時系列基盤モデルは炭素価格を予測できるか:EU ETSと中国炭素市場からのリークフリー証拠
Do Time Series Foundation Models Forecast Carbon Prices? Leakage-Free Evidence from the EU ETS and China’s Carbon Markets (原題)
Ruiting Wu, Jafferi Jamaludin, Jeyraj A-L Selvaraj
🤖 gxceed AI 要約
日本語
Chronos-2とTimesFM-2.5を含む時系列基盤モデルをEU ETS・中国全国ETS・湖北・広東の4市場で1〜60営業日先まで検証した。EU ETSではランダムウォークを有意に上回るモデルはなく、湖北ではTSFMが最小誤差を示す一方、全国市場や広東では精度が劣後した。誤差の大小は日次リターンの一時的成分と価格の持続的変動に対応し、先読みバイアスは分解型モデルの誤差を約半分に見せかける。TSFMの実運用には市場別の検証が不可欠と結論づける。
English
Evaluates time series foundation models (Chronos-2, TimesFM-2.5) against benchmarks across the EU ETS, China's national ETS, and Hubei/Guangdong pilots at 1–60 day horizons using leakage-free rolling-origin design. No model beats the random walk in the EU ETS; TSFMs lead in Hubei but underperform in the national market and Guangdong. Gains track transitory return components, losses track persistent price moves, and look-ahead bias roughly halves apparent errors of decomposition models. TSFMs need market-specific validation before deployment.
Unofficial AI-generated summary based on the public title and abstract. Not an official translation.
📝 gxceed 編集解説 — Why this matters
日本のGX文脈において
炭素価格は日本でもGX経済移行債やカーボンプライシング議論の中心であり、価格予測モデルの信頼性評価は国内ETS・炭素税設計の参考になる。AIによる価格シグナル分析の限界を示す点で実務的示唆が大きい。
In the global GX context
Carbon price forecasting underpins transition finance, hedging, and compliance strategy under EU ETS and emerging Chinese markets. The leakage-free benchmark and market-specific caveats inform how AI forecasting tools should be validated before use in disclosure and risk models.
👥 読者別の含意
🔬研究者:炭素価格予測におけるTSFMの限界と先読みバイアスの影響を定量化した貴重なベンチマーク。
🏢実務担当者:AI価格予測ツールを炭素リスク管理や調達判断に使う前に市場別検証が必要と理解すべき。
🏛政策担当者:炭素市場の価格安定性評価やAI予測の規制利用に際し、市場別の精度検証を求める根拠になる。
📄 Abstract(原文)
Time series foundation models (TSFMs) promise accurate zero-shot forecasts, but their value for carbon markets has been examined only in the European Union Emissions Trading System (EU ETS). We evaluate Chronos-2 and TimesFM-2.5 alongside six benchmark and hybrid models in four markets (the EU ETS, China’s national ETS, and the Hubei and Guangdong pilots) at horizons of 1–60 trading days, using a rolling-origin design without information beyond the forecast origin. No model significantly outperforms the random walk in the EU ETS. TSFMs achieve the lowest point and probabilistic forecast errors in the Hubei pilot, but their point forecasts are less accurate than the random walk in the national market and, over the full test period, in Guangdong. In the national market, Chronos-2 errors reach 22.8% at the 60-day horizon and its 80% intervals cover only 62% of outcomes, whereas TimesFM-2.5 attains the lowest quantile loss despite under-coverage. The national-market and Hubei results hold over a common calendar period and for targets postdating the model weights. Across the markets and periods studied, gains are associated with a large transitory component in daily returns and losses with persistent price moves. Look-ahead bias roughly halves the apparent error of decomposition-based models. TSFMs therefore require market-specific validation before deployment.
🔗 Provenance — このレコードを発見したソース
- openalex https://doi.org/10.3390/en19204999first seen 2026-10-23 05:23:41 · last seen 2026-11-06 04:58:18
- semanticscholar https://doi.org/10.3390/en19204999first seen 2026-10-27 05:19:30 · last seen 2026-11-06 05:20:47
🔔 こうした論文の新着を逃したくない方は キーワードアラート に登録(無料・3キーワードまで)。
gxceed は公開メタデータに基づく研究支援データセットです。要約・翻訳・解説は AI 支援で生成されています。 最終的な解釈・検証は利用者が原典資料に基づいて行うことを前提とします。