Motohiro Yogo

Firm characteristics, based on accounting and financial market data, are commonly used to represent firms in economics and finance. However, investors collectively use a much richer information set beyond firm characteristics, including sources of information that are not readily available to researchers. We show theoretically that portfolio holdings contain all relevant information for asset pricing, which can be recovered under empirically realistic conditions. Such guarantees do not exist for other data sources, such as accounting or text data. We build on recent advances in artificial intelligence (AI) and machine learning (ML) that represent unstructured data (e.g., text, audio, and images) by high-dimensional latent vectors called embeddings. Just as word embeddings leverage the document structure to represent words, asset embeddings leverage portfolio holdings to represent firms. Thus, this paper is a bridge from recent advances in AI and ML to economics and finance. We explore various methods to estimate asset embeddings, including recommender systems, shallow neural network models such as Word2Vec, and transformer models such as BERT. We evaluate the performance of these models on three benchmarks that can be evaluated using a single quarter of data: predicting relative valuations, explaining the comovement of stock returns, and predicting institutional portfolio decisions. We also estimate investor embeddings (i.e., representations of investors and their strategies), which are useful for investor classification, performance evaluation, and detecting crowded trades. We discuss other applications of asset embeddings, including generative portfolios, risk management, and stress testing. Finally, we develop a framework to give an economic narrative to a group of similar firms, by applying large language models to firm-level text data.Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

We generalize the closed-economy neoclassical growth model (CNGM) to allow for costly goods trade and capital flows with imperfect substitutability between countries. We develop a tractable, multi-country, quantitative model that matches key features of the observed data (e.g., gravity equations for trade and capital holdings) and is well suited for analyzing counterfactual policies that affect both goods and capital market integration (e.g., U.S.-China decoupling). We show that goods and capital market integration interact in non-trivial ways to shape impulse responses to counterfactual changes in productivity and goods and capital market frictions and the speed of convergence to steady-state.

Firm characteristics are ubiquitously used in economics. These characteristics are often based on readily-available information such as accounting data, but those reflect only a part of investors' information set. We show that useful information about firm characteristics is embedded in investors’ holdings data and, via market clearing, in prices, returns, and trading data. Based on insights from the recent artificial intelligence (AI) and machine learning (ML) literature, in which unstructured data (e.g., words or speech) are represented as continuous vectors in a potentially high-dimensional space, we propose to learn asset embeddings from investors' holdings data. Indeed, just as documents arrange words that can be used to uncover word structures via embeddings, investors organize assets in portfolios that can be used to uncover firm characteristics that investors deem important via asset embeddings. This broad theme provides a natural bridge to connect recent advances in the fields of AI and ML to finance and economics. Specifically, we show how language models, including transformer models that feature prominently in large language models such as BERT and GPT, can handle numerical information, and in particular holdings data to estimate asset embeddings. We provide initial evidence on the value added of asset embeddings through a series of applications in the context of firm valuations, return comovement, and uncovering asset substitution patterns. As a by-product, the models generate investor embeddings, which can be used to measure investor similarity. We propose a programmatic list of potential applications of asset and investor embeddings to finance and economics more generally.

Life insurers' business model has changed with the growth of insurance products with minimum return guarantees that are exposed to market and interest risks. The interest risk exposure of US and European insurers increased in the low-rate environments after the global financial crisis and the European sovereign debt crisis, respectively. The relative fragility of life insurers is highly persistent across the global financial crisis, the European sovereign debt crisis, and the COVID-19 crisis. European insurers with a higher share of liabilities with minimum return guarantees in 2016 had lower stock returns during the COVID-19 crisis.

Variable annuities, which package mutual funds with minimum return guarantees over long horizons, accounted for $1.5 trillion or 35% of U.S. life insurer liabilities in 2015. Sales decreased and fees increased during the global financial crisis, and insurers made guarantees less generous or stopped offering guarantees to reduce risk exposure. These effects persist in the low‐interest rate environment after the global financial crisis, and variable annuity insurers suffered large equity drawdowns during the COVID‐19 crisis. We develop and estimate a model of insurance markets in which financial frictions and market power determine pricing, contract characteristics, and the degree of market completeness.

Since the mid-1980s, the share of household net worth intermediated by US financial institutions has shifted from defined benefit plans to life insurers an

A large literature in asset pricing decomposes valuation ratios into expected returns and expected growth rates of firm fundamentals to understand why valuation ratios vary across firms and over time. This literature leaves two fundamental questions unanswered: (i) what information do investors attend to in forming their demand beyond prices and (ii) how important are various investors in the price formation process? We use a demand system approach to answer both questions. Empirically, we show that a small set of characteristics explains the majority of variation in a panel of firm-level valuation ratios across countries. We then estimate an international asset demand system using investor-level holdings data in Great Britain and the United States, allowing for flexible substitution patterns within and across countries. We use this framework to measure the contribution of each institutional type in linking characteristics to prices and long-horizon expected returns. Investment advisors, largely driven by their size, are most influential among all institutional types. Conditional on size, hedge funds are the most, and long-term investors (insurance companies and pension funds) are the least influential.