← Papers

The AI boom reshapes, but does not uniformly strengthen, global market linkages: overall connectedness fell modestly after ChatGPT's launch, with AI stocks losing their initial role as net shock transmitters while the S&P 500 stayed the primary source of spillovers.

Artificial Intelligence and Financial Market Connectedness: Evidence from AI-Related Equities, Cryptocurrencies, and Global Assets

Shigeyuki Hamori · May 06, 2026 · FinTech

openalex correlational medium evidence 7/10 relevance DOI Source PDF

Using daily TVP-VAR estimates from 2021–2025, the study finds that overall financial connectedness modestly declined after the public release of ChatGPT in November 2022 while the structure of spillovers changed—AI equities moved from net shock transmitters toward lesser roles, whereas the S&P 500 remained the dominant source of spillovers.

The rapid expansion of artificial intelligence (AI), particularly with the rise of generative AI technologies, has attracted increasing attention in financial markets. This study examines how the recent AI boom relates to changes in the interconnectedness of global financial markets. Using daily data from January 2021 to December 2025, we analyze spillover dynamics among AI-related equities, cryptocurrencies, and traditional financial assets within a time-varying parameter vector autoregression (TVP-VAR) framework. Our findings indicate that the emergence of generative AI is not associated with a uniform increase in financial connectedness. Instead, the overall level of connectedness declines modestly following the public release of ChatGPT by OPENAI in November 2022, while the structure of spillovers undergoes significant changes. In particular, AI-related equities initially act as net transmitters of shocks, but their relative importance diminishes over time. In contrast, broader equity markets, proxied by the S&P 500, remain the dominant source of spillovers throughout the sample period. These results are robust to alternative model specifications, including different lag lengths and forecast horizons. Overall, the findings suggest that the impact of AI on financial markets is better understood as a structural transformation of interconnectedness rather than a simple intensification of linkages. This study contributes to the literature by providing new evidence on how technological innovation reshapes financial spillover networks and highlights the importance of considering both the level and structure of connectedness in assessing systemic risk.

Summary

Main Finding

The rise of generative AI (peaking around the public release of ChatGPT in Nov 2022) did not produce a uniform intensification of financial market linkages. Using daily data from Jan 2021–Dec 2025 and a TVP-VAR spillover framework, the study finds a modest decline in overall connectedness after Nov 2022 but large structural shifts in how shocks propagate: AI-related equities start as net transmitters of shocks but lose relative importance over time, while the broader equity market (S&P 500) remains the dominant source of spillovers across the whole sample. Results are robust to alternative lag lengths and forecast horizons.

Key Points

Sample and assets: daily observations Jan 2021–Dec 2025 covering AI-related equities, cryptocurrencies, and traditional financial assets (including S&P 500 as the broad equity proxy).
Methodology: time-varying parameter VAR used to track dynamic spillovers among asset groups.
Aggregate connectedness: overall level of connectedness falls modestly following the public release of ChatGPT (Nov 2022).
Structural change: the pattern and direction of spillovers change substantially — AI equities initially act as net shock transmitters but their influence weakens over time.
Dominant transmitter: the S&P 500 remains the primary net transmitter of shocks throughout the period.
Robustness: findings hold across alternative model specifications (different lag lengths and forecast horizons).

Data & Methods

Data: daily returns (Jan 2021–Dec 2025) for three asset groups — AI-related equities, cryptocurrencies, and traditional financial assets (with the S&P 500 used as the broad equity benchmark).
Econometric framework: time-varying parameter vector autoregression (TVP-VAR) to capture evolving dynamics and account for nonstationary spillover patterns.
Spillover measurement: connectedness inferred from forecast-error variance decompositions derived from the TVP-VAR (the study adopts the dynamic connectedness approach commonly used in the literature to obtain total, directional, and net spillovers).
Event focus: comparison of connectedness and network structure before and after the public release of ChatGPT (Nov 2022).
Robustness checks: alternative lag lengths and forecast horizons tested to ensure stability of main results.

Implications for AI Economics

Rethinking "AI boom → greater market linkage": technological shocks from AI can reconfigure how risks transmit without necessarily increasing aggregate connectedness; policy and research should distinguish level vs. structure of connectedness.
Systemic risk monitoring: regulators should incorporate dynamic network measures (directional/net spillovers) rather than relying solely on aggregate connectedness indices, since dominant transmitters (e.g., broad equity markets) can maintain systemic influence even if sectoral transmitters (AI equities) ebb.
Asset pricing and portfolio management: investors should consider time-varying transmitter/receiver roles — AI-sector exposures may offer different diversification benefits over time as their spillover role declines.
Market structure and financial innovation: the weakening of AI equities as transmitters may reflect faster information incorporation, changing investor beliefs, or market depth/liquidity evolution; researchers should investigate mechanisms (liquidity, attention, firm fundamentals).
Crypto vs. traditional assets: coexistence of structural shifts suggests heterogeneous responses across asset classes — risk models and stress tests should allow for asymmetric, time-varying links between crypto and conventional markets.
Directions for future research: causal identification of the mechanisms driving the structural reshaping (e.g., investor flows, news/attention, corporate investment), higher-frequency microstructure analysis, and cross-country/global spillover heterogeneity.

Assessment

Paper Typecorrelational Evidence Strengthmedium — The study uses a credible time-varying parameter VAR framework on high-frequency (daily) market data and reports robustness checks across lag lengths and forecast horizons, which supports the descriptive stability of results; however, it remains observational without a clear causal identification strategy and is vulnerable to confounding from concurrent macro, policy, or market-specific events and to choices in how 'AI-related equities' are defined. Methods Rigormedium — Application of TVP-VAR is appropriate for time-varying spillover analysis and the paper reports specification robustness tests, but the methods lack explicit strategies to address endogeneity or alternative drivers (e.g., macro shocks, monetary policy, COVID/residual pandemic effects), limited discussion of variable selection and multiple-testing, and no quasi-experimental design to support causal claims. SampleDaily financial market data from January 2021 to December 2025 covering AI-related equities (a constructed proxy/index of firms tied to AI/generative AI), cryptocurrencies, and traditional financial assets including the S&P 500; analyzed via a time-varying parameter VAR to estimate dynamic spillovers and connectedness measures. Themesinnovation governance GeneralizabilityFindings depend on how 'AI-related equities' are defined and may not hold under alternative stock selection or weighting rules., Sample period (2021–2025) includes several major global shocks (pandemic aftermath, rate cycles, geopolitical events) that may confound attribution to AI-related developments., Daily-frequency market data may not capture intraday dynamics or structural microstructure effects., Cryptocurrency market idiosyncrasies (extreme volatility, regulatory changes) limit transferability to traditional asset classes., Results speak to financial-market interconnectedness and systemic risk, not real-economy outcomes like productivity or labor markets., Global coverage may be uneven and results could be driven by dominant markets (e.g., US), limiting geographic generalizability.

Claims (10)

Claim	Direction	Confidence	Outcome	Details
This study uses daily data from January 2021 to December 2025 to analyze spillover dynamics among AI-related equities, cryptocurrencies, and traditional financial assets within a time-varying parameter vector autoregression (TVP-VAR) framework. Market Structure	null_result	high	spillover dynamics / connectedness among asset classes	0.5
The emergence of generative AI is not associated with a uniform increase in financial connectedness. Market Structure	null_result	high	level of financial connectedness	0.3
The overall level of connectedness declines modestly following the public release of ChatGPT by OpenAI in November 2022. Market Structure	negative	high	aggregate connectedness level	0.3
The structure of spillovers undergoes significant changes over the sample period. Market Structure	mixed	high	structure/topology of spillover network	0.3
AI-related equities initially act as net transmitters of shocks. Market Structure	positive	high	net directional spillovers (net transmitter status)	0.3
The relative importance of AI-related equities as shock transmitters diminishes over time. Market Structure	negative	high	relative contribution of AI equities to spillovers	0.3
Broader equity markets, proxied by the S&P 500, remain the dominant source of spillovers throughout the sample period. Market Structure	positive	high	dominance in net spillover contributions	0.3
These results are robust to alternative model specifications, including different lag lengths and forecast horizons. Market Structure	null_result	high	stability of connectedness findings across model specifications	0.3
The impact of AI on financial markets is better understood as a structural transformation of interconnectedness rather than a simple intensification of linkages. Market Structure	mixed	high	nature of change in financial interconnectedness (structural transformation vs. intensification)	0.3
The study provides new empirical evidence that technological innovation (specifically generative AI) reshapes financial spillover networks and highlights the importance of considering both the level and structure of connectedness in assessing systemic risk. Market Structure	mixed	high	reshaping of spillover networks; relevance for systemic risk assessment	0.3