Abstract:

We introduce artificial intelligence pricing theory (AIPT).  In contrast with the APT's foundational assumption of a low dimensional factor structure in returns, the AIPT conjectures that returns are driven by a large number of factors.  We first verify this conjecture empirically and show that nonlinear models with an exorbitant number of factors (many more than the number of training observations or base assets)  are far more successful in describing the out-of-sample behavior of asset returns than simpler standard models.  We then theoretically characterize the behavior of large factor pricing models, from which we show that the AIPT's "many factors'' conjecture faithfully explains our empirical findings, while the APT's "few factors'' conjecture is contradicted by the data.

Abstract:

Topic models are a simple and popular tool for the statistical analysis of textual data.  Their identification and estimation is typically enabled by assuming the existence of anchor words; that is, words that are exclusive to specific topics. In this paper we show that the existence of anchor words is statistically testable: there exists a hypothesis test with correct size that has nontrivial power. This means that the anchor-word assumption cannot be viewed simply as a convenient normalization. Central to our results is a simple characterization of when a column-stochastic matrix with known nonnegative rank admits a separable factorization. We test for the existence of anchor words in two different datasets derived from the transcripts of the meetings of the Federal Open Market Committee (FOMC) - the body of the Federal Reserve System that sets monetary policy in the United States - and reject the null hypothesis that anchor words exist in one of them.

Abstract:

We study how shrinkage affects portfolio efficiency when the number of assets approaches or exceeds the sample size. Standard methods such as ridge impose uniform shrinkage, treating all assets as ex-ante identical and creating inefficiency when profitability is heterogeneous. Empirically, this “one-size-fits-all’’ design produces a hump-shaped relationship between model complexity and out-of-sample Sharpe ratios: adding assets can paradoxically reduce performance. We introduce shrinkage alignment, showing that efficiency requires matching shrinkage strength to each asset’s true profitability. Building on this insight, we propose Sharpe Ratio Shrinkage (SRS), a data-driven approach that aligns shrinkage intensity with empirical Sharpe ratios. SRS outperforms conventional methods under profitability heterogeneity and restores the "virtue of complexity" in high-dimensional portfolio construction.

Abstract:

We train neural networks to predict the trading volume of institutional investors, allowing us to evaluate the relative importance of hundreds of public information signals. On average, macroeconomic signals and announcement date indicators contribute the most to the model’s explanatory power (58%), while classical firm-level characteristics account for only 35%. We document large heterogeneity in which information sources different institutions pay attention to, and in their responses to the same information. Our results suggest that limited attention and differences of opinion—particularly about the state of the macroeconomy—are the two most salient mechanisms driving institutional trading decisions.

Abstract:

How adverse is the impact of fiscal policy uncertainty on economic and financial variables? To answer this question, we construct a novel cross-country database of news-based fiscal policy uncertainty indicators. Importantly, we track fiscal events that attract global attention, which we refer to as "global" fiscal policy uncertainty. We find that heightened fiscal policy uncertainty triggers contractionary effects, lowering industrial production in both advanced and emerging market economies. It also raises sovereign borrowing costs, generates synchronous movements in global financial variables including risk aversion, and strengthens the US dollar, even after accounting for US monetary policy shocks. 

Abstract:

We study the Latent Dirichlet Allocation model, a popular Bayesian algorithm for text analysis. Our starting point is the generic lack of identification of the model’s parameters, which suggests that the choice of prior matters. We then characterize by how much the posterior mean of a given functional of the model’s parameters varies in response to a change in the prior, and we suggest two algorithms to approximate this range. Both of our algorithms rely on obtaining multiple Nonnegative Matrix Factorizations of either the posterior draws of the corpus’ population term-document frequency matrix or of its sample analogue. The key idea is to maximize/minimize the functional of interest over all these nonnegative matrix factorizations. To illustrate the applicability of our results, we revisit recent work on the effects of increased transparency on discussions regarding monetary policy decisions in the United States.