发表机构
Foresight-AI, Intuit（富睿特远见人工智能实验室，Intuit公司）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
提出Accounting Graph Transformer模型，针对小企业短历史数据完成13项财务KPI的12个月联合预测，在多组测试集上均优于LightGBM等基准模型，为企业财务分析提供集成预测层。
AI 中文摘要
小企业通常仅有12至24个月的会计历史，但其规划和风险工作流程需要对财务报表进行协同预测。本研究针对来自71个月度总账序列的13项损益表、资产负债表、现金流量表及营运资金关键绩效指标（KPIs），开展为期12个月的联合预测研究。我们提出会计图变换器（Accounting Graph Transformer, AGT），该模型将每个总账序列表示为掩码令牌，通过固定会计关系图上的类型注意力机制交换信息，汇集目标特定上下文并与门控三个月近期路径融合。在来自1060家未见过的公司的11993个预测起点上，AGT在三个独立随机种子下的样本加权KPI宏平均绝对误差（MAE）为0.6990±0.0013，而最强基准模型LightGBM的对应值为0.7378±0.0014。在指定随机种子42下，配对公司聚类自举检验得出LightGBM与AGT的差值为0.0395，95%置信区间（CI）为[0.0350,0.0439]。在匹配的种子42对比中，AGT在全部13项KPI上的表现均优于LightGBM、TimeMixer和SOFTS；最终架构消融实验表明，关系注意力、会计拓扑结构及近期路径均能提升验证集和测试集准确率。在2025年1月至5月采样起点的7094家额外未见过的公司上，AGT的MAE为0.7548，而SOFTS为0.7694。一个拥有530万参数的单一模型可生成156个对齐预测，无需针对公司进行特定拟合，为集成规划、流动性及营运资金分析提供了一个预测层。
英文摘要
Small businesses often have only 12-24 months of accounting history, yet planning and risk workflows require coordinated forecasts across financial statements. We study joint 12-month forecasting of 13 income-statement, balance-sheet, cash-flow, and working-capital key performance indicators (KPIs) from 71 monthly ledger series. We introduce the Accounting Graph Transformer (AGT), which represents each ledger series as a masked token, exchanges information through typed attention on a fixed accounting-relation graph, pools target-specific context, and fuses it with a gated three-month recency path. Across 11,993 forecast origins from 1,060 unseen companies, AGT achieves sample-weighted KPI-macro mean absolute error (MAE) $0.6990 \pm 0.0013$ over three independent seeds, compared with $0.7378 \pm 0.0014$ for the strongest baseline, LightGBM. At the pre-specified seed 42, a paired company-clustered bootstrap gives a LightGBM-minus-AGT difference of 0.0395 with 95% confidence interval (CI) $[0.0350,0.0439]$. AGT is best on all 13 KPIs against LightGBM, TimeMixer, and SOFTS in the matched seed-42 comparison, while final-architecture ablations show that relational attention, accounting topology, and the recency path each improve validation and test accuracy. On 7,094 additional unseen companies with origins sampled from January-May 2025, AGT obtains 0.7548 MAE versus 0.7694 for SOFTS. A single 5.3M-parameter model produces 156 aligned forecasts without company-specific fitting, providing one forecasting layer for integrated planning, liquidity, and working-capital analysis.
