

What it does

1. Markets

2. Equity Research (14 sub-tabs per ticker)

3. Discovery (3 sub-tabs)

4. Research (the quant terminal — 4 panes)

Studies

Cross-Section

Alpha Lab

PnL Attribution

Data architecture

Yahoo Finance v8 (primary)

Alpha Vantage (free 25/day budget)

Real-time chart

Stress-tested

Quant primitives

Backtest engine

API routes

Local development

Required environment variables

Deployment

Netlify (recommended — netlify.toml is included)

Vercel

Other Node hosts

Project structure

Bundle size

License

Credits

About

Topics

Resources

License

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

Netlify (recommended — `netlify.toml` is included)

Packages