quant-garage

Single-commodity macro read. Answers "is this commodity in a winning or losing macro setup right now" and names the macro driver that dominates it. Pulls one commodity ETF (default GLD; accepts SLV, USO, DBC, or any commodity ETF) plus the macro context it needs (UUP for the dollar, TIP and IEF for a real-yield proxy, and GDX/SLV for gold's miner and silver co-move set) and computes the drivers that push the commodity around: rolling DXY correlation, real-yield correlation (vs the TIP-minus-IEF spread), miner divergence (gold only), silver co-movement (gold only), and a momentum quintile. The take reads constructive / neutral / headwind and names the dominant variable. It would have flagged a gold drawdown two weeks early via the dollar and real-yield readings. Use when the question is about one commodity specifically ("is gold in trouble", "what's driving crude", "is the dollar the problem for gold"). Runs on any stocks tier (Free Basic works with --sleep 13).

Split one name's relative strength into two clean legs, name vs its sector ETF and sector vs benchmark, so you can tell whether the name is strong because its sector is strong or because it is pulling away from its own sector. Maps the ticker to its SPDR sector ETF (XLK, XLF, XLE, XLV, XLI, XLY, XLP, XLU, XLB, XLRE, XLC) with a --sector override for any name. Computes name-vs-sector, sector-vs-benchmark, and name-vs-benchmark RS in basis points across 5/20/60/120-day windows, a divergence score, and a take that classifies the name as leading its sector, lagging its sector, or diverging. Use when a single name looks strong or weak vs SPY and you need to know if it is name-specific or sector-driven. Requires Stocks Starter (Free Basic works, only three series per run).

Proposes concrete, live-priced option hedges against a single long position and ranks them by cost per dollar of downside protected. Takes a ticker and a position size (shares or notional), pulls the underlying price and the options chain around the horizon expiry, and constructs five standard overlays (covered call, protective put, collar, put spread, ratio put spread) priced from chain mids with net cost, breakeven, max loss, max gain, and net delta/gamma/theta at open. The take line recommends the structure that fits the stated risk tolerance. Use when a concentrated long needs a hedge and you want the actual structures and their live cost, not just "you are over-exposed." Needs Stocks Starter plus the Options Developer add-on. Not advice.

Prescriptive macro scenario analysis on a current book. Given a position book (CSV of ticker,shares) and a scenario stated as flags (--rates-bp, --dxy-pct, --oil-pct, --gld-pct), it regresses each position's daily returns on four macro factor ETFs (TLT for rates, UUP for the dollar/DXY, USO for oil, GLD for gold), translates the scenario into factor return shocks, and reports the expected per-position and book-level P&L with a rough 90% band, plus the dominant position and factor contributors. Use when the question is forward-looking and conditional: "what happens to my book if rates keep rising / the dollar keeps rallying / oil spikes / gold falls." This is the prescriptive counterpart to risk-report and portfolio-review, which are descriptive of the past. Runs on any stocks tier (Free Basic works with --sleep 13).

Pull Kalshi prediction market prices for Fed decisions, CPI, GDP, NFP, and other macro / market events. Report implied probability per outcome, aggregate cross-strike distribution when the series is a laddered strike set (like KXFED-27APR-T4.25, T4.00, T3.75...), expected value, modal outcome, and open interest. Prediction markets now clear enough volume post-2024 to reflect a real market-implied policy path, often diverging from surveyed economist consensus. Uses Kalshi's public read-only API; no authentication required.

Workflow composite for a single ticker heading into an earnings print. Chains earnings-blackout (timing check) + event-study (prior print reaction distribution) + guidance-tracker (management raise/cut track record) + analyst-tracker (sell-side positioning) + mc-portfolio-simulator (P&L distribution at proposed weight over the horizon). Emits a posture verdict (constructive_setup / mixed_setup / avoid_or_hedge / no_imminent_print). Requires Stocks Basic; guidance-tracker and analyst-tracker sections skip gracefully without Benzinga entitlements.

Workflow composite that runs manager-portfolio-diff across a curated cohort of well-known filers (Berkshire, Baupost, Renaissance, Bridgewater, Third Point, Pershing Square, Tiger Global, Scion, Appaloosa) and aggregates the initiations, adds, and exits by issuer. Surfaces cross-fund conviction: names that appeared in >= N funds' new positions this quarter. Requires Stocks Basic. Runs on the free tier.

Take a book (weights per ticker), compute the daily portfolio return series, and run the full tearsheet with deflated Sharpe correction (Bailey & Lopez de Prado) plus rolling IC vs benchmark. Emits a verdict (real_alpha / possibly_alpha / essentially_beta / underperforming / no_edge_evident) based on DSR significance, alpha annualized, and beta. Answers "is this book actually alpha, honestly?" Requires Stocks Basic.

Workflow composite that runs change-point-detector + hurst-exponent on SPY plus the 11 SPDR sector ETFs. Reports per-name the last detected regime shift, current persistence classification (mean_reverting / random_walk / trending), and cross-sector summary (broad_regime_shift / localized_regime_shift / trend_dominated / mean_reversion_dominated / mixed_stable). Requires Stocks Basic. Runs on the free tier.

Workflow composite that runs five filing / ownership skills on a single ticker (8-k-scanner + risk-factor-delta + filing-sentiment + insider-flow + analyst-tracker) and returns a unified fundamental report with a cross-source verdict (predominantly_constructive / predominantly_concerning / mixed / no_clear_signal). Handles entitlement gaps gracefully. Requires Stocks Basic; analyst-tracker section skipped without Benzinga Analyst Ratings.

Estimate net dealer gamma exposure (GEX) for same-day-expiry (or nearest-expiry) SPY / SPX / QQQ / IWM options and identify gamma pins. Uses Black-Scholes gamma applied to reported open interest with a standard dealer positioning assumption (short customer calls, long customer puts). Reports net dealer gamma, gamma regime (long / short), gamma flip strike, and top 5 gamma pin strikes with per-side notional gamma. Motivated by 2024-25 research on how 0DTE options now drive systematic intraday moves through market-maker delta hedging pressure. Requires Options Developer.

Rough-volatility-scaled vol forecast (Bayer-Friz-Gatheral 2016) for a ticker across multiple horizons. Under rough vol, realized vol scales as h^H with H around 0.14 empirically (Livieri et al. 2018), much slower than the sqrt(t) growth of Brownian motion. This dampens long-horizon extrapolation and lifts short-horizon estimates. Reports the rough-vol forecast alongside traditional Brownian scaling and EWMA for direct comparison at each horizon. Requires Stocks Basic. Runs on the free tier.

Estimate the half-life of a candidate signal by computing rolling information coefficient (IC) vs forward returns over a 5-year window and fitting an exponential decay to the IC series. Motivated by 2024-25 factor decay literature showing most published signals have decayed sharply post-publication. Reports the fitted half-life in trading days, the recent vs early IC delta (regime break check), and a full performance tearsheet on the signed-signal PnL. Four built-in signals: momentum, mean_reversion, vol_expansion, trend_break. Requires Stocks Basic. Runs on the free tier.

Bayesian Online Change-Point Detection (BOCPD) on a ticker's daily log returns. Detects points in time where the return-generating distribution changed (regime shift in mean, vol, or both), reports the confidence at each detected boundary, and emits per-segment statistics (annualized return, annualized vol) so the reader can see what changed. Uses Adams and MacKay (2007) BOCPD with a Normal-Gamma prior on (mu, tau) and a Student-t predictive so hyperparameters update in closed form. Requires Stocks Basic. Runs on the free tier.

Estimate the Hurst exponent for a single ticker's daily log returns using rescaled-range (R/S) analysis, and classify the series as mean_reverting (H < 0.45), random_walk (H in [0.45, 0.55]), or trending (H > 0.55). Reports per-block R/S values and a block-bootstrap confidence band around H. Companion to pairs-scanner: pairs handles two-name cointegration, hurst handles single-name persistence. Answers "is this name a mean-reversion setup or a momentum setup?" Requires Stocks Basic. Runs on the free tier.

Track sell-side analyst positioning on a name via Benzinga Analyst Ratings. Pulls every rating event over the lookback window, classifies each as upgrade / downgrade / initiation / reiteration / drop-coverage / PT-change, aggregates the latest rating and price target per firm, and reports the current consensus (median PT across firms plus buy/hold/sell distribution). Uses Massive's built-in Benzinga integration. Requires Stocks Basic + Benzinga Analyst Ratings entitlement.

Monte Carlo forward P&L simulator for a book. Simulates 10,000 correlated return trajectories from the shrunk covariance matrix over a caller-specified horizon (default 60 trading days) and reports the full cumulative-return distribution, max-drawdown distribution, path VaR, and P(loss > X%) at 5/10/20/30% thresholds. Companion to position-sizer. Requires Stocks Basic. Runs on the free tier.

Diff the two most recent quarterly 13-F filings for an institutional investment manager (Berkshire, Baupost, Renaissance, Pershing Square, Tiger Global, Appaloosa, Scion, etc.) using Massive's pre-parsed 13-F endpoint. Reports initiations, exits, adds (>= 25% share change), trims (<= -25%), and portfolio value change. Answers "what did Buffett/Klarman/Burry do last quarter?" Requires Stocks Basic. Runs on the free tier.

Score 10-K narrative sections (Business, Risk Factors) for a ticker using the Loughran-McDonald finance sentiment dictionary and report year-over-year tone shifts by category (negative, uncertain, litigious, modal-weak, modal-strong, constraining). Answers "did management's language get more defensive this year?" Uses Massive's pre-parsed 10-K sections endpoint. Requires Stocks Basic. Runs on the free tier.

Track corporate guidance history for a ticker via Benzinga Corporate Guidance. Classifies each event as raised / lowered / reaffirmed / initiation against the prior figure using the endpoint's built-in previous_min/max fields, groups by fiscal period, and reports the trajectory. Answers "how has management's own view of the year evolved?" Requires Stocks Basic + Benzinga Corporate Guidance add-on (approx $99/mo). Emits a clear NOT_AUTHORIZED tier caveat when the entitlement is missing.

Scan SEC 8-K disclosures across a single ticker or a watchlist using Massive's pre-parsed disclosure taxonomy. Groups the underlying rows by filing (one 8-K carries N tagged Items), ranks by signal bucket (M&A / Restatement / Material agreement / Regulatory / Leadership change / Capital / Earnings / Corporate housekeeping / Other), and surfaces the highest-signal filings at the top with the supporting text quoted. Use when a PM or analyst asks "what material events hit my names this week?" Requires Stocks Basic. Runs on the free tier.

Aggregate SEC Form 4 insider activity for a ticker over a caller-supplied lookback window, classify each transaction by SEC transaction code and Rule 10b5-1 status, separate signal (conviction buys, discretionary sales) from noise (grants, exercises, tax withholding, 10b5-1 sales), detect cluster buys (>= 2 insiders in a 14-day window worth >= $100k), and emit a sentiment label backed by the underlying dollar flow. Use when a PM or fundamental analyst asks "are insiders buying or selling this name?" Uses Massive's pre-parsed Form 4 endpoint. Requires Stocks Basic. Runs on the free tier.

Diff Item 1A Risk Factors between two 10-K filings for a name using Massive's pre-parsed and taxonomy-classified risk-factor endpoint. Reports categories added, categories removed, and categories where the supporting text materially changed (>= 25% length delta) year-over-year. Groups by primary category so the reader sees the shape of what changed, not a flat diff. Use when a PM, credit analyst, or fundamental researcher asks "what did management add to Item 1A this year?" Requires Stocks Basic. Runs on the free tier.

Measure abnormal returns around a corporate event for one or many tickers. Three input modes pick the output shape automatically: single ticker + single event renders a sell-side note (with t-stat vs that name's reaction distribution); many tickers + one event class renders a cross-section table; many events + many tickers renders aggregate statistics. Supports earnings (Benzinga or SEC EDGAR fallback), dividend changes, and computed volume spikes out of the box. Generalizes earnings-drilldown's PEAD work to any event class.

Run vol-target, fractional Kelly, risk parity, and equal-weight position sizes side-by-side on a basket of tickers. Use when a PM has names they want in the book and asks "how much of each?" The script doesn't pick names or predict returns; it shows what each sizing method gives so the PM can pick the one whose worldview matches their conviction. Requires Stocks Starter.

VaR (historical + parametric), Expected Shortfall, max drawdown, beta, tracking error, position variance contributions, concentration metrics, and worst-N historical stress scenarios for a portfolio. Pairs with portfolio-mark (which marks the book — risk-report tells you what could happen to those marks). Use when a PM, risk officer, or quant needs the full risk picture on a current book.

Screen every pair in a basket for cointegration on daily closes and rank the tradeable ones by spread z-score. Runs the Engle-Granger two-step on log prices, tests the residual with a Dickey-Fuller t-stat against MacKinnon 2010 critical values, estimates the Ornstein-Uhlenbeck half-life of mean reversion, and flags out-of-sample regime shifts via a 70/30 residual std ratio. Emits per-pair hedge ratio, ADF t-stat and p-value bucket, half-life, current z-score, and a stability label. Use when a PM or stat-arb desk has a sector basket and asks "which two names are statistically tethered right now, and how wide is the spread." Requires Stocks Starter.

Rates and credit view via ETF proxies (SHV, SHY, IEF, TLT, TIP, LQD, HYG, AGG). Reports returns across 1/5/20/60/120 day windows, price percentile vs trailing year, HYG-LQD credit spread delta and TLT-IEF duration spread delta, plus HYG-benchmark correlation. Derives a regime label (risk_off, credit_stress, goldilocks, reflation, rate_pressure, neutral). Every equity valuation implicitly assumes something about rates; this skill closes the equity-only gap without needing FRED.

Build a clean, point-in-time, ready-to-backtest OHLCV dataset for a US equity universe across an arbitrary date window. Emits parquet plus a manifest plus an edge-case log, with corporate actions reconciled, survivorship treatment documented, holidays and half-days preserved correctly, and any IPO partial coverage or trading halts flagged. The dataset is the artifact a downstream Python/R/Julia backtester consumes; the rendered summary is the operator-readable companion. First skill in the suite that ships dataset output mode.

Reconcile a position file against splits, dividends, and spinoffs to catch breaks before they hit P&L or T+1 settlement. Use when an operator hands over a CSV of positions and asks "are these right after the recent corporate actions." Runs on a free Massive Basic key.

Surface 24h crypto volatility and microstructure anomalies across a universe (default top 10) as a Bloomberg crypto desk / Cheddar-Flow-for-crypto-style stream. Per-name: realized vol spike (vs 30d distribution), volume anomaly (vs 30d avg), cross-exchange basis (max bid-ask divergence across Coinbase / Binance / Kraken / Bitstamp / Bitfinex), and 24h move z-score. Ranked by composite impact. Real-time on Currencies Business; spot-snapshot accurate within seconds across all paid tiers.

Lightweight watchlist scanner. Takes a comma-separated list of tickers plus a forward window and returns each ticker's earnings status (blackout_imminent, blackout_soon, blackout_extended, just_printed, recent_print, clear, unresolved) with the next/most-recent print date and consensus EPS where available. Use before a trading day to spot which positions are about to print. Earnings-drilldown is the deep dive; this is the batch scan.

Produce a sell-side-grade earnings preview for a single ticker: implied vs realized move, beat/miss history, post-earnings drift, peer reaction, and a one-line take. Use when an analyst, PM, or trader is preparing for a specific company's earnings print. Lite mode runs on Stocks Starter; full mode adds IV crush analysis and needs Options Developer.

Run a quant-style multi-factor backtest on a defined US equity universe. For momentum, value, quality, and low-vol factors, compute decile spreads, information coefficients with t-stats, IC decay curves at 1M/3M/6M/12M forward horizons, single-name attribution at the long and short tails, and the factor correlation matrix. Emits FactSet/Axioma factor research-style table output a buy-side quant would hand to a PM. First skill to exercise the flat-files foundation: a 5-year x top-500 daily aggregates pull is ~80,000 ticker-days, done via a few day-bucket downloads instead of 80,000 REST calls.

Daily macro context. Pulls SPY trend (5 buckets via 20/50/200-day SMA stack), VIX state with percentile rank vs the trailing year, breadth proxy from 11 sector ETFs above their own 50-day / 200-day SMAs, and 20-day relative-strength sector leadership. Combines the four blocks into a single composite regime label (risk_on, risk_off, mixed_risk_on, mixed_risk_off, neutral) with explicit reasons[] so the operator sees the evidence, not just the label. Anchor every research session with this; closest competitor (Jow Dones) leads with the same idea. Use when the day's question is "what's the tape doing right now" or "is this still a risk-on regime."

Surface the day's news events that actually moved a stock. For each notable headline across a watchlist (or the broader market), render a Bloomberg news tape / Benzinga Pro-style stream with sentiment, novelty, and the post-publish price reaction. Ranked by impact, capped at top N (default 15-20). The 6am sell-side morning-note prep workflow.

Surface unusual options activity across a watchlist as a Cheddar Flow / FlowAlgo-style stream. Each notable print rendered as a three-line block with kind (sweep vs block), premium, volume vs avg, volume vs OI (opening vs closing), price relative to NBBO, and inferred direction. Use when a trader is scanning for unusual flow, monitoring smart money, or hunting for actionable setups. Real-time on Options Business; ~15-min delayed on Options Developer.

Build a Bloomberg / CapIQ-style comparable companies set for a subject ticker. Identifies peers via a curated override map (with correlation and SIC fallbacks), pulls current multiples (EV/Sales, EV/EBITDA, P/E) for the subject and peers, renders the comp table with median / mean / 25-75 percentile summary stats, runs a regression-adjusted multiples view that controls for growth and profitability, and surfaces a one-line banker read at the bottom. Use when an analyst or banker is preparing pitch materials, a fairness opinion, or a valuation memo. Requires Stocks Starter for financials.

Mark a book of positions to current fair value and flag any name where the mark is low-confidence (stale, wide-spread, illiquid). Two modes: delayed REST snapshots for EOD reporting, live WebSocket stream for intraday risk. Emits a marked-positions table plus an exception block per flagged mark. Use when an operator hands over a CSV and asks "what's this book worth right now" or "which marks am I not sure about."

Rank a watchlist of tickers by relative strength versus a benchmark (default SPY) across multiple lookback windows (default 5/20/60/120 trading days). Emits per-window RS in basis points, total return, a within-watchlist composite percentile rank, and a trend label per name (stable_leader, improving, deteriorating, stable_laggard, mixed). Use when a PM or trader has a watchlist and asks "rank these by RS vs SPY across week/month/quarter/half." Lightweight watchlist ranker, not a universe-wide factor study. Requires Stocks Starter.