Alphanume Learn

Theory is cheap.Run the code.

An interactive course on quantitative trading, taught in real Python against real market data. Lessons are read and run in your browser: no setup, no videos, no toy datasets.

Free. No account needed.

earnings_edge.py

Python · runs in your browser

Sample event history bundled inline, editable before you run it. In the course, the same numbers come live from the Alphanume API.

Every lesson is a working session.

No lectures to sit through. A lesson states a claim, hands you the data, and grades what your code prints. These fragments are from the earnings module's worked example.

1 · Read the argument

Short prose that states a testable hypothesis before any data is touched, so the data can't seduce you later.

“For large-cap names, the earnings move implied by the options market is, on average, larger than the move that gets realized.”

2 · Write and run the code

An editor in the page, wired to live market data from the Alphanume API. You measure the claim yourself.

events["spread"] = (
    events["implied_move_pct"]
    - events["realized_abs_move_pct"]
)
print(f"Mean: {events['spread'].mean():.2f}pp")

3 · Check the result

Output is graded against the expected result, and quizzes attack the reasoning, not the vocabulary.

Mean: 1.34pp

✓ Output matches expected

The whole syllabus, up front.

12 modules · 66 lessons · ~18 hours

Every lesson listed, every description real. One course, taken in order: foundations and tooling first, then volatility, earnings, index structures, corporate events, and the machinery to run it all.

Twelve modules from first principles to a running book: volatility, earnings, 0-DTE, dilution, SPACs, dividends, momentum, and the machinery to trade them systematically.

How Markets Create Repeatable Edges

01Price as Consensus: Information vs. ExpectationEvery price move is one of two things: new information or a shift in expectation. Why only one of them is tradeable systematically.02What Counts as an Edge: Mechanism-First ThinkingStructural vs. statistical edges, and who is forced to act: index trackers, locked-up insiders, redemption mechanics, desperate CFOs.03Recurring, Dated, DisclosedThe three-part test that turns market noise into an event class, and why queryable is the load-bearing word.04The Four-Step Research LoopHypothesis, data, measurement, attack. Stated once, precisely, because the next eleven modules run it repeatedly.05Knowledge Check: EdgesA short quiz covering the foundations module.

The Quant's Toolkit

06Setting Up Python for ResearchA clean, reproducible Python environment — and why most setups quietly rot.07Your First Research DatabaseWhy flat files stop scaling, and how to stand up a SQL database for market data.08Git and GitHub for Research CodeVersion control as a research journal: commits, branches, and not losing a month of work.09Anatomy of a Market-Data APIThe {count, data} envelope, auth, rate limits, and the one line beginners skip and later regret.10From JSON to DataFrame in Five LinesThe ingestion skeleton reused in every later exercise: json_normalize, date parsing, sort.11Knowledge Check: ToolkitA short quiz covering the toolkit module.

Research Methods: How to Trust a Backtest

12Event Windows and Study DesignDefining day zero, choosing windows, and what a population study is and is not.13Abnormal vs. Raw ReturnsWhy 'it went down 10%' means nothing in a month the market fell 12%, with a small worked example.14Survivorship, Delisting, Look-AheadThe three ways backtests lie, and building a point-in-time universe that doesn't.15Reading a Result HonestlyTails, regime concentration, sample size, costs. The permanent checklist later modules invoke by name.16Knowledge Check: MethodsA short quiz covering the research methods module.

Volatility I: The Volatility Risk Premium

17What an Option Price Actually SaysImplied vol as a forward-looking consensus, realized vol as delivered reality, and the premium between them.18Expensive Compared to What?The IV/HV ratio and spread, cross-sectional ranks, and pulling today's richest names.19Expensive for This Name: IV RankIs 63 rich or normal? Rank vs. percentile against the trailing 52-week band, walked on a real ticker.20Vol-of-Vol: How Rough Is the RideStability of vol as a sizing dial: steady names for income, unstable names for snapbacks.21Stacking the Filters: From Coin-Flip to EdgeThe conditioned premium screen: what each filter adds, built end to end.22Knowledge Check: Volatility IA short quiz covering the volatility risk premium module.

Volatility II: Earnings and Scheduled Catalysts

23The Implied Earnings Move: Straddles as ForecastsHow the pre-earnings ATM straddle encodes an expected move, and why magnitude beats direction as a research question.24A Worked Example: Implied vs. RealizedFrom raw earnings-move history to a testable hypothesis, step by step.25Chronic Overpricers: Per-Name Track RecordsSome names overprice every quarter. Hit rates, running averages, and the ranked tally.26Building the Pre-Earnings ScreenCombine track record and current vol context into a watchlist ahead of a reporting week.27Knowledge Check: EarningsA short quiz covering the earnings module.

The Index Game: SPX, 0-DTE, and Regimes

280-DTE Options and the Daily Strike BandWhat same-day options are, what the 10:30 AM strike band is, and why 'never revised' makes the history honest.29Condors, Butterflies, and VerticalsThree containment structures placed off two published strikes, and what the wing parameter costs you.30Measuring Containment: Backtesting the BandHow often does SPX close inside the band? Rolling containment as a health metric.31Band Width as a Volatility Time SeriesWidth as a VIX-lite: percentile regimes, z-scores, and width mean-reversion.32Skew, Drift, and Regime GatesAsymmetry reads, midpoint drift as a slow lean, and the risk-regime flag as the on/off switch.33Knowledge Check: The Index GameA short quiz covering the index module.

Event-Driven I: Dilution and the Short Side

34The Last-Ditch Financing LadderWhy dilution is a one-way message: no other lender said yes, holders get watered down, and management is selling.35S-1 to EFFECT: How New Shares Reach the TapeRegistration vs. effectiveness, the two punishment points, and why not every S-1 is dilutive.36The Mechanics of Selling ShortLocates, borrow fees as daily carry, recalls and buy-ins, dividend liability. The chapter that keeps you out of trouble.37The Dilution StudyPull recent dilutive filings, size the offerings against market cap, and read the drift evidence with effectiveness as the exit clock.38ATM Programs and Shelf RegistrationsThe slow-drip version: how at-the-market programs sell into strength for quarters at a time.39Toxic Financing and Failure ModesDeath-spiral converts, plus the collected failure modes across every dilution flavor.40Knowledge Check: DilutionA short quiz covering the dilution module.

Event-Driven II: De-SPACs, Defaults, and Distress

41The SPAC Machine: Incentives and the ClockWhy sponsors get paid for closing any deal, redemption mechanics, and what that implies for the post-merger cohort.42Fading the CohortBuild the post-completion de-SPAC basket, hold it against an index leg, measure the drift.43Lock-Up Expirations: Supply on a CalendarThe purest calendar event: dates written into the prospectus months ahead, and how you'd source them.44Corporate Defaults: Distress as an Event ClassCredit-like equity behavior without CDS: post-default drift, and keeping longs out of distressed names.45Knowledge Check: De-SPACs and DistressA short quiz covering the de-SPACs and distress module.

Income and Flows: Dividends and Momentum

46Who Really Pays the DividendEx-date mechanics, the drop ratio, why naive capture is a coin flip, and what makes it stop being one.47The Capture Screen: Recovery Odds and the CalendarThe three-stage screen: forward calendar, give-back filter, recovery filter, yield rank.48Why Winners Keep WinningCross-sectional momentum as a documented factor: underreaction, herding, and flows that chase reconstitution.49Replicating the Basket: Rank Tilts and DriftPull the monthly ten, weight by rank slot, measure post-rebalance continuation.50Knowledge Check: Income and FlowsA short quiz covering the income and flows module.

Alternative Data and Daily Screening

51Attention as Data: Wikipedia ViewsPage views, 30-day z-scores, anomaly days, and attention as signal vs. attention as context.52Filing Bursts: Activity Before AnnouncementsCompanies interact with EDGAR more right before something happens. Screen for the spikes.53The Movers Slate: Volatility Screens, Not Crystal BallsThe daily slate, the null-until-close wrinkle, and the honest hit-rate framing.54Slicing Every Signal by SectorJoin sector classification onto any screen and watch how differently sectors behave.55Knowledge Check: Alternative DataA short quiz covering the alternative data module.

Portfolio, Risk, and Temperament

56From Single Event to PortfolioCombining event signals, overlapping signals on one name, and rebalancing cadence.57Sizing, Concentration, and CorrelationPosition limits, sleeve capital allocation, and when uncorrelated sleeves stop being uncorrelated.58The Negative-Skew P&L PathSqueezes, crowded borrow, long quiet stretches punctuated by sharp drawdowns. Temperament as a position.59Gating the Whole BookWire the regime flag and rolling containment into a single sizing layer over everything you've built.60Knowledge Check: Portfolio and RiskA short quiz covering the portfolio and risk module.

Automation and Agents

61Pull, Rank, Format, SendThe production skeleton: one flat file, no framework, a signal that reaches you on its own.62Making the Signal Reach YouSMTP, app passwords, schedulers, UTC gotchas. The take-home recipe for putting a script on a server.63Idea Generation With ClaudeTasking a model like a junior analyst instead of a slot machine.64From Idea to Code, With GuardrailsLetting an agent write research code without letting it grade its own homework.65Capstone: Build and Defend Your Own Event StudyPick any dataset, run the four-step loop end to end, and grade yourself against the honest-results checklist.66Final AssessmentA cumulative, reasoning-heavy check drawing on every module in the course.

Taught from a public track record.

This course is written by the quant behind Alphanume Research and The Quant Galore, where the research is published in the open: hypothesis, data, code, result. Some ideas survive the testing and some don't, and both outcomes get published. The lessons run the same loop, systematized.

Nothing in the course asks to be taken on faith. Every claim is a thing you run.

The first lesson is free and takes about ten minutes.

Read the argument, run the code, and decide with evidence. That's the whole method; it starts working on you in lesson one.