|
on Financial Markets |
| By: | Onur Polat (Institute of Informatics, Hacettepe University, Beytepe Campus, 06800 Çankaya, Ankara, Türkiye); Yuxiang Ye (School of Economics and Finance, University of the Witwatersrand, Johannesburg, South Africa); Rangan Gupta (Department of Economics, University of Pretoria, Private Bag X20, Hatfield 0028, South Africa) |
| Abstract: | This study examines whether decomposing the geopolitical risk (GPR) index into oil- and non-oil-related components yields incremental explanatory power for U.S. industry stock market volatility. Using daily conditional volatilities derived from a GJR-GARCH model with time-varying skewness and kurtosis (GJR-SK) across 49 Fama-French industries, we estimate autoregressive specifications with exogenous inputs (ARX). Each model incorporates rich macro-financial controls, including market portfolio volatility, the Aruoba-Diebold-Scotti business conditions index, and the shadow-rate-augmented federal funds rate. Through a sequence of properly nested F-tests, we address three core empirical questions: (i) whether non-oil GPR adds predictive content beyond the macro baseline; (ii) whether oil-specific GPR provides incremental information once non-oil GPR is controlled for; and (iii) whether a geographic decomposition across eight oil-producing regions enhances explanatory power beyond aggregate metrics. We find that non-oil GPR is incrementally significant in 35 of 49 industries. Conditional on non-oil GPR, aggregate oil-related risk remains incrementally significant in 23 industries, while its regional breakdown increases significance to 29 industries. Estimated coefficients demonstrate marked sign heterogeneity: for oil-intensive sectors, non-oil GPR reduces volatility while oil-specific risk elevates it, supporting a commodity-price information-resolution channel; for non-oil-intensive sectors, the pattern reverses, consistent with the classical risk-aversion channel of Bloom (2009). Regionally, geopolitical risk originating in the Middle East, Venezuela, and Asia systematically amplifies industry volatility, whereas risk originating in Russia and Africa exerts a stabilizing, volatility-dampening effect. |
| Keywords: | stock price volatility, industry sectors, geopolitical risk, oil geopolitical risk, ARX, nested F-test |
| JEL: | C22 G12 F51 Q41 |
| Date: | 2026–09 |
| URL: | https://d.repec.org/n?u=RePEc:pre:wpaper:202626 |
| By: | Borgioli, Stefano; Gallo, Giampiero M.; Ongari, Chiara; Mavillonio, Maria Saveria; Giannetti, Caterina |
| Abstract: | Measuring sentiment from financial news is a central task in economics and finance, yet most existing indicators rely on dictionary-based approaches that infer sentiment from word counts and only partially capture context, negation, and semantic structure. This paper proposes a framework for constructing daily news mood indices using transformer-based language models and evaluates whether they better represent sentiment than dictionary-based alternatives. Using 143, 755 financial news articles from Factiva, we classify sentiment at the sentence level with FinBERT and aggregate these predictions into article-level and daily sentiment measures through alternative normalization schemes. We compare the resulting indices with benchmark measures based on Shapiro et al., 2022 and Barbaglia et al., 2025. A central contribution is the validation of alternative sentiment measures against human judgments. We conducted an incentivized annotation exercise in which 444 participants evaluated a validation subsample of 588 financial news articles. Consensus ratings from independent human evaluations serve as an external benchmark for assessing the quality of automated sentiment measures. Across correlation, regression, and classification exercises, transformer-based measures show stronger agreement with human judgments than vocabulary-based alternatives and perform substantially better in distinguishing positive, neutral, and negative articles. Overall, the results suggest that incorporating contextual information through transformer-based language models produces sentiment measures that more closely reflect human assessments of financial news. JEL Classification: C55, C81, E32, E37, G14 |
| Keywords: | daily mood indices, financial news sentiment, FinBERT, human validation, text-as-data, transformer-based language models |
| Date: | 2026–09 |
| URL: | https://d.repec.org/n?u=RePEc:ecb:ecbwps:20263283 |
| By: | Khder Alakkari (Department of Statistics and Programming, Faculty of Economics, Latakia University, P.O. Box 2230, Syria; Faculty of Dentistry, Al Andalus University for Medical Sciences, Tartus, Syria; Department of Financial and Banking Sciences, Faculty of Economics, Tartous University, Syria); Bushra Ali (Department of Financial and Banking Sciences, Faculty of Economics, Tartous University, Syria); Mariem Brahim (Paris School of Business, Paris 75005, France); Rangan Gupta (Department of Economics, University of Pretoria, Private Bag X20, Hatfield 0028, South Africa) |
| Abstract: | In this paper, it is analyzed whether deregulation can be used as a positive or negative state dependent predictor of U.S. stock market bubbles. We use monthly data between February 1973 and May 2025 to calculate the dynamics of these bubbles using 6 Multi-Scale Log-Periodic Power Law Singularity Confidence Indicators, separated by positive and negative bubbles in short-, medium- and long-term horizons. We then use a multivariate quantile-on-quantile regression model to analyse the joint changing nature of the prediction power of deregulation as the distribution of deregulation intensity and the distribution of bubble confidence are controlled for macroeconomic activity and monetary conditions. The results show the best and most consistent results for medium and long horizon positive bubbles. The long-horizon positive bubble confidence is directly associated with deregulation with larger effects in the upper tail of the bubble distribution. Regulation produces broadly opposite patterns, while proposed deregulation displays considerably stronger predictive content than enacted deregulation, supporting an expectations-based regulatory channel. The findings are robust to alternative controls and bandwidth choices. International evidence for the remaining G7 and BRICS markets is more heterogeneous, reinforcing the U.S.-centered nature of the main results. Overall, deregulation contains state-dependent information about the evolution of equity-market imbalances and may complement conventional indicators used to monitor emerging risks to financial stability. |
| Keywords: | Stock market bubbles, MS-LPPLS confidence indicators, Quantile-on-quantile regression, Regulatory policy, Financial stability |
| JEL: | G01 G12 G18 C22 |
| Date: | 2026–09 |
| URL: | https://d.repec.org/n?u=RePEc:pre:wpaper:202627 |
| By: | Vincent Meh; Jabir Sandhu; Andreas Uthemann |
| Abstract: | We examine whether hedge funds and other asset managers contributed to illiquidity in the Government of Canada bond market at the onset of the war in the Middle East in early 2026. We find no evidence that they did. Hedge funds remained active participants in primary auctions, while the extent to which dealers could offset client transactions in secondary markets stayed within its typical range. Although hedge funds did reduce leveraged positions, their trading remained broadly consistent with their usual relative-value strategies. |
| Keywords: | Financial markets and funds management; Funds management; Market functioning |
| JEL: | G G0 G01 G1 G12 G14 G2 G23 G24 |
| Date: | 2026–09 |
| URL: | https://d.repec.org/n?u=RePEc:bca:bocsap:26-42 |
| By: | Kemal Kirtac |
| Abstract: | Financial language models can transform unstructured firm-specific news into structured decision signals, but financial AI research lacks an integrated deployment framework for evaluating whether those signals remain useful in financial decision systems. Computer science research has developed strong methods for time-series forecasting, text classification, multimodal stock prediction, graph-based market modeling, and machine-learning operations, yet these streams do not provide a domain-specific protocol that jointly tests financial language-model outputs under event-time observability, probability calibration, execution timing, transaction costs, liquidity constraints, capacity limits, operational diagnostics, and statistical inference. We introduce MFAST, a Market-Friction-Aware Sentiment-to-Trading framework that converts timestamped financial text into auditable, reproducible, and market-feasible trading decisions. The application is news-based trading, where firm-specific text must be linked to securities before portfolio decisions can be evaluated. The framework links Refinitiv News Analytics to Center for Research in Security Prices (CRSP) equity data, restricts the primary out-of-sample evaluation to post-release news outside disclosed foundation-model data-freshness periods, and adds a public replication arm using open financial text and public price data. Results show that decoder-only language models outperform encoder baselines and dictionary sentiment in classification, calibration, return prediction, and net portfolio performance, while operational diagnostics reveal trade-offs among accuracy, latency, memory, throughput, and inference cost. The paper shows that credible evaluation of financial language models requires an end-to-end engineering approach combining language understanding, temporal discipline, market-friction-aware deployment, and reproducible validation. |
| Date: | 2026–09 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2609.23703 |
| By: | Yu Peng; Matloob Khushi; Josiah Poon |
| Abstract: | Managing drawdown, the peak-to-trough decline in an investment portfolio's value, is a precondition for long-term survival in practical investment management. However, mainstream stock forecasting methods predominantly optimize returns or Sharpe ratios under the independent and identically distributed (i.i.d.) assumption. Real markets do not follow this assumption, triggering catastrophic drawdowns. We propose a cross-asset state-space trading system (CAST), consisting of two components: The predictor, Cross-Asset Collaborative Kalman Filter (CoKF), estimates each asset's latent state online, coupling all assets through their correlations and adaptively fusing multiple integrated-random-walk orders. The controller, Model Predictive Control (MPC), converts the predictor's forecast into trades, using forecast uncertainty as an explicit risk penalty that controls drawdown. We evaluate CAST on four real-world stock markets over a 15-year test window and show that it consistently occupies the return-drawdown Pareto frontier, achieving strong risk-adjusted performance while maintaining substantially lower maximum drawdown than competitive baselines. A stress test across crisis periods further demonstrates robust behavior under market shocks and distribution shift. Because the predictor and controller interact only through the predicted price path, both are plug-and-play, making CAST a modular, interpretable trading system. The code is available at https://github.com/FanBroWell/CAST |
| Date: | 2026–09 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2609.14205 |
| By: | Sadia Anum; Jabir Sandhu; Adrian Walton |
| Abstract: | We document haircut practices in the Canadian dealer-to-client repo market using transaction level data. We find that haircuts are rarely used across repos for different collateral, maturities and most counterparty types. Notable exceptions include wealth managers and some foreign official institutions, which use haircuts more, likely due to their specific regulations or requirements. We also find no evidence that haircut use increased materially during periods of market stress, including March 2020. |
| Keywords: | Financial markets and funds management; Market structure; Financial system; Financial institutions and intermediation; Financial stability and systemic risk; Financial system regulation and oversight |
| JEL: | G G0 G01 G1 G18 G2 G23 G28 |
| Date: | 2026–09 |
| URL: | https://d.repec.org/n?u=RePEc:bca:bocsap:26-39 |
| By: | Arvind Krishnamurthy; Miguel Chumbo |
| Abstract: | We measure the convenience yield on dollar safe assets over the last 5 years to answer this question. We find that within the set of U.S. dollar assets, safe asset convenience yields are about the same in 2025 as they were in 2019, and these levels are expected to persist for the next decade. Across the globe, the richness of U.S. dollar safe assets, particularly Treasurys, compared to the safe assets of other currencies have fallen considerably. There is also evidence that the U.S. dollar repo is currently the highest convenience dollar safe asset. In short, our answer is, no for domestic liquidity purposes, and yes for global liquidity purposes. |
| JEL: | G12 G15 |
| Date: | 2026–09 |
| URL: | https://d.repec.org/n?u=RePEc:nbr:nberwo:35742 |
| By: | Marco Albori (Bank of Italy); Andrea Zaghini (Bank of Italy) |
| Abstract: | We study how the emergence of generative artificial intelligence (AI) affected corporate bond pricing. Using bond-level issuance data for U.S. non-financial corporations and a difference-in-differences design centered on the launch of ChatGPT, we examine whether debt investors differentiate firms according to their position in the AI value chain. We find that hyperscalers experienced a significant decline in borrowing costs relative to comparable issuers, while software firms faced less favorable financing conditions. The results suggest that credit markets rapidly incorporated expectations regarding the distribution of gains and losses from generative AI. More broadly, the evidence indicates that technological shocks affect not only firm valuations but also the cost of external finance, thereby influencing the allocation of capital across firms during periods of technological transition. |
| Keywords: | artificial intelligence, asset pricing, bonds, corporate finance |
| JEL: | G12 G32 C23 |
| Date: | 2026–09 |
| URL: | https://d.repec.org/n?u=RePEc:bdi:opques:qef_1057_26 |
| By: | Antonio Diez de los Rios |
| Abstract: | We show that the parameters of a broad class of portfolio-balance models can be recovered from estimates of Gaussian dynamic term structure models (GDTSMs) using a two-step estimator that bypasses the fixed-point problem that characterizes portfolio balance models. Specifically, we develop a novel canonical representation of bond supply in terms of factor-mimicking portfolios, whose returns replicate the structural shocks to the pricing factors. We identify these structural shocks by assuming that each shock corresponds to a distinct, mutually exclusive dimension of bond-supply variation. Using U.S. Treasury yields and macroeconomic data, we show the recovered shocks admit interpretations as hedging-risk-premium, risk-bearing-capacity, and monetary policy shocks. |
| Keywords: | Financial markets and funds management; Funds management; Models and tools; Econometric, statistical and computational methods |
| JEL: | E E4 E43 E44 G G1 G12 H H6 H63 |
| Date: | 2026–09 |
| URL: | https://d.repec.org/n?u=RePEc:bca:bocawp:26-33 |