Metraipolis › archive › dc_c2148fdd3729
Long-Horizon Forecasting of Complete Financial Statements with Forma — arXiv:2608.11327
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.11327
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-19T05:39:41.387Z
- SHA-256 of the archived text
c96ef6a4677ae66357f76a11ef25e80b3148cfb51bbe7bc279919ef01e397e60- Stable URL of this page
https://metraipolis.com/documento/dc_c2148fdd3729
How to cite
Long-Horizon Forecasting of Complete Financial Statements with Forma — arXiv:2608.11327 — arXiv (en), https://arxiv.org/abs/2608.11327 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_c2148fdd3729 (Metraipolis archive dc_c2148fdd3729, sha256 c96ef6a4677ae66357f76a11ef25e80b3148cfb51bbe7bc279919ef01e397e60)
BibTeX
@misc{dcc2148fdd3729,
title = {Long-Horizon Forecasting of Complete Financial Statements with Forma — arXiv:2608.11327},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.11327},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_c2148fdd3729; sha256 c96ef6a4677ae66357f76a11ef25e80b3148cfb51bbe7bc279919ef01e397e60},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.11327. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Specialist training beats generalist scale when forecasting financial statements. To our knowledge, no prior work jointly forecasts complete financial statements beyond one year, yet in a discounted-cash-flow valuation most firm value sits past that window. We release ProForma-20Q, a reproducible benchmark for forecasting 78 statement line items 1-20 quarters ahead, for anonymized firms, from past statements and an industry code, scored by change-space $R^2$. On it, Forma, a transformer that reads statements as sets of (account, quarter, value) tuples and maximizes a masked-tuple Gaussian likelihood, beats every competitor we field: classical machine learning, chained gradient boosting, a zero-shot time-series foundation model, and frontier large language models. Its lead widens with horizon, where valuation needs accuracy most, and its Gaussian predictive intervals never under-cover. Forma's forecasts nearly satisfy accounting identities; exact coherence is recoverable at no statistically significant accuracy cost. Its tuple interface supports scenario analysis without retraining, and we show that pinning future revenue paths sharpens the rest of the statement. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-13 in cs.LG). Categorie arXiv: Machine Learning (cs.LG) ; Computational Finance (q-fin.CP). Pagina del lavoro: https://arxiv.org/abs/2608.11327 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-19T05:18:32.456Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-08-13?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).