AI BENCHMARK PROFILE
FinVerse
FinVerse is a finance-domain time-series forecasting benchmark with 116,897 series (171.1M observations), selecting 60,232 as evaluation targets. It defines 11 metric families (78 metrics) tailored to each series' economic meaning.
- Released
- 2026-08-04
- Readiness
- Paper only
- Primary field
- Finance & Economics
Why it matters
Generic forecasting benchmarks use uniform error metrics that may not align with financial decision objectives. This benchmark aims to provide domain-aware evaluation that better reflects real-world utility.
Motivation
As time-series foundation models have emerged, the need for benchmarks that can evaluate their forecasting ability in meaningful ways has become increasingly important.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.