|
on Financial Markets |
| By: | Nikolas Anic (Swiss Finance Institute - University of Zurich; Finreon); Andrea Barbon (University of St. Gallen; University of St.Gallen); Ralf Seiz (University of St.Gallen; Finreon); Carlo Zarattini (Concretum Group) |
| Abstract: | This paper investigates whether large language models (LLMs) can improve cross-sectional momentum strategies by extracting predictive signals from firm-specific news. We combine daily U.S. equity returns for S&P 500 constituents with high-frequency news data and use prompt-engineered queries to ChatGPT that inform the model when a stock is about to enter a momentum portfolio. The LLM evaluates whether recent news supports a continuation of past returns, producing scores that condition both stock selection and portfolio weights. An LLM-enhanced momentum strategy outperforms a standard longonly momentum benchmark, delivering higher Sharpe and Sortino ratios both in-sample and in a truly out-of-sample period after the model's pre-training cutoff. These gains are robust to transaction costs, prompt design, and portfolio constraints, and are strongest for concentrated, high-conviction portfolios. The results suggest that LLMs can serve as effective real-time interpreters of financial news, adding incremental value to established factor-based investment strategies. |
| Keywords: | Large Language Models, Momentum Investing, Textual Analysis, News Sentiment, Artificial Intelligence |
| Date: | 2025–10 |
| URL: | https://d.repec.org/n?u=RePEc:chf:rpseri:rp2594 |
| By: | Lars A. Lochstoer; Stig R. H. Lundeby; Zhaneta K. Tancheva |
| Abstract: | Recent evidence in the psychology literature suggests that individuals' degree of present bias varies with the state of nature and increases under stress. Consistent with this notion, we document, using survey data on individuals’ expected and realized consumption growth, that agents indeed tend to overconsume relative to their expectations and that the degree of such overconsumption is higher in bad times. We analyze the asset pricing implications of this bias and show that even if they control a small fraction of wealth, investors with time-varying degree of present bias cause substantial, priced discount rate risks that have first-order impact on the level and time-variation of asset risk premiums. The mechanism is distinct from that of models with time-varying preference parameters and that of models with biased expectations about aggregate outcomes. |
| JEL: | G12 G41 |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:nbr:nberwo:34453 |
| By: | Corsetti, G.; Lloyd, S.; Marin, E.; Ostry, D. |
| Abstract: | We document a rise in investors' assessment of U.S. risk relative to other G.7 economies since the late 1990s, driven by higher permanent risk but not reflected in currency returns. Using a two-country framework with trade in a rich maturity structure of bonds which earn convenience yields, alongside risky assets and currencies, we establish an equilibrium relationship between cross-border convenience yields, relative country risk and carry-trade returns. Empirically, we identify a cointegrating relationship between relative permanent risk and long-maturity convenience yields. Counterfactual experiments show rising relative permanent risk explains around one-third of declining long-maturity convenience yields over 2002-2006 and 2010-2014. |
| Keywords: | Convenience Yields, Exchange Rates, Long-Run Risk, U.S. Safety, Equity Risk Premium |
| JEL: | F30 F31 G12 |
| Date: | 2025–09–16 |
| URL: | https://d.repec.org/n?u=RePEc:cam:camdae:2570 |
| By: | Dmitrii Vlasiuk; Mikhail Smirnov |
| Abstract: | We test the hypothesis that consecutive intraday price changes in the most liquid U.S. equity ETF (SPY) are conditionally nonrandom. Using NBBO event-time data for about 1, 500 regular trading days, we form for every lag L ordered pairs of a backward price increment ("push") and a forward price increment ("response"), standardize them, and estimate the expected responses on a fine grid of push magnitudes. The resulting lag-by-magnitude maps reveal a persistent structural shift: for short lags (1-5, 000 ticks), expected responses cluster near zero across most push magnitudes, suggesting high short-term efficiency; beyond that range, pronounced tails emerge, indicating that larger historical pushes increasingly correlate with nonzero conditional responses. We also find that large negative pushes are followed by stronger positive responses than equally large positive pushes, consistent with asymmetric liquidity replenishment after sell-side shocks. Decomposition into symmetric and antisymmetric components and the associated dominance curves confirm that short-horizon efficiency is restored only partially. The evidence points to an intraday, lag-resolved anomaly that is invisible in unconditional returns and that can be used to define tradable pockets and risk controls. |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2511.06177 |
| By: | Toby Barter; Zheng Gao; Eva Christodoulaki; Jing Chen; John Cartlidge |
| Abstract: | Bond markets respond differently to macroeconomic news compared to equity markets, yet most sentiment models, including FinBERT, are trained primarily on general financial or equity news data. This mismatch is important because bond prices often move in the opposite direction to economic optimism, making general or equity-based sentiment tools potentially misleading. In this paper, we introduce BondBERT, a transformer-based language model fine-tuned on bond-specific news. BondBERT can act as the perception and reasoning component of a financial decision-support agent, providing sentiment signals that integrate with forecasting models. It is a generalisable framework for adapting transformers to low-volatility, domain-inverse sentiment tasks by compiling and cleaning 30, 000 UK bond market articles (2018--2025) for training, validation, and testing. We compare BondBERT's sentiment predictions against FinBERT, FinGPT, and Instruct-FinGPT using event-based correlation, up/down accuracy analyses, and LSTM forecasting across ten UK sovereign bonds. We find that BondBERT consistently produces positive correlations with bond returns, achieves higher alignment and forecasting accuracy than the three baseline models, with lower normalised RMSE and higher information coefficient. These results demonstrate that domain-specific sentiment adaptation better captures fixed income dynamics, bridging a gap between NLP advances and bond market analytics. |
| Date: | 2025–10 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2511.01869 |
| By: | Ingomar Krohn; Andreas Uthemann; Rishi Vala; Jun Yang |
| Abstract: | We establish an empirical framework that causally identifies how Treasury demand shocks transmit across foreign exchange and global bond markets, providing direct validation of quantity-driven theories of international risk premia. Our identification exploits predetermined auction supply to isolate demand shocks from high-frequency movements in Treasury futures prices around Treasury auctions. A one-standard-deviation increase in Treasury demand causes the U.S. dollar to depreciate by 2 basis points against G9 currencies while generating 10-basis-point increases in foreign bond prices. Effects persist for two weeks, indicating meaningful economic impacts. The transmission mechanism varies systematically across countries: those with lower U.S. short-rate correlations exhibit stronger currency responses but weaker bond effects, while higher-correlation countries show the opposite pattern. This cross-sectional variation provides empirical support for models of segmented markets where global arbitrageurs link exchange rates and bond risk premia. |
| Keywords: | Asset pricing; Exchange rates; International financial markets; Interest rates; Market structure and pricing |
| JEL: | F30 F31 G12 G15 |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:bca:bocawp:25-29 |
| By: | Ángel Estrada García (BANCO DE ESPAÑA); Christian E. Castro (CAIXABANK); Gonzalo Fernández Dionis (GWU) |
| Abstract: | The 2010 sovereign crisis in the euro area brought to light the depth of the monetary union’s structural weaknesses. In particular, it highlighted the dangers of the sovereign-bank nexus – the amplification effect resulting from sovereign debt being held primarily by domestic banks. In response, important changes have been put in place. From a crisis management perspective, new institutions were created, such as the European Stability Mechanism which acts as a lender of last resort for euro area countries in difficulties. From an ex ante perspective, the crisis led to the launch of the banking union, which comprises the Single Resolution Mechanism – including the Single Resolution Fund – and the still-pending European Deposit Insurance Scheme. In addition, a wide array of regulation has been put in place, including Basel III and MREL/TLAC requirements to reduce the need for bailouts during financial crises and therefore limit the use of public funds in resolution processes. Despite this, the regulatory debate on how banking regulation should address this sovereign-bank interdependence continues today. In this paper we review the main regulatory proposals aimed at curtailing both exposure to sovereign risk and ownership concentration - two factors often associated to the broader sovereign-bank nexus. We assess their impact on bank capital and risk-weighted assets and simulate banks’ responses to these measures. We find that these solutions could entail significant side effects for both banks and bond markets, highlighting the importance of completing the monetary union and, in particular, issuing a European safe asset as key measures to mitigate this vulnerability. |
| Keywords: | sovereign debt, banking regulation, safe asset, monetary union |
| JEL: | H63 G21 G28 F45 |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:bde:wpaper:2540 |
| By: | Ardelia L. Amardana (Ca’ Foscari University of Venice); Diana Barro (Ca’ Foscari University of Venice); Marco Corazza (Ca’ Foscari University of Venice) |
| Abstract: | Sustainability in financial markets has gained attention. This study addresses it by enhancing portfolio optimization through as additional inputs alongside price data that can improve stock return prediction. Using LSTM models with RMSProp optimizer performs best in consistency of minimizing prediction errors and given the ability to capture complex pattern between price, greenhouse gas (GHG) emissions and environmental scores (E-Scores). This study uses data from the EURO STOXX 50 between 2016 and 2022, focusing on out-of-sample weekly return predictions in 2022. Four model setups are tested: price-only, and price combined with GHG, E-score, or both. Our findings show that incorporating the E-Score improves price and return predictions in several sectors, whereas some sectors show limited benefit, indicating sustainability information may already be priced in. Additionally, in portfolio optimization shows that models including E-Score gives better performance across different holding periods by setting more effective weightings and aligning closely with our benchmark. This results provides further evidence in the following year 2023 and EURO STOXX 50 ESG performance. |
| Keywords: | Sustainable Indicators; Long Short Term Memory; Return Prediction; Portfolio Optimization |
| JEL: | C45 C53 C63 |
| Date: | 2025 |
| URL: | https://d.repec.org/n?u=RePEc:ven:wpaper:2025:25 |
| By: | Koji Takahashi (Director, Institute for Monetary and Economic Studies, Bank of Japan (E-mail: kouji.takahashi-2@boj.or.jp)); Sumiko Takaoka (Professor, Faculty of Business Administration, Seikei University (E-mail: takaoka@bus.seikei.ac.jp)) |
| Abstract: | This study empirically investigates the effects of the corporate bond purchase program conducted by the Bank of Japan (BOJ) on bond issuance and issuers. By applying a large language model to market news on corporate bond issuance, we identify individual bond issues that attract demand from investors intending to resell them to the BOJ through its purchase program. Using this indicator of demand, we find that the credit spreads of such bonds are more than 20% lower than those without it, and their issuance sizes are approximately 15% larger. Additionally, the effects on corporate bonds with investor demand to utilize the BOJ program are greater than those on bonds that are merely eligible for it. Finally, we show that firms leveraging the BOJ's program to increase bond issuance allocate the raised funds to capital investment while reducing bank borrowing. |
| Keywords: | Corporate bonds, Corporate bond purchase program, Demand uncertainty, Marketing news, Large Language Model |
| JEL: | E43 E52 G12 |
| Date: | 2025–09 |
| URL: | https://d.repec.org/n?u=RePEc:ime:imedps:25-e-11 |