We model competition between banks and a bigtech platform that lend to a merchant with private information and subject to moral hazard. By controlling access to a valuable marketplace for the merchant, the platform enforces partial loan repayments, thus alleviating financing frictions, reducing the risk of strategic default, and contributing to welfare positively. Credit markets become partially segmented, with the platform targeting merchants of low and medium perceived credit quality. However, conditional on observables, the platform lends to better borrowers than banks because bad borrowers self-select into bank loans to avoid the platform's enforcement, causing negative welfare effects in equilibrium.

Using hand-collected data on robo-advisors, I study the impact of robo-advisors on the corresponding high-skilled (financial advisors) labor market. I find that robo-advisors and financial advisors are complements. This complementarity can be explained by the expansion in market for financial services through (1) an increase in financial advisors at firms that directly compete with robo-advisors in terms of services provided (relative to the rest of the firms) and (2) an increase in investor-level demand for financial advisors. I also find that the observed increase in the number of financial advisors is due to a reduction in separations and an increase in hirings. This is associated with an increase in the average experience of a financial advisor with no e

We propose a new framework for understanding non-fungible tokens (NFTs), crypto-assets that typically represent digital artwork. We posit that NFTs are digital Veblen goods: consumers demand them partly because other consumers do. Demand for NFT collections is thus fragile; issuers respond by underpricing their NFTs in primary markets, creating profit opportunities for "scalpers." We construct a simple model of NFT markets emphasizing social forces on demand and verify its predictions empirically. Our results have implications for redesigning NFT primary markets and for interpreting NFT returns.

This study evaluates the impact of BigTech credit on consumer spending, utilizing a unique dataset from a prominent BigTech ecosystem. In a nearly randomized context, we observe a 19% monthly increase in online spending among credit recipients. This increase is more pronounced for individuals with limited access to traditional financial credit, highlighting the role of BigTech credit in supporting financial inclusion. Moreover, the impact of credit is more notable in areas with more advanced logistics, illustrating the synergy between the financial and non-financial sectors of BigTech firms. Our analysis indicates that the uptick in consumption can be attributed to an increased frequency of purchases rather than to higher order values. Examining order-item level data, we find that credit recipients diversify their buying to include a wider variety of products and brands. Importantly, the provision of credit does not lead to a corresponding increase in discretionary spending or item pricing, and the heightened spending is not associated with increased delinquency, suggesting no overspending associated with BigTech credit.

The integration of algorithmic trading and reinforcement learning, known as AI-powered trading, has significantly impacted capital markets. This study utilizes a model of imperfect competition among informed speculators with asymmetric information to explore the implications of AI-powered trading strategies on speculators' market power, information rents, price informativeness, market liquidity, and mispricing. Our results demonstrate that informed AI speculators, even though they are ``unaware'' of collusion, can autonomously learn to employ collusive trading strategies. These collusive strategies allow them to achieve supra-competitive trading profits by strategically under-reacting to information, even without any form of agreement or communication, let alone interactions that might violate traditional antitrust regulations. Algorithmic collusion emerges from two distinct mechanisms. The first mechanism is through the adoption of price-trigger strategies (``artificial intelligence''), while the second stems from homogenized learning biases (``artificial stupidity''). The former mechanism is evident only in scenarios with limited price efficiency and noise trading risk. In contrast, the latter persists even under conditions of high price efficiency or large noise trading risk. As a result, in a market with prevalent AI-powered trading, both price informativeness and market liquidity can suffer, reflecting the influence of both artificial intelligence and stupidity.

Oracles are software components that enable data exchange between siloed blockchains and external environments, enhancing smart contract capabilities and platform interoperability. Using both hand-collected data from hundreds of DeFi protocols and market data for oracle networks, we find that oracle integration is positively associated with total value locked and platform/protocol valuation, triggered by positive network effects in adoption and usage. Our study reveals symbiotic gains from enhanced interoperability across protocols on a given chain and, depending on the mass of integrated protocols, among integrated chains. We also show that oracle integration improves risk-sharing and mitigates contagion; integrated protocols are more resilient than nonintegrated protocols during turbulent periods in crypto markets. We draw parallels between oracle integration and international economics, offering initial insights for regulators, entrepreneurs, and practitioners in the emerging space of decentralized finance.

Combining transaction-level data with survey-based information from a large consumer panel, we show that on average consumers form excessively high expectations about future income relative to ex-post realizations after unexpected positive income shocks. This systematic bias in expectations leads to higher current consumption and debt accumulation as well as a higher likelihood of subsequent default on consumer debt. A consumption-saving model with defaultable unsecured debt and diagnostic Kalman filtering with consumers who over-extrapolate income shocks rationalizes these findings. The model predicts excessive leverage and higher subsequent default rates compared to a rational expectations benchmark. Over-extrapolation of income expectations can contribute to explaining state-dependent household debt cycles.

We study the performance of IRA pension plans from 2004 through 2018. We document novel evidence of large return heterogeneity across income groups in the US, and provide estimates of its impact on wealth inequality. High-income individuals substantially outperform low-income ones, and this return differential is almost three times as large in “tax-free” Roth IRAs. These returns cannot be matched by equity market returns, but are consistent with high-income individuals having exposure to private assets.

We challenge two central tenets of lifecycle investing: (i) investors should diversify across stocks and bonds and (ii) the young should hold more stocks than the old. An even mix of 50% domestic stocks and 50% international stocks held throughout one’s lifetime vastly outperforms age-based, stock-bond strategies in building wealth, supporting retirement consumption, preserving capital, and generating bequests. These findings are based on a lifecycle model that features dynamic processes for labor earnings, Social Security benefits, and mortality and captures the salient time-series and cross-sectional properties of long-horizon asset class returns. Given the sheer magnitude of US retirement savings, we estimate that Americans could realize trillions of dollars in welfare gains by adopting the all-equity strategy.

Partial Ownership (PO), which allows households to buy a fraction of a home and rent the remainder, is increasing in many countries with housing affordability challenges. We incorporate an existing for-profit PO contract into a life-cycle model to quantify its impact on homeownership, households’ welfare, and its implications for financial stability. We have the following results: 1) PO increases homeownership rates. 2) Willingness to pay increases with housing unaffordability and is highest among low-income and renting households. 3) PO increases aggregate debt as renters become partial owners but also reduces the average leverage ratios as indebted homeowners become partial owners.

We propose a structural model of constant gain learning about future earnings growth that incorporates preferences for the timing of cash flows. As implied by the model, a cross-sectional decomposition using survey forecasts shows that high price-earnings ratios are accounted for by both low expected returns and overly high expected earnings growth. The model quantitatively matches a number of asset pricing moments, as learning about growth interacts strongly with the preference for the timing of cash flows, and provides insights on the roles of risk premia and mispricing in the cross-section of stocks. The magnitudes and timing of the comovement between prices, earnings growth surprises, and anomaly returns are all consistent with a gradual learning process rather than expectations being highly sensitive to the most recent realization. Large earnings growth surprises do not immediately translate into large one-period returns, but instead are gradually reflected in future returns over time.

Student loans with income-contingent repayment insure borrowers against income risk but can reduce their incentives to earn more. Using a change in Australia's income-contingent repayment schedule, I show that borrowers reduce their labor supply to lower their repayments. These responses are larger among borrowers with more hourly flexibility, a lower probability of repayment, and tighter liquidity constraints. I use these responses to estimate a dynamic model of labor supply with frictions that generate imperfect adjustment. My estimates imply that the labor supply responses to income-contingent repayment decrease the optimal amount of insurance but are too small to justify fixed repayment contracts. Moving from a fixed repayment contract to a constrained-optimal income-contingent loan increases welfare by the equivalent of a 1.3% increase in lifetime consumption at no additional fiscal cost.

What explains cross-sectional dispersion in stock valuation ratios? We find that 75% of dispersion in price-earnings ratios is reflected in differences in future returns, while only 25% is reflected in differences in future earnings growth. This holds at both the portfolio-level and the firm-level. We reconcile these conclusions with previous literature which has found a strong relation between prices and future profitability. Our results support models in which the cross-section of price-earnings ratios is driven mainly by discount rates or mispricing rather than future earnings growth. Evaluating six models of the value premium, we find that most models struggle to match our results, however, models with long-lived differences in risk exposure or gradual learning about parameters perform the best. The lack of earnings growth differences at long horizons provides new evidence in favor of long-run return predictability. We also show a similar dominance of predicted returns for explaining the dispersion in return surprises.

We develop an approach to determine whether a particular predictor represents a proxy for fundamental risk. We build on the assumption that risk-based predictors should be linked to new information about economic conditions. We show that most predictors forecast returns on either days with macroeconomic announcements or the remaining days, indicating that sources of return predictability differ across predictors: few are driven by fundamental risk; most have other origins. We show that Shiller’s excess volatility is confined to non-announcement days, suggesting that the ability to forecast stock market’s noise component underlies much of the predictability documented in the literature.

Present-value logic says that aggregate stock prices are driven by discount-rate and cash-flow expectations. Dividends and net repurchases are both cash flows between the firm and household sectors. Aggregate dividend-price ratios do not forecast dividend growth, but do robustly forecast future buybacks and issuance. Long-run variance decompositions say that discount-rate and cash-flow expectations contribute equally to aggregate dividend-price-ratio variation.

We show that an increase in passive exchange-traded fund (ETF) ownership leads to stronger and more persistent return reversals. Exploiting exogenous changes due to index reconstitutions, we further show that more passive ownership causes higher bid-ask spreads, more exposure to aggregate liquidity shocks, more idiosyncratic volatility, and higher tail risk. We examine potential drivers of these results and show that higher passive ETF ownership reduces the importance of firm-specific information for returns but increases the importance of transitory noise and a firm's exposure to market-wide sentiment shocks.

This paper addresses a key question in economic forecasting: does pure noise truly lack predictive power? Economists typically conduct variable selection to eliminate noises from predictors. Yet, we prove a compelling result that in most economic forecasts, the inclusion of noises in predictions yields greater benefits than its exclusion. Furthermore, if the total number of predictors is not sufficiently large, intentionally adding more noises yields superior forecast performance, out- performing benchmark predictors relying on dimension reduction. The intuition lies in economic predictive signals being densely distributed among regression coef- ficients, maintaining modest forecast bias while diversifying away overall variance, even when a significant proportion of predictors constitute pure noises. One of our empirical demonstrations shows that intentionally adding 300 ∼ 6, 000 pure noises to the Welch and Goyal (2008) dataset achieves a noteworthy 10% out-of-sample R2 accuracy in forecasting the annual U.S. equity premium. The performance sur- passes the majority of sophisticated machine learning models.

Numerous trading strategies examined in options research exhibit remarkably high mean returns and Sharpe ratios. We show some of these seemingly ``good deals'' are due to look-ahead biases. These biases stem from using information unavailable at the portfolio formation time to filter out observations suspected of being noisy or erroneous. Our results suggest that elevated Sharpe ratios may serve as potential indicators of such look-ahead biases. Furthermore, deviating from previous literature findings, we show that illiquidity is not strongly priced in stock options and that only a small set of stock characteristics are in fact associated with option expected returns.