Papers

Current portfolio construction methods are either agnostic to the effects of idiosyncratic shocks (standard factor models) or to the latent data structure driving systematic returns (recent graph-based approaches). This presents an opportunity to combine the complementary market aspects captured by the factor and graph domains, allowing asset allocations to operate directly on the underlying market structure, rather than on its observed co-movement or its finite-sample artefacts. In this work, we introduce the Mutually-INformed Graph-Locality and Exposures framework (MINGLE), which mutually regularises the factor and graph domains by redefining graph locality through systematic factor exposure profiles, rather than via observed co-movements. This is formalised through a unified Alternating Direction Method of Multipliers (ADMM) framework that jointly learns a latent factor representation and its induced graph topology directly from market returns. The resulting exposure-similarity graph aligns more closely with established economic sectors than conventional correlation-based graphs. Portfolios constructed from this representation are shown to consistently outperform their correlation-based counterparts across a range of volatility regimes and transaction cost levels. For rigour, paired statistical testing confirms that these gains stem from the reconciliation of the graph and factor domains.

I measure whether the rise of services is ultimately supported by household consumption or by non-household demand through production networks, investment, exports, and government. Using harmonized input-output tables for 76 economies over 1995-2022, I attribute service value added to its ultimate domestic final-demand category and measure the household orientation of each service economy in both levels and growth. The conventional service share does not reveal the type: it is negatively correlated with household orientation, and only about four in ten tertiarization episodes are majority-household. Among the 65 economies whose service share rose over 1995-2018, 77 percent became less household-oriented; the share is 75 percent under a two-percentage-point material-change threshold. Yet household-supported production becomes more service-intensive with income because households' aggregate position falls faster. A one-type benchmark has no purchaser margin; a two-type extension gives the condition for this reversal. An exact Leontief accounting and Shapley decomposition allocate a median 73 percent of service-share growth to production-system channels, which are the largest contribution in 80 percent of economies with materially rising shares. The decline in household orientation is instead allocated mainly to aggregate final-demand structure. The service share records how much service activity exists, not which economic transformation produced it.

Motivated by the rapid growth of data centers, we develop a model to evaluate bringyour-own-capacity (BYOC) mandates and flexibility accreditation in capacity markets for new large loads with shared supply-chain constraints. With efficient pricing, BYOC mainly reallocates procurement between grid-built and self-built capacity and therefore has little welfare effect, while flexibility delivers a modest gain by reducing the effective capacity requirement. Under administrative price caps, mandates can improve static welfare by forcing data centers to internalize the full cost of capacity. The welfare ranking of the two instruments depends on supply-chain stress. At low or moderate stress, only the flexibility instrument raises welfare. Under severe stress with capped prices, the welfare gain from the BYOC obligation can exceed the gross flexibility benefit. The two instruments differ in their effects on a neighboring market: a unilateral BYOC mandate can crowd out its capacity investment, while flexibility produces essentially no spillover at our calibrated benchmark. Finally, applying current capacity non-performance penalties to flexible loads may lead to financial incentives that are too weak to induce truthful flexibility reporting.

We study double auctions for markets in which goods are valuable in bundles, such as data, model weights, and fine-tuned AI assets. A key friction in such markets is No Assembly: a platform may be unable, for legal or technical reasons, to combine components supplied by different sellers into a single bundle. We formulate a combinatorial buyer's-bid double auction under this constraint. Under explicit stability and price-influence conditions (maintained in general, and for two goods derived from local price-taking and a feedback bound), each bundle submarket inherits the large-market discipline of single-good double auctions: bid shading vanishes, and clearing prices concentrate on competitive levels and track the common value (price discovery). The key incentive step, that bidding on a bundle creates no first-order strategic distortion beyond the single-good logic, is proved for two goods; for larger item sets it remains a maintained condition. Multi-agent reinforcement-learning simulations decompose the welfare loss and indicate that No Assembly, not strategic shading, is the binding finite-market friction, with both losses small in moderately thick markets and declining with complementarity amongst goods.

Using official government data sets of USA and France we analyze the occurrence/frequency/popularity distributions of given names on a time scale of more than 100 years. These distributions are characterized through the Lorenz and Pareto curves broadly used in the analysis of wealth inequality in the world. These curves remain stable during the considered time period with the Gini coefficient remaining in the narrow range 0.85-0.95. As for the case of wealth inequality, we show that the distributions of names are well described by the Rayleigh-Jeans (RJ) thermalization and condensation phenomenon well studied in various physical systems. The RJ thermalization results from two integrals of motion being analogous to energy and probability norm conservation in physical systems with energy states corresponding to popularity levels of names. Time correlations between names are also determined showing their stability until the middle of the twentieth century and a significant change after that.

The enormous growth in datasets, both in number and size, has prompted investors to adapt to new ways for assimilating information. Normatively, the approach has been to integrate such datasets into pricing formulations and assess the performance of portfolios created thereafter. However, such approaches underestimate their influence in portfolio investments by limiting their impact to pricing only. While being theoretically valid, this results in a potential sub-optimal performance in the presence of real-life decision constraints, and a blind spot for performance attribution. We start by analysing investment decisions from a knowledge perspective, which unfurls a new structure. We then propose a FinTech process termed Knowledge Optimisation that aims to integrate the influence of knowledge components that could be related to data, models, or business units that extract information. A 3-stage process, namely, decision structure, portfolio selection, and performance assessment is designed. We present an alternative to the ex-ante Sharpe Ratio, integrating a term for knowledge units. Through scenario analysis involving portfolio investment situations, we illustrate the utility. By design, the process improves the importance of knowledge in investment decisions.

Neural and numerical policy solvers can produce feasible controls even when the optimal rule and its binding constraints are unavailable. A primal-dual bracket certifies value loss, but it does not locate the optimal policy or explain which constraints genuinely bind. We show that, on the same declared simulation grid, one bracket can support both conclusions. For polyhedral controls, an exact conditional budget identity rewrites the residual as a pathwise nonnegative terminal Fenchel defect plus date-by-constraint complementary-slackness terms. A canonical Doob compensation removes the budget martingale that obscures small residuals. Bellman-primitive curvature conditions then yield an occupancy-weighted policy region with the sharp O(sqrt(G)) radius, while a paired constraint relaxation lower-bounds the optimal multiplier and certifies a binding face. A finite-sample resolution theorem quantifies the path budget needed to certify a target policy tolerance or face. Locked one-asset and two-asset audits cover every external policy error and make no false face declaration. An exact-wrapper stress test remains tight through 50 assets, while a separate state-dependent pilot identifies learned-dual tightness as the high-dimensional bottleneck. Reference solutions enter only after certification.

Predicting financial asset returns remains one of the most difficult challenges in empirical finance, driven by the low signal-to-noise ratio and the semi-strong form of market efficiency. While deep learning models, especially LSTM networks, have shown promise in capturing temporal dependencies, standard architectures often struggle to account for the cross-sectional heterogeneity of asset returns. This paper proposes a novel architectural extension to the basic LSTM model designed to improve both predictive accuracy and model interpretability. The framework integrates macro-financial covariates to capture broader economic signals and learnable sector embeddings to encompass heterogeneity by sector. The trading strategy involves constructing a long-short portfolio based on daily directional forecasts for each S&P 500 constituent, targeting stocks expected to under- or outperform the cross-sectional median return of the S&P 500. Model Performance is evaluated against three competitive benchmarks: a basic LSTM, a Random Forest model and a traditional market buy-and-hold strategy. The empirical results demonstrate that the LSTM with sector embeddings outperforms all benchmarks across key risk and return metrics. By utilizing sector embeddings, the model explicitly incorporates cross-sectional heterogeneity, allowing it to adapt to varying industry dynamics within the market. To address the black-box nature of deep learning, I use latent space visualizations to analyse how the model differentiates between sectors, providing insights into the internal representation of the sectors in the LSTM. The impact of the sector information can be quantified using a novel contribution metric by inspecting the weights of the LSTM. The predictive signal is driven by a short-term reversal factor and an industry momentum factor.

This study provides a method to quantify the benefits of shifting passenger traffic from air to high-speed rail from the perspective of flight-delay cost reduction. We first estimate the number of flight reductions for airport origin-destination pairs based on the high-speed rail ridership forecasts provided in the California High-Speed Rail 2020 Business Plan, and then distribute these flight reductions to quarter-hour intervals. Lasso models are applied to estimate the impact of reduced queuing delays at SFO, LAX, and SAN on arrival delays at the national Core 29 airports. These delay reductions are then monetized using aircraft operating costs and the value of passenger time. We evaluate alternative airport-capacity and flight-schedule scenarios, as well as multiple percentiles of probabilistic high-speed rail ridership forecasts. The resulting estimates indicate flight-delay cost savings of $51-88 million in 2018 dollars in 2029 and $235-392 million in 2018 dollars in 2033.

This article studies a dynamic corporate risk management problem by considering the decision-making of risk-averse managers who exert costly effort and select project risk. We study how a Value-at-Risk (VaR) constraint affects managerial decisions and the distribution of firm value when the manager's objective is non-concave with a fixed salary and options. By the concavification technique, we analyze the optimal terminal firm value on the concave envelope of the objective function. Applying the quantile formulation and the martingale approach, we can derive explicit solutions for optimal effort, terminal firm value, and project choice. The optimal terminal firm value can be divided into nine cases by carefully discussing the choices of VaR floor and tail probability. Compared with the benchmark case, we find that a VaR manager will smooth terminal firm value across states, reducing it in good states while supporting it in adverse states. Moreover, a VaR requirement generally improves downside protection and reduces bankruptcy probability when the VaR floor is low or moderate. However, when the VaR floor is sufficiently high, it can increase bankruptcy probability and induce gambling-for-recovery behavior in adverse states. Our sensitivity analysis indicates that greater managerial effort uniformly improves firm value. Moreover, more incentive options make managers more responsible, leading to a smoother terminal firm value across states. In contrast, a high fixed salary makes the manager less responsible and ultimately causes a more dispersed firm value.

Remote work has fundamentally reshaped urban economic life, and the spatial organisation of activity across cities. However, access to flexible work is distributed unevenly across industries, income groups, and genders, creating disparities in health risks, social mixing, and economic opportunity. Understanding where workplace dependence (WPD) is concentrated is therefore important, yet its distribution across urban areas remains poorly understood. Here we pair fine-grained hourly population data with detailed company records in a large European city to examine how location, industry composition, and socio-economic characteristics shape physical workplace attendance. By comparing workplace activity during periods of low versus high COVID-19 restrictions, we identify the determinants of WPD. We find that while industry and firm productivity are key drivers, the relationship between WPD, income, and gender is highly contingent on distance from the city center. Near the centre, female-majority and income-diverse locations show the highest WPD, consistent with a residual, place-bound service workforce. These findings reveal a spatially contingent 'service trap' at the urban core, extending remote-work inequalities beyond individuals to the urban ecosystem as a whole.

Intraday market manipulation is hard to detect because its footprint is brief, buried in millions of quotes, and statistically similar to ordinary volatility. Detectors reach high recall only by flagging so many other days that measured precision collapses, producing alerts no regulator can act on. We show that this manipulation leaves a distinctive dynamic signature: a pump-and-crash pattern visible in the velocity of market state, rather than its level. We build a minute-level detection pipeline, strictly partitioned in time, based on smoothed state velocity: option-Delta velocity for index options and price velocity for equities. We explain every alert with SHAP attribution. We hold the test period strictly out-of-sample and fix all thresholds before evaluation. On the locked Indian BANKNIFTY index-options test, the plain autoencoder recovers 10 of 10 regulator-identified manipulation days. Conditioning detection on market regimes inferred by a hidden Markov model yields an instructive negative result. The regimes are descriptively distinct, but using them trades recall for precision. Under the closed-world assumption that unlabeled days are normal, precision remains near 25%. The same dynamic appears in thinly traded U.S. equities (SEC v. Patel). The shape of the signature survives the transfer; its velocity magnitude does not. A pump-reversal shape score ranks the complaint's alleged manipulation days with AUC 0.91 (ARQQ) and 0.81 (ACY). On the ARQQ worked example, the score peaks inside the complaint's documented minute window. Finally, exact SHAP attribution over every alert shows that unconfirmed alerts share the regulator-identified days' attribution profile (cosine similarity 0.99). The precision ceiling is consistent with incomplete enforcement labels rather than detector failure. What transfers across markets and instrument types is the dynamic signature itself.

Counterfactual analysis aims to predict potential outcomes under hypothetical scenarios, offering valuable insights for decision-making. This paper investigates the application of large language models (LLMs), specifically the GPT-3.5 model, for counterfactual analysis. We focus on the online lending context, where the counterfactual return on investment (ROI) is crucial for evaluating different interest rate schemes. We begin by assessing the predictive performance of GPT and comparing it with advanced machine learning algorithms. The results show that prompt engineering can significantly enhance GPT's predictions, with the R-squared increasing from 1.97% to 2.84%, closely approaching the 3.48% achieved by gradient-boosted regression. Subsequently, we utilize GPT to generate counterfactual ROIs under a set of alternative interest rates. GPT exhibits logical coherence and causal reasoning in its responses. The findings underscore the potential of LLMs as effective tools for counterfactual analysis in online lending, suggesting broader applications for LLMs in various predictive and decision-making contexts.

We study utility maximization for high-frequency trading in fractional Brownian motion models. We first consider exponential utility in the discretized arithmetic model. Using spectral methods for stationary Gaussian sequences, we derive the asymptotic growth rate of the optimal certainty equivalent and the scaling limit of the optimal positions. We then consider power and logarithmic utility in a positive-price model with positive wealth and investment fractions in $[0,1]$. In this case, the logarithm of the optimal certainty equivalent grows at rate $n^{1-H}$ for $H\neq1/2$, with a coefficient determined by the one-sided prediction error of fractional Gaussian noise. We also give a simple binary strategy which attains this leading rate, without claiming exact finite-horizon optimality.

Equality before the law is a human right, and access to high-quality legal aid for indigent defendants is essential to enforce it. In a context where all defendants have access to a lawyer, I study the impact of denying legal aid on court outcomes. I combine double machine learning and a new administrative dataset linking aid to court outcomes in New South Wales, Australia, to learn the assignment function, whose inputs are known. I find that applicants who fail the means test and hire private lawyers are 10 percentage points less likely to be incarcerated than if they passed and relied on legal aid. Given an average incarceration length of nearly four years, this gap is significant. However, I find evidence suggesting that they spend more time in jail if they are incarcerated. A government preference for broad access to aid over allocated time per case could explain this pattern. Keywords: Indigent Defense, Crime, Criminal Justice. JEL: I30, K14, H44.

Cross-correlations between financial signals are neither scale-free nor amplitude-independent: they vary with the time scale over which they are measured and with the magnitude of the fluctuations that dominate the average. We exploit this structure to construct a portfolio allocation model in which the risk functional is the signed fluctuation function of multifractal cross-correlation analysis (MFCCA), indexed by a scale $s$ and a fluctuation order $q$. Unlike MFDCCA-type criteria, which rectify local detrended covariances before aggregation, MFCCA retains their sign, so that co-moving and counter-moving components contribute to risk with opposite signs; for $q=2$ the resulting quadratic form coincides with the detrended fluctuation function of the portfolio series itself, recovering the mean–variance criterion as a scale-dependent limit. Using two-component ARFIMA and Markov-switching multifractal processes, we show that prescribed multiscale and multifractal dependence is transmitted into the optimal weights, and that sign preservation contributes more to the reduction of tail risk than aggregation over fluctuation orders. Applied to financial multi-assets, the criterion lowers drawdown, Value-at-Risk, and expected shortfall relative to the mean–variance benchmark at every required return, in and out of sample, without any loss in realized portfolio return. The construction maps signed multiscale interaction structures onto resource-allocation decisions, and applies to any complex system whose components interact across heterogeneous scales with amplitude-dependent coupling.

The United States is a currency union where multiple risk-sharing mechanisms— migration, fiscal transfers, income diversification and credit markets—buffer consumption from local income fluctuations. We show that risk sharing not only directly smooths consumption but also indirectly stabilizes income by dampening the local multiplier. Combining causal estimates from regional military buildups with a multi-region quantitative model, we find that current levels of risk-sharing cut state-level consumption volatility by a factor of four. Crucially, the indirect stabilization of income accounts for nearly half of this effect, implying substantially larger benefits from integration than conventional measures suggest.

The impact of web datasets on market prices has suggested the development of new sources of information, such as social media and web portals, indicating the possibility of an emergent phenomenon. We propose a defining perspective, termed open information, that adds to the existing types of public and private information. We demonstrate their existence and justify material significance for pricing. In this respect, we present statistical hypotheses to test for a web dataset, GDELT, integrated for carbon pricing, that is represented by EU Allowance spot prices. Tests are designed with VAR and GARCH-X formulations, and return forecasting. The outcomes cannot rule out the material existence of open information. The result is significant for providing a conceptual basis to integrate a vast number of web datasets as alternative data in investment decisions.

This paper explores how financial institutions pass interest rate risk through to product markets using the life insurance industry as a setting. We show theoretically that it is optimal for insurers to distort product issuance across maturities to offset duration gaps. We examine insurers exogenously exposed to interest rate risk through their variable annuity liabilities after the 2008 financial crisis. Consistent with our mechanism, exposed insurers developed negative duration gaps, increased markups on long-duration products, and rebalanced product issuance toward shorter-duration products to hedge. This response reduced long-duration life insurance coverage by 12.1% of GDP between 2005 and 2023.

Control policies optimized in simulation can perform poorly in the real system when the parameters $x$ of the simulator are estimated from limited data but the resulting parameter uncertainty is not represented inside the simulation. A common way to incorporate such ambiguity is to simulate each trajectory of the system under a randomly drawn value for $x$. Since the policy cannot observe the drawn value, it must initially choose controls that perform well across many possible parameter values. However, if the policy progressively observes the system, it can often gradually infer the value of $x$, so that ambiguity vanishes. Over time, the policy then specializes to its estimate of $x$ and loses its robustness. This is undesirable in many real systems, where latent factors are expected to shift. In financial markets, for example, a policy hedging a derivative payoff should remain robust to changes in the volatility regime. To induce such continual robustness, we propose training policies in simulators where ambiguity varies with the system's state but does not systematically decay over time. We formalize this requirement as stationary ambiguity: the simulator should induce a stationary filter process over the latent state. We show how to construct such simulators and demonstrate, on hedging problems, that policies trained under stationary ambiguity preserve robustness to latent factors over time, leading to strong performance on real market data. As a modeling principle, stationary ambiguity informs many simulator design decisions: which models make realistic simulators, how their parameters should be randomized, and how simulator and policy should be initialized. While our experiments focus on hedging, stationary ambiguity may also be useful for other sequential control problems driven by exogenous stochastic processes with shifting latent structure.

Forecasting systems used in the Comprehensive Capital Analysis and Review (CCAR) and Current Expected Credit Losses (CECL) processes combine portfolio data, macroeconomic scenarios, model specifications, business assump- tions, and management adjustments. When the forecast changes from one run to the next, practitioners need an attribu- tion that reconciles to the total change without depending on an arbitrary sequence of input replacements. This paper formulates forecast-gap attribution as a cooperative game and examines several approaches: the exact Shapley value, hierarchical or nested Shapley values, Integrated Gradients, Gradient SHAP, Permutation SHAP, and Kernel SHAP. We compare their allocation rules, computational costs, implementation requirements, and limitations in production forecasting systems. The analysis provides a practical framework for choosing an attribution method according to the number and type of inputs, the feasibility of hybrid forecast runs, and the need for interpretability, reproducibility, and governance.

This paper studies the investment and insurance strategies of defined-contribution (DC) pension plans under the mean-variance framework. We consider a stochastic environment with time-varying interest rates, contributions, and mortality risk. The DC plan members are allowed to decide their bond and stock allocations, as well as their life insurance coverage. Adopting the martingale approach, we derive the closed-form optimal strategies and the mean-variance efficient frontier. Further numerical analysis investigates how mortality improvements affect investment and insurance decisions, as well as the sensitivity of the optimal decision to market parameters. Our analysis suggests that longevity raises expectations of future contributions, allowing pension members to adopt a less risky investment strategy. Meanwhile, insurance strategy shifts toward early adulthood to protect the high value of future income and decreases significantly at later ages. Moreover, we conduct sensitivity analyses on the target expected wealth, market price of risk, and contribution growth. These findings provide practical guidance for pension members on investment and offer insights for the design of DC pension plans.

The transmission of financial distress depends on the structure of exposures, the distribution of institutional buffers, and the duration of stress. We develop a dynamic model in which losses flow through weighted directed networks while counterparties are distressed, with recovery and repeated threshold crossings incorporated into the same system. A representation with \(K\) exposure factors closes the \(N\)-institution dynamics through \(K\) feedback coordinates, preserving heterogeneous sender and receiver roles. For bounded Lipschitz losses, Wasserstein stability and aligned \(L^1\) kernel estimates connect finite populations, factor models, and directed-kernel limits, with separate bounds for sampling and exposure approximation. For the hard threshold, every bounded nonnegative kernel admits a greatest cumulative-distress solution selected by vanishing positive-side regularization. The mass of institutions near the threshold governs stability: an Osgood condition along one reference path yields uniqueness, quantitative perturbation bounds, and convergence under sampled latent labels. Rank-one examples establish sharpness among uniqueness criteria based on threshold-layer mass, and a branchwise condition verifies the regularity directly from the initial profile and kernel. Numerical experiments examine how network structure and threshold concentration affect the dynamics. An application to disclosed EBA sovereign holdings constructs common-exposure factors and quantifies sensitivity to portfolio perturbations.

On two-sided content platforms, symmetric two-sided isolation (assigning matched fractions of creators and viewers to isolated treatment and control submarkets) is widely used for creator-side and cold-start experiments because it removes cross-arm marketplace interference. Isolation, however, thins each viewer's candidate catalog, and intuition suggests the resulting engagement cost should fade as the platform grows: a small fraction of a vast catalog is still vast. We show that, in an order-statistics model of engagement, whether this intuition holds depends on the upper tail of match quality. Extreme-value theory yields tail-class loss laws with a sharp dichotomy: for light or bounded tails the loss vanishes as the candidate pool grows, whereas under heavy tails it converges to a size-independent constant, so expanding the candidate pool, even by orders of magnitude, does not asymptotically eliminate the cost. Evidence from two production experiments on a platform with millions of active creators is consistent with this picture: a pure A/A traffic sweep reveals a measurable, depth-graded engagement cost; a one-sided catalog ablation independently shows that per-viewer thinning contributes to the loss; and a tail index calibrated on the small exploration pool predicts an effect consistent with the one observed in the far larger full-catalog ablation. Isolation thus carries a price that experimenters should budget for, like any other cost. We give practitioners a preflight procedure that estimates it before launch, sizes traffic accordingly, and recommends a fallback design when the predicted cost exceeds a chosen tolerance.

In this paper, I study to what extent countries differ in their preferences for quality and their technologies for improving quality. The paper also quantifies the contribution of those differences to the differences in gains from trade across countries. I adopt ANTONIADES, which allows endogenous quality choice of firms. This paper extends that framework into multi-country and multi-sector setting. The new feature is that bilateral trade liberalization yields spillover effects on the competitiveness of another country. This paper structurally estimates the model under multi-country and multi-sector framework. It is found that richer countries generally have stronger valuation for quality. The quantification demonstrates that variations in the strength of quality preferences and in the technology of improving quality across countries add to their heterogeneities in market competitiveness. Subsequently, I simulate a 5% increase in the trade barriers. If the quality channel is shut down, countries with stronger preferences for quality have larger degrees of underestimations in their losses from the trade barrier. Finally, gains from a universal rise in quality preference are unequal among countries, with larger economies generally gaining more than smaller economies.

Informed traders are supposed to need anonymity: they profit by hiding among the uninformed. A decentralized exchange now publishes the counterparty. Every committed order, cancellation, rejection, and fill carries a persistent pseudonymous wallet address. We reconstruct the full-depth limit order book from a record of 17.1 billion messages and 14.3 million aggressive orders by 147,113 wallets, covering $84.3 billion in taker notional. We report three findings. First, informativeness is a persistent wallet attribute. Wallets ranked by the price movement following their aggressive orders retain that ordering across adjacent ten-day windows, with a rank correlation of 0.52. Second, the ranking predicts returns. Adding the live activity of the highest-ranked wallets to a standard anonymous benchmark of prices, quotes, and order flow raises the out-of-sample R2 for one-second returns to 12.31%, a 13.2% gain (t = 9.2) that is 1.6 times the largest of 200 activity-matched placebo cohorts. Third, measured at realized trades rather than at every sampled moment, the increment grows from 1.43 to 2.47 percentage points of R2. Public wallet histories therefore carry short-horizon price information that anonymous order-book data leave unmeasured.

Risk-aware Q-learning (RaQL) provides a model-free, two-timescale estimator for dynamic risk objectives, but its finite-budget behavior remains fragile: fixed inner-loop hyperparameters can produce unstable value estimates, persistent Bellman residuals, and inefficient sample reuse. This paper proposes an adaptive training controller for Conditional Value-at-Risk (CVaR) RaQL and evaluates it on a daily Bitcoin trading task. The controller preserves the original CVaR estimator and Bellman fixed point; instead, it redesigns the training procedure through six coordinated mechanisms: per-cell inner-step sizing, outer-rate-matched decay synchronization, a short early correction for the VaR-like inner variable, a coverage-first-then-greedy sample allocation rule, progressive suffix aggregation of mature inner estimates, and data-driven calibration of key scales from online-observable quantities. Across 20 random seeds and 856,000 inner-transition samples, the controller reduces the mean empirical CVaR Bellman residual by approximately 85% relative to the fixed-parameter baseline (MeanBEQ: 1.2202 to 0.1854; MeanBEV: 1.1624 to 0.0535) and maintains stability across CVaR levels, discount factors, and training budgets. On the chronological out-of-sample test set, the learned policy attains a Sharpe ratio of 0.9281 with a maximum drawdown of 6.46% after transaction costs. Although buy-and-hold yields a higher cumulative return (35.43% vs. 23.61%), the adaptive policy achieves far lower volatility (9.57% vs. 47.93%), drawdown, and CVaR loss. These results demonstrate that adaptive finite-budget training design, applied solely to the training procedure without altering the risk objective, can materially improve the reliability and risk-adjusted performance of risk-aware Q-learning in financial applications.

Language models can propose many plausible trading factors, but an autonomous research system must also allocate its evaluation budget, verify its own evidence, and preserve how each candidate was produced. We present AgonAlpha, an architecture that searches over frozen research artifacts—hypotheses, executable expressions, platform evidence, rationales, and review status—rather than formulas alone. To our knowledge, AgonAlpha is the first alpha-mining system to combine verified artifact search, a fresh-context adversarial reviewer with re-execution and veto authority, and pending-aware parallel budget allocation, together with a complete public evidence trail. Independent deployments on WorldQuant BRAIN produced SPECTACULAR-grade alphas across five users and six model backends, with Fitness reaching 9.50 and Sharpe reaching 3.48, while retaining prompt-to-expression provenance for every submission.

This paper develops a unified mathematical theory of implied, local, and learned volatility surfaces. Total variance $w_t(k,\tau)=\tau\sigma_t^2(k,\tau)$ is an infinite-dimensional state constrained by positivity, calendar monotonicity, and the butterfly differential inequality. We establish the topology and tangent geometry of this arbitrage set and prove that a nondegenerate Gaussian shock at an active constraint exits with probability tending to one half. Exact invariance therefore requires tangency, reflection, or confinement to an arbitrage-free manifold. We separate this static invariance problem from dynamic no-arbitrage, derive the Musiela maturity-transport identity, and identify the additional fixed-contract martingale restriction. We formulate Hilbert-space dynamics, prove an exact modal reduction with closed-form truncation error, derive Karhunen–Loève factors, identify the portfolio derivative as a vega field, and obtain the covariance-optimal hedge $\alpha^\ast=(H^\ast C H)^{-1}H^\ast C\nu$. The local-volatility chart completes the geometry: Dupire local variance is the ratio $a=\partial_\tau w/g[w]$ of the calendar and butterfly constraint functionals. Neural operators provide arbitrage-free universal approximation through simplex and cone heads. Normalizing-flow maps then add tractable conditional densities: we derive exact change-of-variables formulae for exponential local-variance flows and invertible stick-breaking price-simplex flows, while stating the quasi-invariance conditions required in genuine function space. Finally, fading-signature fields encode surface history and yield autonomous finite-dimensional controlled dynamics. The result is one framework for representation, dynamics, arbitrage, dimension reduction, likelihood-based learning, simulation, and hedging, together with a falsifiable empirical protocol.

Financial sentiment classifiers are commonly evaluated against human labels, but strong linguistic performance does not necessarily imply economically useful return predictability. This study separates these questions through two experiments. First, we construct a unified three-class benchmark from five financial text datasets and compare TF–IDF Naive Bayes, off-the-shelf FinBERT and Financial-RoBERTa encoders, zero-shot Qwen2.5-7B, and QLoRA-adapted Qwen2.5-7B, LLaMA3-8B, and Mistral-7B models. Mistral-7B achieves the best test accuracy (0.8840) and macro-F1 (0.8771), while QLoRA raises Qwen2.5's macro-F1 from 0.7274 to 0.8615. An inverse-frequency class-weighted loss does not improve Qwen2.5. Second, we evaluate economic validity on a temporally separate 2019 Benzinga sample containing 10,637 unique headlines and 13,115 headline–stock observations for a fixed S&P~100 universe. Model probabilities are converted into continuous sentiment scores, aggregated by stock and signal date, and aligned with next-session returns over one-, two-, three-, and five-day horizons. All seven downstream models produce positive but small mean rank information coefficients at the one-day horizon; the largest is 0.0143 for FinBERT. None of the 28 model–horizon tests remains significant after Newey–West inference and false-discovery-rate correction. Portfolio results likewise fail to establish a robust advantage for the best-performing classifiers. The findings show that QLoRA is effective for financial sentiment adaptation, while also documenting a clear gap between classification accuracy and tradable cross-sectional signals.