Datasets:
You need to agree to share your contact information to access this dataset
This repository is publicly accessible, but you have to accept the conditions to access its files and content.
These files are model OUTPUTS derived from Compustat via WRDS under The University of Texas at Austin's academic subscription. They are released under the Forma Non-Commercial Research Licence (WRDS-Conditioned) v1.0 — see LICENSE.md. Access requires that you hold your own current Compustat/WRDS licence, use the files for non-commercial academic research only, and do not redistribute them. No raw Compustat values are distributed here; rebuilding the ground truth requires your own WRDS access.
Log in or Sign Up to review the conditions and access this dataset content.
Forma ProForma-20Q Forecasts in USD ($M)
Probabilistic forecasts of complete quarterly financial statements for U.S. public firms, denominated in millions of USD, at horizons of 1–20 quarters.
This is the dollar-space companion to the standardized-space release. Forma internally predicts a per-account, per-quarter asinh z-score; that is the right space for scoring and the wrong space for almost everything else. This dataset inverts the full normalization chain so the forecasts can be merged against Compustat, analyst forecasts, or prices without reimplementing the pipeline.
| Rows | 472,695,966 (0 null) |
| Firms × accounts × horizons | ~1.17M firm-quarters × 78 accounts × 20 horizons |
| Origin quarters | 2008Q1 – 2024Q3 (targets to 2029Q3) |
| Size | 12.43 GB, Hive-partitioned by target (78 partitions, ~169 MB each) |
| Model | Forma, Gaussian β-NLL, transformer-only — 5-seed equal-weight mixture |
Quick start
import pandas as pd
# one account only -- 169 MB, not 12.43 GB
fc = pd.read_parquet("hf://datasets/forma-lab-mccombs/forma-usd-forecasts/target=revtq/")
fc = fc[fc.forecast_horizon == 4]
fc[["firm_id", "target_quarter", "pred_p05_musd", "pred_p50_musd", "pred_p95_musd"]].head()
Schema
| column | meaning |
|---|---|
firm_id |
gvkey, zero-padded to 6 |
quarter |
forecast origin (q0), calendar quarter end |
target_quarter |
quarter being forecast = quarter + forecast_horizon |
target |
Compustat-style account name (the partition key) |
forecast_horizon |
1–20 quarters |
pred_p05/p25/p50/p75/p95_musd |
predictive quantiles, $M |
pred_mean_musd |
predictive mean, $M |
Use pred_p50_musd as the point forecast. The mean is exact but is not the
median — the gap is the skew, and it is large in the tail (see below).
Why quantiles and not a standard deviation
The back-transform V = sinh(a + b·Z)·S/k is strictly monotone, so quantiles map
through it exactly and interval coverage is identical in z-space and dollars.
Moments also have closed forms, but the second moment carries exp(2b²) with
b = σ_reg·σ_model, whose p99 is 3.11 and max 5.69 — so exp(2b²) reaches 1e28.
A Gaussian in z-space is lognormal-tailed in dollars; a standard deviation
would be meaningless for the top few percent of cells and would silently dominate
any pooled statistic. It is deliberately not shipped.
Three traps when merging
- Join on
target_quarter, notquarter.quarteris the origin. - 25 of 78 targets are not Compustat columns. 20 are quarterly flows derived
by differencing YTD items (
oancfq←oancfy, …); 5 are composites. Merging the YTD column directly is silently wrong — Q4 medianoancfq/oancfy= 0.308. Use the suppliedbuild_usd_truth.py. - Filter
forecast_horizonor your panel multiplies ~10× (median 10, max 20 origins forecast the same target quarter).
Full recipe, including I/B/E/S information-set alignment (rdq lags the fiscal
quarter end by a median of 37 days) — see FORECAST_USD_RELEASE.md.
Calibration
Inherited from the model and unchanged by the transform: the 90% interval is close to nominal (0.905–0.939 measured), the central 50% interval is empirically over-wide (~0.58–0.69 vs 0.50). A monotone transform cannot alter this — it is a property of Forma, not of the conversion.
Related releases
| Model code (no weights) | forma-lab-mccombs/forma-release (Apache-2.0) |
| Benchmark + builder | forma-lab-mccombs/proforma-20q (Apache-2.0) |
| Standardized-space forecasts + sample mask | forma-lab-mccombs/proforma-20q-artifacts (same licence, gated) |
| Trained weights + regularization stats | forma-lab-mccombs/forma (non-commercial, no WRDS condition) |
Licence and data provenance
Released under the Forma Non-Commercial Research Licence (WRDS-Conditioned)
v1.0 (LICENSE.md) — not a CC licence. Non-commercial academic research
only; you must hold your own current Compustat/WRDS licence; no
redistribution; attribution required.
CC BY-NC was considered and rejected: it restricts commercial use but says nothing about holding a WRDS licence, which is the actual constraint here, and CC's "no additional restrictions" term makes bolting that on ambiguous.
These are model outputs, not Compustat data — no raw Compustat values are
distributed. Access is gated for a concrete reason: combined with the
regularization statistics and the standardized-space forecasts, the dollar values
permit closed-form reconstruction of the scale deflator (|ltq| + |seqq| + 1e-3)
via S = V·k / sinh(μ + σ·z) — measured at a median error of 0.13%, within 1%
for 100% of firm-quarters once the ~1,120 estimates each firm-quarter yields are
averaged. Omitting the deflator column does not prevent this; gating does.
Derived from Compustat under UT Austin's academic subscription. S&P Global Market Intelligence retains all rights in the underlying Compustat data. Nothing here is investment advice.
Citation
Cite the companion paper and this dataset; see CITATION.cff in
forma-release.
- Downloads last month
- -