|
on Forecasting |
| By: | Katarzyna Maciejowska; Arkadiusz Lipiecki; Bartosz Uniejewski |
| Abstract: | In recent years, a rapid development of forecasting methods has led to an increase in the accuracy of predictions. In the literature, forecasts are typically evaluated using metrics such as Root Mean Squared Error (RMSE) and Mean Absolute Error (MAE). While appropriate for statistical assessment, these measures do not adequately reflect the economic value of forecasts. This study addresses the decision-making problem faced by a battery energy storage system, which must determine optimal charging and discharging times based on day-ahead electricity price forecasts. To explore the relationship between forecast accuracy and economic value, we generate a pool of 192 forecasts. These are evaluated using seven statistical metrics that go beyond RMSE and MAE, capturing various characteristics of the predictions and associated errors. We calculate the dynamic correlation between the statistical measures and gained profits to reveal that both RMSE and MAE are only weakly correlated with revenue. In contrast, measures that assess the alignment between predicted and actual daily price curves have a stronger relationship with profitability and are thus more effective for selecting optimal forecasts. |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2511.13616 |
| By: | M. Stocker; W. Ma{\l}gorzewicz; M. Fontana; S. Ben Taieb |
| Abstract: | Conformal prediction is a powerful post-hoc framework for uncertainty quantification that provides distribution-free coverage guarantees. However, these guarantees crucially rely on the assumption of exchangeability. This assumption is fundamentally violated in time series data, where temporal dependence and distributional shifts are pervasive. As a result, classical split-conformal methods may yield prediction intervals that fail to maintain nominal validity. This review unifies recent advances in conformal forecasting methods specifically designed to address nonexchangeable data. We first present a theoretical foundation, deriving finite-sample guarantees for split-conformal prediction under mild weak-dependence conditions. We then survey and classify state-of-the-art approaches that mitigate serial dependence by reweighting calibration data, dynamically updating residual distributions, or adaptively tuning target coverage levels in real time. Finally, we present a comprehensive simulation study that compares these techniques in terms of empirical coverage, interval width, and computational cost, highlighting practical trade-offs and open research directions. |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2511.13608 |
| By: | Ping Wu; Dan Zhu |
| Abstract: | Financial markets are interconnected, with micro-currents propagating across global markets and shaping economic trends. This paper moves beyond traditional stock market indices to examine cross-sectional return distributions-15 in our empirical application, each representing a distinct global market. To facilitate this analysis, we develop a matrix functional VAR method with interpretable factors extracted from cross-sectional return distributions. Our approach extends the existing framework from modeling a single function to multiple functions, allowing for a richer representation of cross-sectional dependencies. By jointly modeling these distributions with U.S. macroeconomic indicators, we uncover the predictive power of financial market in forecasting macro-economic dynamics. Our findings reveal that U.S. contractionary monetary policy not only lowers global stock returns, as traditionally understood, but also dampens cross-sectional return kurtosis, highlighting an overlooked policy transmission. This framework enables conditional forecasting, equipping policymakers with a flexible tool to assess macro-financial linkages under different economic scenarios. |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2511.17140 |
| By: | Ryan Engel; Yu Chen; Pawel Polak; Ioana Boier |
| Abstract: | Conditional Autoencoders (CAEs) offer a flexible, interpretable approach for estimating latent asset-pricing factors from firm characteristics. However, existing studies usually limit the latent factor dimension to around K=5 due to concerns that larger K can degrade performance. To overcome this challenge, we propose a scalable framework that couples a high-dimensional CAE with an uncertainty-aware factor selection procedure. We employ three models for quantile prediction: zero-shot Chronos, a pretrained time-series foundation model (ZS-Chronos), gradient-boosted quantile regression trees using XGBoost and RAPIDS (Q-Boost), and an I.I.D bootstrap-based sample mean model (IID-BS). For each model, we rank factors by forecast uncertainty and retain the top-k most predictable factors for portfolio construction, where k denotes the selected subset of factors. This pruning strategy delivers substantial gains in risk-adjusted performance across all forecasting models. Furthermore, due to each model's uncorrelated predictions, a performance-weighted ensemble consistently outperforms individual models with higher Sharpe, Sortino, and Omega ratios. |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2511.17462 |
| By: | Kéa Baret; Frédérique Bec; Marion Cochard |
| Abstract: | We propose a simple, simulation based framework for stochastic debt sustainability analysis. Estimating a parsimonious vector autoregression (frequentist and Bayesian) on quarterly French data (1990:Q1–2023:Q4) for the debt's key drivers, we generate predictive fan charts and probability statements for debt to GDP outcomes. Median VAR projections are close to a hypothetical deterministic baseline derived from the deterministic debt sustainability analysis framework. Assuming this illustrative central scenario, historical relationships estimated by our VAR models imply a corresponding confidence band around the debt trajectory. The BVAR yields slightly wider cones and lower tail probabilities than the frequentist VAR, with cone widths between those reported by the European Commission and the ECB. Our analysis, which does not reflect the most recent developments in public finance, suggests that an ambitious fiscal consolidation effort would be required to materially enhance the prospects of stabilizing the debt-to-GDP ratio over the medium term. |
| Keywords: | Debt Sustainability, Stochastic Analysis, VAR Model, Bayesian Forecasting, Density Forecasts |
| JEL: | C3 E6 H6 |
| Date: | 2025 |
| URL: | https://d.repec.org/n?u=RePEc:bfr:banfra:1019 |
| By: | Mainak Singha |
| Abstract: | We document a high-performing cross-sectional equity factor that achieves out-of-sample Sharpe ratios above 13 through regime-conditional signal activation. The strategy combines value and short-term reversal signals only during stock-specific drift regimes, defined as periods when individual stocks show more than 60 percent positive days in trailing 63-day windows. Under these conditions, the factor delivers annualized returns of 158.6 percent with 12.0 percent volatility and a maximum drawdown of minus 11.9 percent. Using rigorous walk-forward validation across 20 years of S&P 500 data (2004 to 2024), we show performance roughly 13 times stronger than market benchmarks on a risk-adjusted basis, produced entirely out-of-sample with frozen parameters. The factor passes extensive robustness tests, including 1, 000 randomization trials with p-values below 0.001, and maintains Sharpe ratios above 7 even under 30 percent parameter perturbations. Exposure to standard risk factors is negligible, with total R-squared values below 3 percent. We provide mechanistic evidence that drift regimes reshape market microstructure by amplifying behavioral biases, altering liquidity patterns, and creating conditions where cross-sectional price discovery becomes systematically exploitable. Conservative capacity estimates indicate deployable capital of 100 to 500 million dollars before noticeable performance degradation. |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2511.12490 |
| By: | Julian Lütticke; Lukas Lautenschlaeger; Wolfgang Schäfers |
| Abstract: | The role of investor sentiment in real estate investment trust (REIT) markets is well-documented. However, traditional sentiment indicators often fail to fully capture real-time market dynamics. This study explores the potential of GenAI-extracted sentiment in forecasting U.S. REIT returns by leveraging large language models (LLMs) to analyze textual data from news media sources. The hypothesis underpinning this study is that LLMs can process textual data in a manner analogous to that of humans. The novel sentiment score is integrated into a machine learning model to predict REIT returns. The analysis differentiates between overall index returns and sector-specific REIT performance, thereby providing a more granular view of sentiment-driven market behavior. In addition to traditional statistical metrics the model performance is assessed by evaluating an active trading strategy based on sentiment signals. This strategy is benchmarked against a buy-and-hold approach to determine whether sentiment-based predictions can systematically outperform the market. The findings contribute to the growing field of AI-driven financial forecasting and offer valuable insights for investors and policymakers in the indirect real estate sector. |
| Keywords: | Generative AI; Large Language Model; News Sentiment; REIT |
| JEL: | R3 |
| Date: | 2025–01–01 |
| URL: | https://d.repec.org/n?u=RePEc:arz:wpaper:eres2025_242 |
| By: | Carlos Madeira |
| Abstract: | Using hand-collected data with biographical information on central bank governors and board members in over 200 countries, I obtain experience-based forecasts for GDP growth and inflation based on an adaptive learning model estimated from their lifetime macroeconomic data. I show life experience influences the monetary policy rates, even after accounting for other macroeconomics observables in the empirical Taylor rule. The role of personal experience is lower in advanced economies and for central bankers with treasury experience. Furthermore, life experience influences the tone of speeches for monetary policy, financial stability and climate concerns. Weather disasters experience reduces climate concerns and NGFS membership. |
| Keywords: | monetary policy, fiscal policy, experience effects, forecasting, learning, beliefs |
| JEL: | D83 D84 E37 E50 E60 E70 |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:bis:biswps:1304 |
| By: | Hongyang Yang; Xiao-Yang Liu; Qingwei Wu |
| Abstract: | Stock recommendation is vital to investment companies and investors. However, no single stock selection strategy will always win while analysts may not have enough time to check all S&P 500 stocks (the Standard & Poor's 500). In this paper, we propose a practical scheme that recommends stocks from S&P 500 using machine learning. Our basic idea is to buy and hold the top 20% stocks dynamically. First, we select representative stock indicators with good explanatory power. Secondly, we take five frequently used machine learning methods, including linear regression, ridge regression, stepwise regression, random forest and generalized boosted regression, to model stock indicators and quarterly log-return in a rolling window. Thirdly, we choose the model with the lowest Mean Square Error in each period to rank stocks. Finally, we test the selected stocks by conducting portfolio allocation methods such as equally weighted, mean-variance, and minimum-variance. Our empirical results show that the proposed scheme outperforms the long-only strategy on the S&P 500 index in terms of Sharpe ratio and cumulative returns. This work is fully open-sourced at \href{https://github.com/AI4Finance-Foun dation/Dynamic-Stock-Recommendation-Mach ine_Learning-Published-Paper-IEEE}{GitHu b}. |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2511.12129 |
| By: | Owen Gabourys; Farrukh Suvankulov; Mathieu Utting |
| Abstract: | This paper replicates and extends the work of Altig et al. (2022) on firms’ subjective sales growth expectations using Canadian survey data from the Bank of Canada’s Business Leaders’ Pulse. We examine the formation, uncertainty and predictive validity of firm-level sales growth forecasts using subjective probability distributions from business leaders at a one-year-ahead horizon. The replication work performed here confirms several findings from Altig et al. (2022), including that expected sales growth predicts realized sales growth, subjective uncertainty predicts forecast errors and firms frequently revise their expectations, usually by small amounts. We also find that subjective uncertainty predicts the magnitude of forecast revisions and follows a V-shaped relationship with past sales growth. We extend the original analysis by further demonstrating that firms with weaker recent performance assign greater weight to future weak growth scenarios, and subsequently that these firms are more likely to underperform, suggesting expectations are grounded in real conditions. The results presented in this paper reinforce the value of firm-level survey data for macroeconomic forecasting and policy analysis and help validate the Business Leaders’ Pulse as a reliable source of firm-level expectations data. |
| Keywords: | Firm dynamics Monetary policy and uncertainty |
| JEL: | C8 C83 D D2 D22 |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:bca:bocadp:25-15 |
| By: | Tuukka Huhtala; Steven Bourassa; Martin Hoesli; Wilma Nissilä; Elias Oikarinen |
| Abstract: | Econometric tests of house price bubbles based on time series explosiveness have become popular in empirical research. These tests typically have good ex-post performance in identifying bubble periods, but their ability to predict large house price declines ex-ante remains an open question. We study the most popular versions of these tests and assess their usefulness as real-time early warning indicators of large house price declines, a feature valuable for policymakers and investors alike. Using a panel of MSA-level data from the U.S., we estimate local housing market bubble periods indicated by each test and assess their ex-ante accuracy in predicting large house price declines. Consistent with previous studies, we find considerable heterogeneity in bubble periods across locations. Although there are complications with real-time interpretation of bubble signals, they are useful in predicting large house price declines. |
| Keywords: | house price bubbles; U.S. housing markets |
| JEL: | R3 |
| Date: | 2025–01–01 |
| URL: | https://d.repec.org/n?u=RePEc:arz:wpaper:eres2025_127 |
| By: | Della Corte, Pasquale; Gao, Can; Preve, Daniel P. A.; Valente, Giorgio |
| Abstract: | This paper investigates the long-horizon predictive variance of an international bond strategy where a U.S. investor holds unhedged positions in constant-maturity long-term foreign bonds funded at domestic short-term interest rates. Using over two centuries of data from major economies, the study finds that predictive variance grows with the investment horizon, driven primarily by uncertainties in interest rate differentials and exchange rate returns, which outweigh mean reversion effects. The analysis, incorporating both observable and unobservable predictors, highlights that unobservable predictors linked to shifts in monetary and exchange rate regimes are the dominant source of long-term risk, offering fresh insights into international bond investment strategies. |
| Keywords: | Currency risk, Long-term bonds, Predictability, Long-term investments |
| JEL: | F31 G12 G15 |
| Date: | 2025 |
| URL: | https://d.repec.org/n?u=RePEc:zbw:safewp:331899 |
| By: | Martin Geiger; Iacovos Sterghides; Marios Zachariadis |
| Abstract: | Utilizing a very large household-level dataset of inflation expectations for twelve euro-area economies, we attempt to assess the formation and accuracy of inflation expectations following major disruptions of the macroeconomy which we identify during the period from 2004:1 to 2025:02. We find that these adverse events tend to increase the degree of inaccuracy in inflation expectations. We also find that this happens because inflation expectations tend to go up in response to these shocks relative to the 12-month ahead inflation realizations, which offers direct evidence of overestimation of inflation. This is consistent with overreaction of inflation expectations in response to inflationary news and with inflation-as-a-bad behavioral patterns in response to adverse non-inflationary shocks. We infer that such behavioral biases appear to have played an important role in the formation of inflation expectations in the euro-area following adverse shocks during the past two decades. |
| Keywords: | forecast errors, behavioral bias, macroeconomic shocks. |
| JEL: | D84 E31 E70 |
| Date: | 2025–11–18 |
| URL: | https://d.repec.org/n?u=RePEc:ucy:cypeua:04-2025 |
| By: | Karen Martinus; Jane Zheng |
| Abstract: | Following a strong period of logistics rental growth in the UK, we are entering a period of more moderate rent growth. It is likely that certain locations and asset types will outperform but in this new environment, gauging future rental growth prospects is increasingly important to investors, developers, and managers. This research aims to provide a granular understanding of the UK logistics market by leveraging geospatial analysis, machine learning, and feature engineering techniques. By integrating high-resolution demographic projections, attributes of existing industrial properties, and rent trends, we develop top-down market analyses and predictive models to estimate rental prices and rent growth potential for individual properties and locations. In doing so, we believe our research has the potential to aid in investment decision making. Additionally, the findings will contribute to a more refined understanding of spatial dependencies and economic drivers within the logistics sector, offering a scalable framework for market forecasting and investment strategy optimization. |
| Keywords: | Forecasting Title change requested on May 15 by email to help; Geospatial; Logistics; Machine Learning |
| JEL: | R3 |
| Date: | 2025–01–01 |
| URL: | https://d.repec.org/n?u=RePEc:arz:wpaper:eres2025_259 |
| By: | Ferreira Batista Martins, Igor (Örebro University School of Business); Virbickaitè, Audronè (CUNEF University, Madrid, Spain); Nguyen, Hoang (Linköping University); Freitas Lopes, Hedibert (nsper Institute of Education and Research, Sao Paulo, Brazil) |
| Abstract: | We propose a high-frequency stochastic volatility model that integrates persistent component, intraday periodicity, and volume-driven time-of-day effects. By allowing intraday volatility patterns to respond to lagged trading activity, the model captures economically and statistically relevant departures from traditional intraday seasonality effects. We find that the volumedriven component accounts for a substantial share of intraday volatility for futures data across equity indexes, currencies, and commodities. Out-of-sample, our forecasts achieve near-zero intercepts, unit slopes, and the highest R2 values in Mincer-Zarnowitz regressions, while horserace regressions indicate that competing forecasts add little information once our predictions are included. These statistical improvements translate into economically meaningful gains, as volatility-managed portfolio strategies based on our model consistently improve Sharpe ratios. Our results highlight the value of incorporating lagged trading activity into high-frequency volatility models. |
| Keywords: | Intraday volatility; high-frequency; volume; periodicity. |
| JEL: | C11 C22 C53 C58 |
| Date: | 2025–11–21 |
| URL: | https://d.repec.org/n?u=RePEc:hhs:oruesi:2025_014 |
| By: | Neville, Ruth (UCL); Rowe, Francisco (University of Liverpool); Zagheni, Emilio |
| Abstract: | International student mobility (ISM) is a critical component of the global migration system, with profound implications for higher education financing, soft power, and geopolitical relations. Within this, the UK is the second leading destination for international students, but its position is under mounting pressure. Using administrative data from the UK’s Colleges and Admissions Service (UCAS) covering applications from 86 countries between 2010-2024, we apply machine learning forecasting to project undergraduate applications through to 2030. Our approach produces disaggregated, country-level forecasts based on the timeliest data. Our findings reveal evidence of the emergence of an increasingly concentrated system: China, India, and the EU, despite the EU’s dramatic post-Brexit decline, will comprise of over 60% of applications by 2030. This trend reflects a concentration paradox, where growth in successful applications from these blocs is expected to slow but reliance intensifies, increasing system vulnerability to policy volatility and geopolitical shocks. These results underscore vulnerabilities in the UK higher education system and hold important insight for migration governance systems and institutions. |
| Date: | 2025–11–10 |
| URL: | https://d.repec.org/n?u=RePEc:osf:socarx:zkpm4_v1 |
| By: | Alain Coen; Aurelie Desfleurs |
| Abstract: | The aim of this study is to investigate the potential impact of US REITs geographic concentration on the accuracy and bias of financial analysts’ earnings (and FFO) forecasts. Using a unique property-level dataset, we analyze from 2000 to 2023 the impact of geographic concentration on the relative performances of real estate investments trusts (REITs). We use different metrics to measure the level of geographic concentration. First, we document the coverage, the accuracy and the bias of financial analysts’ earnings forecasts on «concentrated» and «diversified» REITs. Our results report that the level of accuracy and the level of optimism are statistically different for these two categories, and statistically related to concentration indices. Second, we focus on the different geographic concentration indices as potential determinants of financial analysts’ forecasts accuracy and bias. Our empirical results shed new light on the relative importance of the level of geographic concentration, or home bias at home, since the early 2000’s and the US REITs maturity era, on the complexity of financial analysts’ forecasts, suggesting implications for asset managers, investors and policymakers. |
| Keywords: | Financial analystsâ; REITs |
| JEL: | R3 |
| Date: | 2025–01–01 |
| URL: | https://d.repec.org/n?u=RePEc:arz:wpaper:eres2025_172 |