|
on Computational Economics |
| By: | Paulo Júlio |
| Abstract: | This paper assesses whether pretrained time-series foundation models and locally trained deep neural networks can make a useful contribution to central bank forecasting in a small, data-constrained economy. We compare TimesFM, Chronos-2, Moirai 2.0, and several locally estimated architectures with conventional statistical and econometric benchmarks and Banco de Portugal official projections. The applications cover monthly inflation, direct quarterly macroeconomic forecasting, and mixed-frequency nowcasting. All headline exercises use historical-vintage information and final data as evaluation truth. Foundation models are generally more competitive than locally trained neural networks and, for selected targets and horizons, match or outperform simple benchmarks and official projections. These gains are not uniform. Official projections remain particularly strong when judgment, granular or advance information, and exceptional economic events matter, while persistent or irregular series can leave little scope for deep-learning improvements. Locally trained and native mixed-frequency models nevertheless add useful signals for selected real and labor-market aggregates. The results therefore favor using deep-learning forecasts as systematic, reproducible complements to official projections rather than as universal replacements. |
| JEL: | C45 C53 E27 E31 E37 |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:ptu:wpaper:w202612 |
| By: | Irene Aldridge; Lourenco Belleau-Bhowmik; James Chan; Josh Chang; Kelly Deng; Pranav Dhingra; Xinda Ding; Haodong Fang; Satya Hanumanchi; Carter He; Shuang He; Yadong Hou; Vallab Karanam; Rithvik Kasarla; Tishya Khanna; Jihyun Kim; Sarah Konety; Abhinav Kotta; Anderson Lo; Jake Luther; Omar Mian; Ishan Patwardhan; Matteo Perona; Chenyue Qi; Shashank Ramachandran; Shubham Rateria; Jake Swain; David Thai; Michael Twersky; Susan Wu; Dean Xu; Helenna Yin; Jiachun Zheng |
| Abstract: | Automated markets increasingly host autonomous, learning trading agents and are powered by formulaic automated market makers (AMMs). We study how heterogeneous agent behavior shapes stress, risk-adjusted performance, liquidity, and volatility in these types of markets. Three behavioral archetypes are modeled as agentic reinforcement learning (RL) systems: a rational arbitrageur, a manipulator, and a retail herd-follower. Each archetype is defined by its reward function within a common, asset-agnostic exposure-control environment, and each is trained with a value-based Deep Q-Network (DQN) and with REINFORCE using a learned baseline. A pilot on hourly cryptocurrency data (AAVE, ETC, SUI) motivates the hypotheses. Herd-followers produce the most frequent but shortest stress episodes, manipulators the fewest but most persistent, and rational agents the most stable rolling Sharpe ratios despite higher return volatility. SHAP attributions tie these patterns to drawdown and cross-asset signals, short-term volatility and position state, and momentum, respectively. The results motivate behavior-specific regulation rather than uniform limits on volatility or turnover. |
| Date: | 2026–10 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2610.22522 |
| By: | Arka P. Bandyopadhyay |
| Abstract: | I use a deep neural network (DNN) on a proprietary panel of commercial mortgages spanning 1998--2021 to separate liquidity-constrained default from strategic default -- two competing, observationally similar explanations for delinquency that are difficult to disentangle with standard parametric models. Conditional on delinquency, property income (net operating income, or NOI) is a more important determinant of the default rate than leverage (loan-to-value, or LTV), despite LTV's status as the workhorse variable in the commercial mortgage default literature. I recover borrower motive from otherwise observationally equivalent delinquency states through a joint analysis of default rates across NOI and LTV buckets, and show that the resulting thresholds are stable across the 2008 financial crisis and the 2020--2021 COVID-19 shock to commercial real estate. Because commercial borrowers are firms rather than households, and because responsibility for a missed payment is diffused across an organization rather than borne by a single decision-maker, the moral and reputational frictions that discourage strategic default in the residential setting are attenuated -- making commercial mortgages a setting in which strategic behavior should be comparatively easier to detect, and comparatively costly to ignore for investors, servicers, and regulators. |
| Date: | 2026–09 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2610.23901 |
| By: | Skander Lejmi |
| Abstract: | A market maker provides liquidity by quoting prices to buy and sell an asset, earning the spread while managing inventory risk. We study market makers trained with multi-agent reinforcement learning on prediction markets, specifically Polymarket, whose thin order books pose distinct challenges. Three challenges underpin this research: adverse selection may cause a policy trained on real data to stop quoting; the consequences of a quote can arrive many steps after the action; and prediction-market L2 data does not identify individual orders needed for simulation. We address these challenges by reconstructing order-level LOBSTER-format data from Polymarket and training agents in JaxMARL-HFT on 417 real markets. We enrich the baseline market maker's observation space with the book and recent flow, redesign its actions and inventory controls, and propose a cooperative multi-agent decomposition into a quoter and a risk manager trained using recurrent Independent Proximal Policy Optimisation (IPPO) with RNN policies. We investigate reward shaping, particularly reward blending and post-fill markout, to mitigate credit-assignment difficulties between the agents. Our redesigned single-agent market maker quotes more frequently and incurs smaller held-out losses than the retrained JaxMARL-HFT baseline, although the contributions of jointly changed observations, actions and risk constraints are not separately identified. Decomposition alone performs poorly; careful reward design substantially improves the desk relative to the unshaped split and yields our best result. Models are evaluated on realised end-of-episode portfolio value on held-out markets. None establishes positive average portfolio value in periodic held-out evaluation, and each configuration was trained with one seed. The environment, data and protocol lay a foundation for makers that quote many linked markets at once. |
| Date: | 2026–10 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2610.26789 |
| By: | Martin Bichler; Philip Hierhager; Kassian K\"ock; Fabian Pieroth |
| Abstract: | Repeated Bertrand competition is a stringent benchmark for algorithmic pricing: in homogeneous-goods markets, standard learning procedures typically select the stage-game Nash outcome with prices near marginal cost. We show that LLM pricing agents can escape this benchmark, depending on deployment decisions. LLMs outperform standard online learners through faster adaptation, but this advantage largely disappears against full-information learners. An extracted surrogate algorithm illustrates how LLMs in specific deployments profit from regularly undercutting slower learners. Agents that post one price every period from raw market histories that include rival prices can sustain supracompetitive regimes in self-play. Agents that periodically emit probability distributions from summarized histories of their own profits instead move toward the competitive region in self-play. Markets mixing these interfaces can generate dynamic regimes in which LLMs establish a high-price anchor and regularly undercut it to capture demand. These regimes can keep both competitors' prices above marginal cost. Deployment choices shape these outcomes, shifting the focus from stationary equilibria to dynamic pricing regimes. |
| Date: | 2026–10 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2610.21711 |
| By: | Carlos Montes-Gald\'on |
| Abstract: | Most empirical models of trend inflation treat it as a latent random walk driven by unobserved shocks, so they cannot say which economic conditions move it. Deep-$\pi^*$ makes the innovations to the trend a possibly nonlinear function of a large number of macroeconomic indicators, estimated with neural networks, so that Shapley values can be used to attribute every movement of the trend to the indicators behind it. Two complementary loss functions define a current trend, which fits inflation today, and a forward-looking trend, which best predicts inflation over the next year. In US data, employment-related variables were the largest contributors to the post-pandemic rise of the current trend. However, the forward-looking trend rose earlier, with real activity, employment, and money and credit as the main drivers, and peaked eight to twelve months before the current trend. Both trends looked through most of the movements in energy and commodity prices. The contributions of the indicators are nonlinear, with sizeable interactions between them. Re-estimated every month during the surge, the model forecast inflation more accurately than the standard UC-SV model and a random walk, and in early 2021, when professional forecasters expected inflation of 3% or less, the forward-looking trend was already signalling an increase in persistent inflation. |
| Date: | 2026–10 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2610.26298 |
| By: | Christopher Clayton; Antonio Coppola; Vanessa Hu; Matteo Maggiori; Jesse Schreger; Serdil Tinda |
| Abstract: | The increasing use of geoeconomic pressure is reshaping the global economy. We introduce a methodology using large language models (LLMs) to systematically measure the application of and response to geoeconomic pressure from large textual corpora, classifying which governments apply pressure to which foreign targets, using which in struments, firms, and products. Most importantly, we use LLMs to analyze how and why firms respond to the pressure. We focus on export controls, sanctions, and tariffs as the main instruments of pressure on firms. We find that export controls induce the targeted firms to undertake domestic R&D to build a domestic alternative technology. Western sanctions on Russia cause supply chain re-arrangements, the major beneficiary of which are firms in third-party countries, like China and India, which step into the market vacuum created by the broken business relationships. The threat and impositions of U.S. tariffs cause losses for the majority of U.S. firms, which report absorbing the cost shock on their inputs via their profit margin. Foreign firms are also negatively affected, mostly via the related macroeconomic uncertainty and the cost of relocating production. |
| Keywords: | artificial intelligence, LLMs, economic coercion, economic security, economic statecraft, economic dependency, chokepoints, sanctions, export controls, tariffs |
| JEL: | C4 F3 F4 F5 G3 |
| Date: | 2026–10 |
| URL: | https://d.repec.org/n?u=RePEc:bis:biswps:1386 |
| By: | Oguzhan Cepni (Center for Economics and Policy Studies, Istinye University, Istanbul, Turkiye; Ostim Technical University, Ankara, Turkiye; Department of Economics, Copenhagen Business School, Denmark); Rangan Gupta (Department of Economics, University of Pretoria, Private Bag X20, Hatfield 0028, South Africa) |
| Abstract: | We examine whether panel machine learning and mixed-frequency indicators improve fore- casts of quarterly real GDP growth across the 50 U.S. states. Using weekly, monthly and quarterly predictors, we compare six models under state-specific, global and grouped estimation at horizons of one to four quarters. The evaluation covers 2015--2026 and uses final- vintage data aligned with an assumed publication calendar. Pooled forecasts have roughly half the mean squared forecast error of a random walk benchmark over the full sample. Outside the pandemic window, the globally pooled forecast combination reduces MSFE by approximately 45 percent--56 percent relative to the random walk, with statistically significant improvements at each forecast horizon. Pooled implementations generally outperform separate state estimation at longer horizons, while differences among pooled structures are small. Model rankings vary across horizons and evaluation samples, and forecast accuracy differs across states. Extreme pandemic observations affect subsequent forecasts as they move through the mixed-frequency lag vector; excluding the collapse and rebound targets from estimation and limiting extreme predictors reduces this deterioration. The findings support combining timely regional indicators within pooled forecasting systems and evaluating their performance across horizons, states and economic conditions. |
| Keywords: | regional forecasting, panel data, mixed-frequency data, machine learning, forecast evaluation, GDP growth |
| JEL: | C53 C55 E37 R11 |
| Date: | 2026–10 |
| URL: | https://d.repec.org/n?u=RePEc:pre:wpaper:202632 |
| By: | Mingyu Guo; Jiayuan Liu; Vincent Conitzer |
| Abstract: | We aim to design revenue-maximizing single-item auctions that are deterministic, strategy-proof, and ex post individually rational --- the quintessential fundamental model for optimal mechanism design. While Myerson's seminal work solved this model for independent bidders, straightforward extensions to correlated settings often result in non-monotonic, and therefore non-strategy-proof allocations. Critically, optimal allocation design for correlated bidders is NP-hard to approximate, rendering theoretical optimality guarantees in this domain computationally intractable. We establish a new standard of empirical rigor for this theoretically hard model by proposing an empirical pipeline. We synthesize neural interpolation of Myerson's greedy allocation based on marginal profits (the correlated variant of virtual valuation), formal neural verification to enforce exact strategy-proofness, and a constructive repair procedure. Empirically, our method is highly effective, consistently achieves near-optimal revenue across a wide range of distributions, including synthetic, adversarial, and real-world (consumer, financial, and industrial) datasets. Compared to existing manual and neural baselines, our approach shows substantial improvement, often reducing the revenue gap by an order of magnitude. Our study is also the first to introduce (unattainable) greedy revenue as a rigorous upper bound for empirical benchmarking, providing a definitive quantitative performance measure where true optima remain theoretically elusive. We demonstrate the generality of our approach by extending it to multi-unit auctions with unit demand and integrating our verification techniques into RegretNet to achieve exact strategy-proofness. |
| Date: | 2026–10 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2610.24240 |
| By: | Varun Gupta; Vijay Kamble |
| Abstract: | When using a large language model (LLM) to simulate a user population, how should we combine individual behavioral histories to predict the fraction of users who take a particular action? We study this question in e-commerce, predicting observed product-level conversion rates from browsing sessions without task-specific model training. We vary two design choices, each involving a bias-variance tradeoff. Summarizing a product's sessions before inference reduces cost but can discard predictive detail, adding bias, whereas predicting outcomes for individual sessions and averaging afterward preserves that detail but requires many calls to reduce variance. Evaluating products together in one prompt can help the model infer what is typical in the setting, reducing bias, but also makes each prediction depend on which other products appear alongside it, adding noise. We evaluate both choices across five LLMs and two datasets. Our main finding is that predictions from individual sessions outperform predictions from summaries when products are evaluated separately, but summaries lead when products are evaluated together. Comparative summaries achieve the highest Spearman correlation with observed rates in 12 of 15 settings, with no alternative significantly better. Comparison thus helps summaries far more than session predictions, overcoming the advantage of behavioral detail. We interpret this difference through a Bayesian model of the second tradeoff, which suggests that comparison helps more when the inputs in a prompt reveal more about differences across products, consistent with our results for four types of comparative context. Comparative summaries also incur a small fraction of the inference cost of the session methods. With the best model for each outcome, they attain 66-89% of the correlation of the best supervised benchmark trained on thousands of labeled products. |
| Date: | 2026–10 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2610.23934 |
| By: | Kevin Foley; Jonathan Hartadi; Shivesh Prakash; Charishma Takkallapalli; Swapnil Vatsal |
| Abstract: | Language models that read financial news are usually trained on text from the same years used to test them, so a backtest can reward hindsight. We build news topics from NoLBERT, an encoder pretrained only on text published between 1976 and 1995, adapted into a sentence encoder by contrastive fine-tuning on passages of the same article, with no labels or returns. We compare it with LDA, MiniLM and FinBERT readers on 1.26 million articles from 2012 to 2024, convert each reader's topics into attention shocks and stock exposures, and forecast the returns of the 1, 000 largest US stocks with eight models at three rebalancing frequencies. Models are fitted on 2013-2019, choices are made on 2020-2021, and 2022-2024 is held out. The experiment with the best validation score loses money at every frequency. Measuring attention shocks against the past 63 trading days instead of 5 raises the test rank IC of every reader. NoLBERT gives the most stable topics and forecasts about as well as LDA. An equal-weight combination of all readers and models earns a monthly net Sharpe ratio of 1.12, significant against zero before adjusting for the search but not against a price-only benchmark. |
| Date: | 2026–10 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2610.22745 |