HOOD forecast model
and what it got wrong.
Robinhood Markets is one of 25 stocks GioModel models. This page is the public record: the inputs, the method, the evidence gates, the validation metrics, and past forecasts graded against what actually happened. Current frozen ranges are the only part held for members.
Company overview
- Company
- HOOD Robinhood Markets
- Sector
- Financials vs XLF
- Market cap
- 83.4B
- Model status
- 2 / 9 evidence gates passed
Reference fundamentals as of Aug 5, 2026: gross margin 91.9%, operating margin 46%, P/E 41.1, P/S 17.16. These are context for reading the model, not model inputs — the admitted inputs are listed below. Price history comes from Nasdaq.
How the HOOD forecast is generated
A probability model: given only information available at the close, it estimates the chance that the raw closing price is higher five sessions later. It outputs a probability, never a price target, and is scored on how well-calibrated that probability is rather than on returns.
- 01Freeze the history
A point-in-time panel of completed HOOD sessions is fixed and hashed — this release froze on Jul 29, 2026. Nothing after that date can influence the numbers, which is what makes a later grade meaningful.
- 02Build inputs that could have existed
Every feature is computed from a strictly prior window and tested for prefix invariance, so a value never changes when later data arrives. The admitted inputs are listed below.
- 03Fit and score through time
Direction classifier candidates are compared inside purged, embargoed walk-forward splits across 685 evaluation origins from Oct 26, 2023 to Jul 22, 2026. The block used to score a model is never the block used to pick it.
- 04Publish, then grade in public
The release states which of the nine gates it passed and which it did not. Thirty days later the frozen numbers appear on the public track record and are graded on the exact trading session the horizon lands on.
What data actually goes in
Not everything a model could use is admitted. Each input is marked with how far it has actually been taken — admitted means it feeds the published forecast; anything else is disclosed rather than quietly implied.
| Category | Input | Status |
|---|---|---|
| Fundamental | Revenue growth | PLANNED |
| Fundamental | Margins | PLANNED |
| Fundamental | Earnings | PLANNED |
| Fundamental | Valuation | PLANNED |
| Technical | Momentum | ADMITTED |
| Technical | Moving averages | ADMITTED |
| Technical | Volatility and volume | ADMITTED |
| Market | Interest rates | PLANNED |
| Market | Sector trends | ADMITTED |
| Market | Market regime | ADMITTED |
| Portfolio | Correlation | PORTFOLIO LAYER |
| Portfolio | Position size | LIVE ARITHMETIC |
| Portfolio | Risk contribution | PORTFOLIO LAYER |
Validation metrics
Purged expanding walk-forward evidence already inspected during model development; not a sealed terminal holdout.
- Evaluation origins
- 685
- Direction accuracy
- 60.9% vs 59.7% always-one-class
- 80% band coverage
- —
- Mean WIS vs baseline
- — lower is better
A historical win means the forecast direction matched the realized direction at the stated horizon. It is not a trading return. Horizons scored: 5 sessions. Fold record: 8 of 11 on brier score versus fold-local prior. Calibration error 4.29%. A win here is a scoring outcome on historical data, never a trade.
Evidence gates: 2 of 9
The Phase B challenger passes 2 of 9 gates on development evidence only. The fold-local prior remains active. The challenger and simulation are research context, not a live trading signal.
Every gate below is reproduced from HOOD's own frozen release artifact — the rule as written before the results were seen, the measured value, and the verdict. Gate numbers are local to this protocol: each model family predeclared its own set, so a G3 here is not a G3 on another model's page. Compare the protocols side by side.
This model is graded out of 9, not out of a single site-wide number. This family still publishes a directional claim, so it keeps the gate that tests it and is graded out of nine. The position-forecast family dropped that gate on 15 August 2026 together with the claim itself, and is graded out of eight — a gate is only ever removed alongside the claim it was there to check.
- G1Brier improvement and its 20-session lower bound are positive — open
Frozen threshold: Point > 0; lower > 0
Observed: +0.004739; 95% lower −0.002580
- G2ROC AUC and its block-bootstrap lower bound clear no-skill — open
Frozen threshold: AUC ≥ 0.52; lower > 0.50
Observed: 0.503030; lower 0.4347
- G3Family-wise circular-shift result — open
Frozen threshold: p < 0.05
Observed: p = 0.6880
- G4Challenger beats frozen Core Ridge robustly — open
Frozen threshold: Point > 0; lower > 0
Observed: +0.003331; lower −0.003398
- G5Calibration slope — open
Frozen threshold: 0.70 to 1.30
Observed: 0.266303
- G6Absolute calibration intercept — open
Frozen threshold: ≤ 0.10
Observed: 0.312902
- G7Five-bin expected calibration error — passed
Frozen threshold: < 5%
Observed: 0.042856 · 4.29%
- G8Outer folds with positive Brier improvement — passed
Frozen threshold: ≥ 60%
Observed: 8 / 11 · 72.7%
- G9Block-length and supported-regime sensitivity — open
Frozen threshold: All lower bounds > 0; no regime skill < −5%
Observed: 10d −0.002417; 20d −0.002580; 40d −0.001978
Historical Evaluation Ledger — Development/OOS Evidence
These origins were evaluated during model development, before the public protocol began. They are out-of-sample within the development panel, but they were not frozen and published ahead of their outcome, so they cannot demonstrate what this model does on genuinely unseen data. Treat them as construction evidence.
The forward record — releases frozen and hashed before the market moved, then graded on their exact resolution session — is kept separately on the Public Forward Track Record. Nothing there has been disclosed yet: the first unlock is 2026-07-29.
Past evaluation origins for HOOD, each graded against the realized move. Only origins older than 0 days appear; anything newer sits behind the same delay as the live release.
| Origin | Horizon | Forecast median | Realized | Direction | Result |
|---|---|---|---|---|---|
| 2026-07-22 | 5 sessions | — | −14.01% | UPactual DOWN | LOSS |
| 2026-07-21 | 5 sessions | — | −12.79% | UPactual DOWN | LOSS |
| 2026-07-20 | 5 sessions | — | −3.66% | UPactual DOWN | LOSS |
| 2026-07-17 | 5 sessions | — | −5.05% | UPactual DOWN | LOSS |
| 2026-07-16 | 5 sessions | — | −4.19% | UPactual DOWN | LOSS |
| 2026-07-15 | 5 sessions | — | −9.57% | UPactual DOWN | LOSS |
| 2026-07-14 | 5 sessions | — | −6.25% | UPactual DOWN | LOSS |
| 2026-07-13 | 5 sessions | — | −9.63% | UPactual DOWN | LOSS |
| 2026-07-10 | 5 sessions | — | −10.73% | DOWNactual DOWN | WIN |
| 2026-07-09 | 5 sessions | — | −7.90% | UPactual DOWN | LOSS |
The Jul 29, 2026 HOOD release is now public in full on the track record, graded on its exact resolution sessions. That delay is the whole product boundary: a forecast that is already thirty days old cannot be traded on profitably, but it can be checked — which is what makes the record worth reading. See every release and its grade.
NO-GO does not mean the research contains no information. It means GioModel refuses to convert incomplete evidence into a trading recommendation. The HOOD model publishes its ranges, probabilities and diagnostics to members, and publishes every gate it has not passed to everyone — but it does not claim to have earned directional deployment, and the deployment gate is independent of payment. Nothing on this page is investment advice, a price target, or a recommendation to buy or sell Robinhood Markets. Read the full method and the research disclaimer.