← Papers

Firms that pair AI and Big Data with strong governance tend to reduce market uncertainty in crises, but benefits are uneven across industries and hinge on regulation and data maturity; misapplied models or blind automation can amplify volatility.

An Empirical Study on the Impact of the Integration of AI and Big Data on Market Uncertainty in the Context of Economic Turbulence

YiLv Ge · Fetched March 15, 2026 · Economics and Public Policy

semantic_scholar correlational medium evidence 8/10 relevance DOI Source PDF

Strategic adoption of AI and Big Data is associated with materially lower measured market uncertainty during economic turbulence, but the effect depends on industry, regulation and data governance maturity and can reverse if models are biased or over-relied upon.

This research investigates the impact of integrating Artificial Intelligence (AI) and Big Data analytics on market uncertainty specifically during periods of economic turbulence. It addresses the gap in existing literature concerning the empirical evidence of this relationship, particularly during economic downturns. The study employs a mixed-methods approach, combining quantitative analysis of market data with qualitative case studies of firms implementing AI and Big Data solutions. Findings reveal that the strategic integration of AI and Big Data can significantly reduce market uncertainty by enhancing predictive modeling capabilities and enabling proactive risk management strategies. However, the effectiveness of these technologies is contingent upon moderating factors such as industry type, the stringency of the regulatory environment, and the maturity of organizational data governance frameworks. Furthermore, the research identifies potential pitfalls, including algorithmic bias and over-reliance on data-driven insights without adequate human oversight, which can exacerbate uncertainty. The study culminates in a practical framework designed to guide businesses in effectively leveraging AI and Big Data to navigate market volatility. The implications extend to policy recommendations for fostering responsible AI adoption and data utilization to mitigate economic risks.

Summary

Main Finding

The paper provides empirical evidence that integrating AI and Big Data analytics reduces market uncertainty during periods of economic turbulence. The effect is economically meaningful and robust to fixed-effects, 2SLS, and GMM estimators, but is moderated by industry data intensity, regulatory stringency, and firms’ data/governance capabilities. The study also highlights risks—algorithmic bias, over‑reliance on opaque models, and insufficient human oversight—that can counteract benefits.

Key Points

Core result: Higher AI/Big Data adoption is associated with lower market-level and firm-level uncertainty (VIX and stock return volatility), especially in turbulent periods.
Moderators:
- Stronger uncertainty reduction in data-rich, structured-data industries (finance, e‑commerce).
- Weaker effect where data-privacy/regulatory stringency is high.
- Larger benefits for firms with mature IT infrastructure, data literacy, and governance.
Risks and limits: algorithmic bias, model overfitting to historical regimes, black‑box opacity, and potential systemic consequences if unchecked.
Practical output: a framework for firms to leverage AI/Big Data for volatility navigation and policy recommendations for responsible AI adoption.

Data & Methods

Sample and period: Firm-level financial data from Compustat and CRSP, 2010–2023; macro market uncertainty via VIX.
Dependent variables: VIX (market uncertainty) and firm-level stock return volatility (std. dev. of daily returns over one year).
Main AI/Big Data adoption proxies:
- CapEx on software and IT infrastructure scaled by total assets.
- Textual frequency of AI/Big Data keywords in annual reports (10‑K textual analysis).
Qualitative supplements: industry reports (Gartner, McKinsey) and a survey of 200 firms capturing organizational capabilities and perceived AI effectiveness.
Econometric strategy:
- Panel regressions with firm fixed effects.
- Endogeneity addressed with two-stage least squares (2SLS) using external instruments (paper reports an exogenous technology-adoption shock instrument; specifics not detailed in text) and validated via over-identification tests.
- Robustness checks using GMM and alternative uncertainty measures.
Robustness and diagnostics: alternative uncertainty indices, checks for heteroskedasticity/multicollinearity, and sensitivity analyses across industries and regulatory contexts.

Limitations (as reported or implied) - Adoption measures are indirect (capex and keyword counts) and may misclassify true AI capability. - Instrument(s) for 2SLS are not fully described in the provided text—validity depends on instrument exogeneity. - Survey size (200 firms) provides useful context but limits external generalizability. - Results are aggregate; heterogeneous firm behavior and potential selection/survivorship biases warrant caution.

Implications for AI Economics

For theory: Empirically supports that AI/Big Data can reduce informational frictions and volatility; suggests research should incorporate regime shifts and tail-risk environments rather than assuming stationarity.
For empirical work: Encourages improved measurement of AI adoption (beyond capex/text counts), stronger causal identification (natural experiments, policy discontinuities), cross‑country and sectoral comparisons, and analysis of systemic risk channels from widespread algorithmic use.
For firms: Invest in data governance, build internal AI literacy and human-in-the-loop processes, prioritize model interpretability and stress testing to capture non‑stationary shocks.
For policymakers: Balance enabling data-driven innovation with privacy and safety; consider:
- Standards for AI transparency and model risk management (analogous to financial stress tests).
- Regulatory sandboxes to assess real‑world performance in turbulent conditions.
- Incentives for firm-level investments in data governance and workforce retraining.
For macro/market stability: Widespread, well‑governed adoption of AI/Big Data can improve market predictability and resilience, but unmanaged diffusion of opaque models could amplify systemic volatility—monitoring and macroprudential policy design should account for algorithmic commonalities and correlated strategies.

Suggested next research avenues (concise) - Stronger causal designs exploiting exogenous shocks to AI adoption (policy changes, infrastructure rollouts). - Sector-specific analyses of algorithmic trading vs. operational AI impacts. - Long-run studies on whether AI adoption changes the distribution of tail‑risk events. - Measurement work to create validated firm-level AI capability indices.

Assessment

Paper Typecorrelational Evidence Strengthmedium — The paper combines quantitative associations showing lower measured uncertainty where AI/Big Data adoption is higher with qualitative case evidence on mechanisms, which increases plausibility; however, adoption is likely endogenous, confounding and reverse causality are possible, and no clear exogenous instrument or natural experiment is reported to establish causality. Methods Rigormedium — Uses mixed methods appropriately—statistical analysis of market/firm-level data plus detailed firm case studies—and reports robustness checks, but lacks a strong identification strategy (e.g., instrumental variables, differences-in-differences with plausibly exogenous shocks) and the qualitative sample may be small/selected, limiting internal rigor. SampleMarket-level time-series (volatility/uncertainty indices) and firm- or sector-level indicators of AI/Big Data adoption and performance around turbulent periods, supplemented by qualitative case studies of firms from multiple industries (examples include finance and manufacturing) examined for governance practices, regulatory context, and implementation details; exact sample sizes, countries and time periods are not specified in the summary. Themesadoption governance innovation org_design IdentificationAssociational analysis linking measures of AI/Big Data adoption to market-level uncertainty metrics during downturns (time-series/cross-sectional regressions with controls and robustness checks), combined with qualitative case studies for process-level triangulation; no strong quasi-experimental or exogenous variation deployed to claim clean causal identification. GeneralizabilityIndustry heterogeneity: effect varies across sectors (finance vs manufacturing) so results may not generalize uniformly., Selection bias: adopting firms may differ systematically (size, resources, risk tolerance) from non-adopters., Regulatory and institutional context: findings may not hold across jurisdictions with different regulation., Firm maturity and data governance: benefits likely concentrated among firms with advanced data practices, not SMEs or early adopters., Temporal dependence: results observed during specific downturns may not generalize to different kinds of shocks or calm periods., Measurement limitations: proxies for 'AI/Big Data adoption' and 'uncertainty' may not capture all relevant variation.

Claims (11)

Claim	Direction	Confidence	Outcome	Details
Strategic integration of AI and Big Data can significantly reduce market uncertainty during periods of economic turbulence. Market Structure	positive	medium	Market uncertainty (reduction in uncertainty / volatility)	0.18
The reduction in market uncertainty occurs through enhanced predictive modeling capabilities enabled by AI and Big Data. Decision Quality	positive	medium	Predictive modeling performance (as a mediator) and downstream market uncertainty	0.18
AI and Big Data enable proactive risk management strategies that contribute to lowering market uncertainty. Decision Quality	positive	medium	Use of proactive risk management strategies and associated change in market uncertainty	0.18
The effectiveness of AI and Big Data in reducing market uncertainty is contingent upon industry type. Market Structure	mixed	medium	Degree of uncertainty reduction conditional on industry	heterogeneous by industry 0.18
The stringency of the regulatory environment moderates how effectively AI and Big Data reduce market uncertainty. Governance And Regulation	mixed	medium	Market uncertainty reduction conditional on regulatory stringency	moderation by regulatory stringency reported 0.18
The maturity of an organization's data governance framework influences the success of AI and Big Data in lowering market uncertainty. Organizational Efficiency	mixed	medium	Market uncertainty reduction conditional on data governance maturity	moderation by data-governance maturity reported 0.18
Algorithmic bias is a potential pitfall of using AI and Big Data that can exacerbate market uncertainty. Ai Safety And Ethics	negative	medium	Increase in market uncertainty (or risk) attributable to algorithmic bias	identified risk (algorithmic bias can increase uncertainty) 0.18
Over-reliance on data-driven insights without adequate human oversight can worsen market uncertainty. Ai Safety And Ethics	negative	medium	Increase in market uncertainty associated with reduced human oversight	identified risk (reduced human oversight can worsen uncertainty) 0.18
The research produced a practical framework to guide businesses in effectively leveraging AI and Big Data to navigate market volatility. Organizational Efficiency	positive	low	Availability of a practical framework (effectiveness of the framework not demonstrated in the summary)	0.09
The study's implications include policy recommendations to foster responsible AI adoption and data utilization to mitigate economic risks. Governance And Regulation	positive	low	Policy guidance for responsible AI adoption (impact on economic risk mitigation not empirically tested in the summary)	0.09
There is a gap in the existing literature regarding empirical evidence about the relationship between AI/Big Data use and market uncertainty during economic downturns. Research Productivity	null_result	medium	Existence of an empirical evidence gap in the literature	0.18