From Compute to Code
Market Value Creation and Destruction Across Three Generative-AI Shocks in US Equities
DOI:
https://doi.org/10.51094/jxiv.6987キーワード:
generative artificial intelligence、 event study、 firm value、 creative destruction、 software、 DeepSeek、 agentic AI抄録
Generative artificial intelligence is often described as a source of creative destruction, but the market evidence comes from single events studied one at a time. This article compares three shocks in the same framework: the release of ChatGPT (30 November 2022), the DeepSeek-R1 repricing (27 January 2025) and the release of agentic workflow plugins (30 January 2026). Using point-in-time S&P 500 membership, market-model abnormal returns, split-consistent market values from SEC filings and the exclusion of firms that released earnings inside the event window, the study measures how much market value each shock created and destroyed, where it fell, and how concentrated it was. The ChatGPT release produced no unusual reallocation in its first week (2.6% of sample value; percentile 68.3 of 224 placebo dates). DeepSeek-R1 reallocated 4.0% (percentile 95.1), with losses in compute hardware (−7.4% on days 0–1) and power infrastructure. The agentic plugins reallocated 5.6% (percentile 99.6), with losses in software (−9.4%) and information and professional services (−11.4%) and gains in the rest of the index. The locus of destruction thus moved from the suppliers of compute to the sellers of codified knowledge work. Losses were more concentrated than gains in all three events, contrary to the pre-specified hypothesis, although this asymmetry is not unusual relative to placebo dates. Firm-level reactions reversed between the 2025 and 2026 shocks.
利益相反に関する開示
著者には申告すべき金銭的な利益相反はない。本論文の作成に使用した生成AIツール(Claude)は、本研究で分析する第3のイベントの製品を公開したAnthropic社が開発したものである。同社は本研究に関与しておらず、資金提供やレビューも行っていない。ダウンロード *前日までの集計結果を表示します
引用文献
Akcigit, U. (2026). Creative destruction in economic growth. Scandinavian Journal of Economics, 128(3), 509–541. https://doi.org/10.1111/sjoe.70037
Beckmann, L., & Hark, P. F. (2024). ChatGPT and the banking business: Insights from the US stock market on potential implications for banks. Finance Research Letters, 63, 105237. https://doi.org/10.1016/j.frl.2024.105237
Brown, S. J., & Warner, J. B. (1985). Using daily stock returns: The case of event studies. Journal of Financial Economics, 14(1), 3–31. https://doi.org/10.1016/0304-405X(85)90042-X
Campbell, J. Y., Lo, A. W., & MacKinlay, A. C. (1997). The econometrics of financial markets. Princeton University Press.
DeepSeek-AI. (2025). DeepSeek-R1: Incentivizing reasoning capability in LLMs via reinforcement learning. arXiv:2501.12948. https://doi.org/10.48550/arXiv.2501.12948
Dong, L., Liu, J., Yang, J., & Zhang, X. (2025). The deployment of general large language models and corporate value: Evidence from the stock market. Finance Research Letters, 86, 108625. https://doi.org/10.1016/j.frl.2025.108625
Eisfeldt, A. L., Schubert, G., & Zhang, M. B. (2023). Generative AI and firm values (NBER Working Paper No. 31222). National Bureau of Economic Research. https://doi.org/10.3386/w31222
Fama, E. F., & French, K. R. (1993). Common risk factors in the returns on stocks and bonds. Journal of Financial Economics, 33(1), 3–56. https://doi.org/10.1016/0304-405X(93)90023-5
Fortune. (2026, February 6). Anthropic's Claude triggered a trillion-dollar selloff. https://fortune.com/2026/02/06/anthropic-claude-opus-4-6-stock-selloff-new-upgrade/
Jung, H., Sharma, A., & Nicolau, J. L. (2026). GenAI in tourism: Who wins, who loses? Tourism Management, 114, 105357. https://doi.org/10.1016/j.tourman.2025.105357
Kogan, L., Papanikolaou, D., Seru, A., & Stoffman, N. (2017). Technological innovation, resource allocation, and growth. Quarterly Journal of Economics, 132(2), 665–712. https://doi.org/10.1093/qje/qjw040
Kolari, J. W., & Pynnönen, S. (2010). Event study testing with cross-sectional correlation of abnormal returns. Review of Financial Studies, 23(11), 3996–4025. https://doi.org/10.1093/rfs/hhq072
MacKinlay, A. C. (1997). Event studies in economics and finance. Journal of Economic Literature, 35(1), 13–39.
Norbäck, P.-J., & Persson, L. (2024). Why generative AI can make creative destruction more creative but less destructive. Small Business Economics, 63(1), 349–377. https://doi.org/10.1007/s11187-023-00829-4
OpenAI. (2022, November 30). Introducing ChatGPT. https://openai.com/blog/chatgpt
Qian, Z., Peng, J., & Li, J. (2026). The impact of generative AI announcements on suppliers: Evidence from the stock market. Production and Operations Management, 35(5), 2037–2053. https://doi.org/10.1177/10591478251398333
Schumpeter, J. A. (1942). Capitalism, socialism and democracy. Harper & Brothers.
Shehzad, K., Ali, F., & Zaman, U. (2026). AI rivalry and tail-risk contagion: Evidence from DeepSeek's creative destruction shock. Research in International Business and Finance, 86, 103378. https://doi.org/10.1016/j.ribaf.2026.103378
Wahyono, B., Rapih, S., & Boungou, W. (2023). Unleashing the wordsmith: Analysing the stock market reactions to the launch of ChatGPT in the US education sector. Finance Research Letters, 58, 104576. https://doi.org/10.1016/j.frl.2023.104576
Wu, Q., Zhuang, Q., Liu, Y., & Han, L. (2024). Technology shock of ChatGPT, social attention and firm value: Evidence from China. Technology in Society, 79, 102756. https://doi.org/10.1016/j.techsoc.2024.102756
Yang, X. (2025). AI competition and firm value: Evidence from DeepSeek's disruption. Finance Research Letters, 80, 107447. https://doi.org/10.1016/j.frl.2025.107447
Zeng, L. (2026a). Low-cost AI and the value of compute scarcity: Evidence from the DeepSeek shock. Economics Letters, 268, 113160. https://doi.org/10.1016/j.econlet.2026.113160
Zeng, L. (2026b). Narrative premia at a global LLM shock: Cross-market repricing of corporate disclosure in China and the United States. Emerging Markets Finance and Trade. https://doi.org/10.1080/1540496X.2026.2746475
ダウンロード
公開済
投稿日時: 2026-10-27 14:57:46 UTC
公開日時: 2026-11-16 06:36:29 UTC
ライセンス
Copyright(c)2026
Merlet Echavarria, Alfredo Agustín
この作品は、Creative Commons Attribution 4.0 International Licenseの下でライセンスされています。
