|
on Forecasting |
| By: | Tien Thanh Thach |
| Abstract: | Transformers have shown remarkable success in sequence modeling, yet their direct application to financial time series remains challenging due to noisy signals, short-memory dynamics, and distributional shifts. This paper proposes a modified Transformer architecture for one-step stock index forecasting, combined with advanced learning-rate scheduling and a novel Shifted Data Augmentation (SDA) technique. We evaluate the proposed framework on two benchmark stock index datasets, VN30 and S&P 500. Experimental results demonstrate that cosine annealing with warmup consistently improves forecasting accuracy over the generalized inverse-power scheduler. Furthermore, SDA substantially reduces forecasting errors and run-to-run variability while improving robustness to hyperparameter selection. The combination of cosine annealing scheduling and SDA achieved the best performance on both datasets, indicating that data augmentation can play a more important role than increasing model complexity in Transformer-based financial forecasting. These findings provide a practical and computationally efficient approach for robust stock index forecasting in noisy financial environments. |
| Date: | 2026–06 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2606.15701 |
| By: | Likai Chen; Weining Wang |
| Abstract: | Forecasting is a central goal of time-series analysis. This review centers on three major developments in recent AI-based time-series forecasting: transformers, large pretrained models for zero-shot forecasting, and diffusion-based generative forecasters. We connect these methods to the econometric tradition built around the vector autoregression (VAR) through a common object: the conditional distribution of the future given the past. The review is organized around three long-standing challenges: high dimensionality, nonstationarity, and nonlinearity. We argue that modern methods make progress by expanding the classical forecasting template: they allow more flexible dynamics, use larger information sets and training corpora, and represent richer predictive distributions. Yet they often lack the inferential and structural tools that make classical models useful for testing, explanation, and policy analysis. We close by outlining open problems where econometric tools remain important. |
| Date: | 2026–01–30 |
| URL: | https://d.repec.org/n?u=RePEc:bri:uobdis:26/838 |
| By: | Li Chen (Paderborn University) |
| Abstract: | To address enterprise adoption challenges beyond model accuracy, including usability for non-expert users, trust and explainability, and cost efficiency, this paper proposes a hybrid architecture, in which a business AI agent acts as an orchestrator and a time series Model Context Protocol(MCP) server provides reusable forecasting capabilities along a seven-stage forecasting lifecycle. In the proposed architecture, the agent interprets the user request, reasons over the available context, invokes appropriate MCP tools, and translates structured tool outputs into business-oriented explanations. The implemented time series MCP server is evaluated on three real-world datasets under two experimental settings: a zero-shot forecasting setup, in which the agent compares the available forecasting tools, and a diagnostic-aware setup, in which the agent first analyzes data quality, seasonality, stationarity, structural breaks, and influencing factors before selecting a forecasting strategy. The results show that forecasts produced through the MCP server can reach plausible quality levels across all datasets. The diagnostic-aware workflow improved forecast accuracy and explanation quality. The experiments further show that no single model family dominates across all settings: automatic ARIMA achieved the highest aggregate accuracy score but required the highest runtime, while Chronos-2 and Toto 2.0 provided competitive accuracy-runtime trade-offs. Exponential smoothing remained a fast and interpretable baseline. The findings suggest that an MCP-based architecture can make heterogeneous forecasting methods accessible, auditable, and cost-aware for enterprise AI agents, while still requiring clear task specification, governance, and human oversight. |
| Keywords: | time series forecasting, Model Context Protocol, AI agents, AutoML, time series foundation models, enterprise analytics, explainable forecasting, design-based research |
| JEL: | C01 C02 |
| Date: | 2026–06 |
| URL: | https://d.repec.org/n?u=RePEc:pdn:ciepap:178 |
| By: | Shujie Li (Paderborn University); Yuanhua Feng (Paderborn University) |
| Abstract: | Macroeconomic time series forecasting is crucial for guiding government policy decisions, business strategies, and understanding economic trends. However, predicting macroeconomic variables remains a significant challenge. The complexity of economic systems, insufficient data, high levels of volatility complicate the task of accurate forecasting. To enhance forecasting accuracy, we propose two novel models to capture both linear and nonlinear dynamics. First, we generalize the random walk model by incorporating a drift term, which is estimated using a simple neural network model. Second, a hybrid model is introduced to combine local linear regression and the neural network model. Additionally, we adopt other models from Fritz et al. (2024) for combination. These models are combined using a simple averaging method. Our results demonstrate that the newly proposed neural network-based models produce the lowest average MASE. Additionally, model combination is an effective strategy for enhancing the performance of GDP forecasting in most countries and it is less risky than relying on a single model. |
| Keywords: | nonparametric approaches, combination of forecasting, NNAR, Random Walk |
| JEL: | C14 C51 |
| Date: | 2026–03 |
| URL: | https://d.repec.org/n?u=RePEc:pdn:ciepap:172 |
| By: | Xu, Yongdeng (Cardiff University, Cardiff, UK); Lyu, Juyi (Loughborough University, UK); Lu, Wenna (Cardiff Metropolitan University, Cardiff, UK) |
| Abstract: | This paper evaluates an Adaptive LASSO-MGARCH model for multivariate volatility forecasting, with an application to green and conventional bonds, equities, energy commodities, and EU carbon allowances. By introducing coefficient-specific adaptive penalisation directly into the multivariate GARCH variance equations, the model delivers a sparse and data-driven volatility spillover structure while preserving positive definiteness of the conditional covariance matrix. Using daily data on green and conventional bonds, equities, energy commodities, and carbon allowances, we show that adaptive regularisation substantially reduces model complexity and improves economic interpretability relative to an unpenalised MGARCH benchmark. Out-of-sample forecasting experiments at multiple horizons demonstrate that the Adaptive LASSO-MGARCH model consistently achieves lower covariance forecast losses, and statistical tests based on the White reality check confirm that these improvements are significant across alternative loss functions. |
| Keywords: | Adaptive LASSO; Multivariate GARCH; Volatility Forecasting; High-Dimensional; Green Finance |
| JEL: | C32 C58 G17 |
| Date: | 2026–03 |
| URL: | https://d.repec.org/n?u=RePEc:cdf:wpaper:2026/4 |
| By: | Alam, M. Jahangir; Boyle, Shane; Li, Huiyu; Sekhposyan, Tatevik |
| Abstract: | Recent research suggests that generic large language models (LLMs) can match the accuracy of traditional methods when forecasting macroeconomic variables in pseudo out-of-sample settings generated via prompts. This paper assesses the out-of-sample forecasting accuracy of LLMs by eliciting real-time forecasts of U.S. inflation from ChatGPT. We find that out-of-sample predictions are largely inaccurate and stale, even though forecasts generated in pseudo out-of-sample environments are comparable to existing benchmarks. Our results underscore the importance of out-of-sample benchmarking for LLM predictions. |
| Date: | 2026–01 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:21057 |
| By: | Kyriakopoulou, Dimitra |
| Abstract: | We develop a ridge-regularized Factor-Augmented Vector Autoregression (FAVAR) framework for modeling macro-fiscal dynamics in settings where the number of predictors is large relative to the available sample. Rather than relying on an unrestricted factor decomposition, the high-dimensional information set is summarized through two economically interpretable latent factors, namely a real activity factor and a nominal-financial factor, extracted from separate data blocks. These factors are embedded in a regularized VAR together with key observable macro-fiscal variables, including real GDP growth, inflation, the primary balance, and the interest-rate-growth differential, allowing macroeconomic conditions to be linked to sovereign debt dynamics through a nonlinear accounting identity. In an application to Greece, the extracted factors capture major business-cycle, inflationary, and financing episodes, while impulse responses are economically plausible and dynamically stable. Recursive out-of-sample forecasting exercises show that the model performs comparatively well for inflation and financing conditions at selected horizons, while simple autoregressive benchmarks remain difficult to outperform for persistent real activity variables. Overall, the results highlight the usefulness of combining structured factor extraction with ridge regularization to obtain a parsimonious and interpretable framework for macro-fiscal modelling, scenario analysis, and debt sustainability assessment. |
| Keywords: | FAVAR, Ridge Regression, Forecasting, High-Dimensional Data, Fiscal Policy, Debt Dynamics, Macro–Fiscal Modelling |
| JEL: | C32 C38 C53 C55 E62 H63 |
| Date: | 2025–12 |
| URL: | https://d.repec.org/n?u=RePEc:pra:mprapa:129519 |
| By: | Li Chen (Paderborn University); Yuanhua Feng (Paderborn University) |
| Abstract: | Building upon our previous work that integrated a semi-parametric ARMA model into the SAP ecosystem, this paper introduces an enhanced forecasting application for SAP Analytics Cloud (SAC), termed deseatsForecast. The application leverages a data-driven seasonal semiparametric ARMA (S-Semi-ARMA) algorithm and novelly addresses two critical gaps in the practical deployment of advanced semiparametric models within enterprise environments. Specifically, the proposed deseatsForecast application enables robust estimation of slowly-changing seasonal patterns jointly with trend components through a data-driven Iterative Plug-In (IPI) algorithm for bandwidth selection. Secondly, the application provides native support for panel data structures, thereby extending its applicability to multidimensional business datasets commonly encountered in enterprise settings. The paper begins with a review of the data-driven S-Semi-ARMA model and the estimation procedures for trend, seasonal, and residual components. Subsequently, forecasting techniques based on the S-Semi-ARMA framework are presented, followed by a brief description of the architecture and design of the deseatsForecast application, with particular emphasis on its extensions relative to the smootsForecast application. Finally, the forecasting application is empirically validated using OECD passenger car registration data for multiple countries and a comparative study against SAP’s autoML-based forecasting approach is conducted. The empirical results demonstrate consistently strong forecast performance of the deseatsForecast application and highlight its superior forecast accuracy and transparency compared with the current autoML approach in SAP. |
| Keywords: | Time series forecasting, semiparametric algorithm, forecasting accuracy, deseats, SAP Analytics Cloud, enterprise adoption |
| JEL: | C01 C02 |
| Date: | 2026–01 |
| URL: | https://d.repec.org/n?u=RePEc:pdn:ciepap:177 |
| By: | Christopher W. Karvetski; Sheldon S. Huang; Simas Ku\v{c}inskas; Nadja Flechner; Jingyu Hu; Philip Tetlock; Ezra Karger |
| Abstract: | Decision-makers routinely rely on expert judgments accompanied by written explanations, yet explanation quality is difficult to measure at scale. Forecasting tournaments offer a natural testing ground: probabilistic judgments are paired with natural-language rationales and scored against realized outcomes. We introduce Explanation Quality Markers (EQMs), a set of sixty theory-guided reasoning patterns scored by large language models (LLMs). In a pre-registered analysis of over 55, 000 forecast-rationale pairs from a multiyear forecasting tournament, EQMs predict accuracy at both the forecast and forecaster levels, consistently outperforming pre-LLM text-analysis methods. More than 90% of statistically significant pattern-level EQM-accuracy correlations match our directional hypotheses. The signal is asymmetric: EQMs identify likely underperformers more reliably than they distinguish the very best forecasters. Benchmarked against traditional indicators of forecasting skill, EQMs are the strongest predictor at the forecast level and competitive at the forecaster level, though weaker than prior accuracy. Human ratings of rationale quality are less consistently correlated with accuracy and place disproportionate weight on rationale length. Results transfer to an independent forecasting study. EQMs provide a scalable, interpretable method for extracting judgment-relevant information from written explanations. |
| Date: | 2026–06 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2606.30987 |
| By: | Kwon, Byeungchun; Park, Taejin; Rungcharoenkitkul, Phurichai; Smets, Frank |
| Abstract: | Macroeconomic indicators provide quantitative signals that must be pieced together and interpreted by economists. We propose a reversed approach of parsing press narratives directly using Large Language Models (LLM) to recover growth and inflation sentiment indices. A key advantage of this LLM-based approach is the ability to decompose aggregate sentiment into its drivers, readily enabling an interpretation of macroeconomic dynamics. Our sentiment indices track hard-data counterparts closely, providing an accurate, near real-time picture of the macroeconomy. Their components–demand, supply, and deeper structural forces–are intuitive and consistent with prior model-based studies. Incorporating sentiment indices improves the forecasting performance of simple statistical models, pointing to information unspanned by traditional data. |
| JEL: | E30 E44 E60 C55 C82 |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20828 |
| By: | Tobias Adrian; Domenico Giannone; Matteo Luciani; Mike West |
| Abstract: | Central banks monitor macroeconomic risk through two traditions: scenario analysis, regularly used since the mid-1990s, and distributional forecasting, practiced since the late 1960s. The two are complementary but separate: scenarios provide narratives without probabilities, while predictive distributions provide probabilities with limited economic interpretation. Treating baseline forecasts and scenarios as conditional predictive densities, and distributional forecasts as reference predictive distributions, places both within a common framework and clarifies their roles. The Scenario Synthesis assigns weights to scenarios consistent with the reference distribution, offering a practical and reproducible tool for risk assessment and policy deliberation under deep uncertainty. |
| Date: | 2026–06 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2606.16708 |
| By: | Degui Li; Yuying Sun; Boyao Wu |
| Abstract: | In this paper, we introduce a flexible time-varying multi-layer network vector autoregression (VAR) model framework for large-scale time series, allowing agents in dynamic systems to interact through multiple channels and incorporating multiple adjacency matrices to capture network spillover effects. We propose a penalized model averaging method to determine a time-varying optimal combination of multi-layer network VAR candidate models whose number may be divergent. Under some regularity conditions, the asymptotic properties such as asymptotic optimality and convergence rates of the proposed time-varying weight estimation are derived in the contexts of both the in-sample fitting and out-of-sample prediction. In addition, we extend the conformal prediction method to construct prediction bands for locally stationary time series. Monte-Carlo simulation studies and an empirical application to forecast CPI inflation by combining multiple network information are given to illustrate reliable finite-sample estimation and predictive performance of the developed methodology. |
| Date: | 2026–06 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2606.25292 |
| By: | Born, Benjamin; Lamersdorf, Nora; Schuster, Jana-Lynn; Steffen, Sascha |
| Abstract: | Using modern natural language processing, we construct a high-frequency inflation expectations index from German-language tweets. This index closely tracks realized inflation and aligns even more closely with household survey expectations. It also improves short-run forecasts relative to standard benchmarks. In response to monetary policy tightening, the index declines within about a week, with the effects concentrated in tweets by private individuals and during the recent period of elevated inflation. Using 117 million online transactions from German retailers, we show that higher inflation expectations are followed by lower household spending on discretionary goods. By linking these shifts in demand to stock returns, we find that, during periods of elevated inflation, firms operating in discretionary sectors experience significantly lower stock returns when inflation expectations rise. Thus, our Twitter-based index provides market participants and policymakers with a timely tool to monitor inflation sentiment and its economic consequences. |
| Keywords: | Inflation expectations; Social media; Large Language Models; Nlp; Household consumption; Stock returns; Monetary policy |
| JEL: | E31 D84 E58 C45 C81 |
| Date: | 2025–12 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20977 |
| By: | Gondauri, Davit |
| Abstract: | This study develops and empirically audits a Global Riemann-Zeta FPAS+ζ inflation-forecasting framework for world aggregate inflation. The model augments a structural FPAS baseline with a normalized Riemann-zeta critical-line cyclical signal transformed into a signed and alpha-calibrated forecast correction. The Riemann-zeta function is used only as an analogical and computational cyclical signal, not as a proof, test, or verification of the Riemann Hypothesis. The empirical architecture is layered: the principal annual validation layer covers 1970-2024 (N = 55), the annual robustness layer covers 1980-2024 (N = 45), the compact crisis-era annual comparison layer covers 2005-2024 (N = 20), the quarterly dynamic-validation layer covers 2000Q1-2024Q4 (N = 100), and the monthly spectral-diagnostics layer covers 2005M1-2024M12 (N = 240). The methodology constructs a composite global inflation-pressure index, a nonlinear macro-financial tau argument, a normalized zeta-cycle, a zeta deviation from neutrality, and an alpha-weighted FPAS+ζ forecast. Forecast performance is evaluated against FPAS and ARIMA through RMSE, MAE, sMAPE, rolling no-look-ahead validation, Mincer-Zarnowitz calibration, forecast-encompassing tests, Diebold-Mariano and Clark-West comparisons, residual diagnostics, HMM/Markov-switching regimes, Fourier/wavelet coherence, and spectral-entropy screens. The corrected results show that FPAS+ζ improves the reported forecast hierarchy: in the N = 55 annual layer, RMSE is 0.421 compared with 0.612 for FPAS and 1.081 for ARIMA; in the N = 45 robustness layer, RMSE is 0.398 compared with 0.578 and 0.983; and in the N = 20 crisis-era layer, RMSE is 0.233 compared with 0.443 and 1.913. Zeta deviations, composite pressure, F_zeta, W_zeta, and regime diagnostics concentrate around the 2008 global financial crisis, the 2020 pandemic shock, and the 2022 energy-food-monetary tightening shock. The contribution is a bounded mathematical-econometric forecast-diagnostic architecture for global inflation modelling; it does not claim theorem proof, unrestricted causality, residual-problem elimination, or universal forecast dominance across all economies, samples, and horizons. |
| Keywords: | global inflation forecasting, FPAS, Riemann-zeta function, FPAS+ζ, hybrid forecasting, macro-financial pressure, zeta-cycle, critical-line signal, alpha calibration, ARIMA benchmark, Mincer-Zarnowitz calibration, Diebold-Mariano test, Clark-West test, HMM regimes, Markov-switching, spectral diagnostics, wavelet coherence, layered validation |
| JEL: | C02 C22 C32 C43 C51 C52 C53 C58 E31 E37 E44 E47 E52 E58 F41 F47 G01 G17 |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:zbw:esprep:341672 |
| By: | Martin, Ian; Shi, Ran |
| Abstract: | We derive option-implied bounds on the probability of a crash in an individual stock, and argue a priori that the lower bound should be close to the truth. The lower bound successfully forecasts crashes both in and out of sample. Crucially, our theory-based approach avoids the "crying wolf" problem faced by risk-neutral crash probabilities, which severely overstate crash risk during crisis periods. Despite having no free parameters, the lower bound outperforms elastic net, ridge, and Lasso models that flexibly but atheoretically combine stock characteristics, risk-neutral probabilities and the bound itself, because such models overfit during crisis periods. |
| JEL: | G12 G13 G17 G01 |
| Date: | 2026–03 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:21236 |
| By: | Li Chen (Paderborn University); Yuanhua Feng (Paderborn University) |
| Abstract: | Motivated by more and more semi- or nonparametric models applied in time series forecasting and their demonstrated superior performance in many empirical researches, this paper explores the adoption and integration of a semiparametric ARMA model in an enterprise system landscape. We begin by reviewing basic construction of the semiparametric ARMA model, the iterative plug-in algorithm for estimating the trend component of trend stationary times series, forecast techniques and quality measurements, which were well researched and published with the R package smoots. Subsequently, we showcase a novel approach to adopt the semiparametric ARMA model in a forecast application based on SAP Analytics Cloud (SAC), which leverages the platform’s strengths in system integrity, state-of-the-art user interface (UI) design as well as seamless connection to a R engine with smoots package embedded. The forecast application addresses key challenges in terms of cost efficiency, user experience, and the requirement for in-house statistical or machine learning expertise while adopting such statistical algorithms in enterprise context. Finally, we empirically evaluate the forecast quality of the integrated semiparametric ARMA model using real-world data, demonstrating promising results overall. |
| Keywords: | Time series forecasting, semiparametric algorithm, forecasting accuracy, smoots package, SAP Analytics Cloud, enterprise adoption |
| JEL: | C01 C02 |
| Date: | 2025–08 |
| URL: | https://d.repec.org/n?u=RePEc:pdn:ciepap:176 |
| By: | Hauzenberger, Niko; Marcellino, Massimiliano; Pfarrhofer, Michael; Stelzer, Anna |
| Abstract: | We develop Bayesian machine learning methods for mixed frequency data. This involves handling frequency mismatches and specifying functional relationships between (possibly many) predictors and low frequency dependent variables. We use Gaussian Processes (GPs) in direct nonlinear predictive regressions, and compress higher frequency variables in a structured way. This yields a set of kernels for GPs with distinct properties and implications. We evaluate the proposed framework in an out-of-sample exercise focusing on quarterly US GDP growth and inflation. Our approach leverages high-dimensional mixed frequency data in a computationally efficient way, and offers robustness and gains in predictive accuracy along several dimensions. |
| JEL: | C11 C22 C53 E31 E37 |
| Date: | 2026–02 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:21214 |
| By: | Kremens, Lukas; Varela, Liliana |
| Abstract: | Short-horizon exchange rate forecasts systematically - and incorrectly - predict a partial reversal of their previous errors, both in consensus and forecaster-by-forecaster measures. This pattern spans almost two-thirds (two-fifths) of the variation in consensus (individual) forecasts and explains their poor predictive performance at short horizons. We decompose short-term forecasts into three orthogonal components correlated with (i) long-term forecasts, (ii) past errors, and (iii) residual noise. The first two components account for three-quarters of forecast variance and strongly predict realizations. But the error-loading component predicts in the wrong direction, offsetting predictive information in the long-term component, and renders the overall forecast uninformative. |
| Keywords: | forecasts |
| JEL: | F31 G15 G17 |
| Date: | 2026–03 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:21258 |
| By: | Mao Guan; Qian Chen |
| Abstract: | Forecasting benchmarks for retrieval-augmented LLMs routinely confound model capability with information leakage: features labeled with a target's timestamp are often not observable at the system's decision time. We study leakage-controlled equity factor ranking with a retrieval-augmented 7B open-source LLM forecaster. At each month-end from 2023-04 to 2026-03, the forecaster observes only decision-time information: lag-shifted FRED macro variables, recent macro-event summaries, and the Cleveland Fed's archived daily CPI nowcast for unreleased current-month inflation. A macro-analog retrieval module selects historical states, a critic LLM compresses them into one tactical rule, and an actor LLM maps the current state and recent rules into scores for seven U.S. equity style factors. The full pipeline obtains a median monthly Spearman rank IC of +0.154, with positive means across three non-overlapping contiguous 12-month subwindows; the mean IC remains statistically underpowered, with a bootstrap 95% confidence interval that includes zero. Non-LLM baselines under the same decision-time constraint demonstrate that a kNN macro-analog model recovers a comparable median IC, indicating that real-time inflation information and macro-similar retrieval explain much of the median signal. The LLM pipeline retains higher mean IC and a stronger long-short allocation sanity check, suggesting that any marginal benefit is concentrated in the extreme rankings that drive long-short portfolio formation. A descriptive audit of the 36 critic rules and per-month case studies appears in the appendix. |
| Date: | 2026–06 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2606.22719 |
| By: | Adam, Klaus; Kuang, Pei; Xie, Shihan |
| Abstract: | We observe a rich set of public information signals available to participants in the Survey of Professional Forecasters (SPF) and decompose individual forecast revisions into those due to public information and a remainder due to residual information. We find that SPF forecasters overreact to residual information at almost all forecast horizons and for almost all forecast variables. In addition, forecasts are overly anchored to prior beliefs for all variables at all forecast horizons. We show analytically that overconfidence in private information qualitatively generates both of these features. It also implies that forecast errors correlate positively with past forecast revisions at the consensus level, but negatively at the individual level, as documented previously in the literature. Estimating Bayesian updating models on SPF data, we show that overconfidence in private information also replicates the observed patterns quantitatively. All estimated models display strong and statistically significant overconfidence in private information. |
| JEL: | C53 D83 D84 E31 |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20806 |
| By: | Joel P. Flynn; Maksim Meinert; Karthik Sastry |
| Abstract: | With uncertainty about persistence, we show that forecasts necessarily become more persistent and over-react at long horizons. For these reasons, correctly specified and Bayesian forecasts may under-react at short horizons and over-react at long horizons. These results provide a unified explanation for several asset pricing and forecasting puzzles, including: (i) the excess responsiveness of long-horizon rates to short rates, (ii) the dominance of apparent term premia for long-term rates, (iii) the ex post predictability of bond yields, (iv) the excess volatility of long-horizon forward prices, (v) the excess persistence of long-horizon forecasts, and (vi) the over-reaction of long-horizon forecasts. |
| JEL: | E0 E17 G12 G13 |
| Date: | 2026–07 |
| URL: | https://d.repec.org/n?u=RePEc:nbr:nberwo:35411 |
| By: | Cimadomo, Jacopo; Giannone, Domenico; Lenza, Michele; Monti, Francesca; Sokol, Andrej |
| Abstract: | We design a Bayesian Mixed-Frequency vector autoregression (VAR) model for fiscal monitoring, i.e., to nowcast the government deficit-to-GDP ratio in real time and provide a narrative for its dynamics. The model incorporates both monthly cash and quarterly accrual fiscal indicators, together with other high-frequency macroeconomic and financial variables, as well as real GDP and the GDP deflator. Our model produces timely monthly density nowcasts of the annual deficit ratio, while governments and official institutions generally only publish their point predictions bi-annually. Based on a database of real-time vintages of macroeconomic, financial and fiscal variables for Italy, we show that the nowcasts of the annual deficit to GDP ratio of our model are similarly or more accurate than those of the European Commission, depending on the month in which the nowcast is produced. Our scenario analysis compares the dynamics of the deficit ratio associated with a monetary and a typical recession, finding a more muted response in the latter case. |
| Keywords: | Nowcasting; Mixed-frequency; Monetary policy shock |
| JEL: | C11 E52 E62 E63 H68 |
| Date: | 2026–02 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:21160 |
| By: | Simone Cuonzo; Nina Deliu |
| Abstract: | This paper extends classical conformal frameworks for constructing prediction intervals with global marginal coverage $1-\alpha$ to intervals that provide explicitly calibrated guarantees for the upper and lower tails separately. Focusing on split conformal prediction, we first construct lower and upper one-sided conformal intervals that achieve marginal validity, and then derive the induced two-sided interval by intersection. Theoretical results prove both tail-specific and global marginal coverage of the induced two-sided interval. Results are presented first for the exchangeable setting, where coverage has finite-sample guarantees, and then for non-exchangeable data, where guarantees are asymptotic. Simulation studies show that the proposed approach achieves improved directional calibration relative to classical two-sided intervals, especially relevant in skewed data. Finally, the benefit of the proposed framework is showcased in a financial application, where one aims for return maximization while seeking strict control on the left tail. |
| Date: | 2026–06 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2606.18199 |
| By: | Saboin, José Luis; Guerrero, Diego; Mazzocca, Angelo |
| Abstract: | This paper introduces a novel real GDP growth nowcasting strategy based on a distribution of nowcast values derived from a large dataset and multiple variable combinations, specifications and estimators. We exploit structured and unstructured data from The Bahamas, using thirteen estimators from the econometrics and machine learning literature. The study begins describing our dataset with over three hundred variables from the national statistics system, satellite nighttime lights data for key geographic locations in the archipelago, and internet search trends. Next, we nowcast the seasonally-adjusted annualized quarter-on-quarter growth rate of real gross domestic product using multiple combinations of variables, specifications and estimators. Overall, the ensemble method produces a distribution of nowcasts that outperform a human-designed benchmark. This paper contributes a novel idea of exploiting a distribution of estimates rather than point-values from a determined specification or estimator. |
| Keywords: | Nowcasting;Tourism;machine learning;Time Series;Google Trends;nighttime lights;GDP Growth |
| JEL: | C53 C55 E37 O47 C38 L83 O54 |
| Date: | 2026–05 |
| URL: | https://d.repec.org/n?u=RePEc:idb:brikps:14635 |
| By: | Jordi Llorens-Terrazas; Mika Meitz |
| Abstract: | We propose a flexible framework for modeling the predictive distributions of nonlinear, possibly multivariate time series. Our approach expresses a general predictive distribution in an appropriate generative representation that is based on a folklore result from measure theoretic probability. This representation provides a direct simulation-based approximation to the predictive distribution, enabling straightforward computation of forecasts for the conditional mean and variance, fan charts, value at risk, expected shortfall, joint tail risks, and other quantities of interest. We estimate this generative representation using a version of conditional generative adversarial networks and provide a formal statistical analysis of estimation under weak temporal dependence. Specifically, estimation is expressed as a particular minimax problem and we establish consistency of its approximate solutions in Hausdorff distance. The empirical relevance of the approach is illustrated using applications to equity returns, realized variance, and realized covariances. The proposed method is also computationally manageable, with estimation in our applications taking approximately one minute on a standard laptop. |
| Date: | 2026–06 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2606.16773 |