Research

A Unified Model of High-throughput Blockchains,  with Yuxuan Lu, and Xi Chen

Abstract: We develop a model of coordination and allocation of decentralized multi-sided markets, in which our theoretical analysis is promisingly optimizing the decentralized transaction packaging process at high-throughput blockchains or Web 3.0 platforms. In contrast to the stylized centralized platform, the decentralized platform is powered by blockchain technology, which allows for secure and transparent Peer-to-Peer transactions among users. Traditional single-chain-based blockchains suffer from the well-known blockchain trilemma. Beyond the single-chain-based scheme, decentralized high-throughput blockchains adopt parallel protocols to reconcile the blockchain trilemma, implementing any tasking and desired allocation. However, unneglectable network latency may induce partial observability, resulting in incoordination and misallocation issues for the decentralized transaction packaging process at the current high-throughput blockchain protocols. To address this problem, we consider a strategic coordination mechanism for the decentralized transaction packaging process by using a game-theoretic approach. Under a tractable two-period model, we find a Bayesian Nash equilibrium of the miner's strategic transaction packaging under partial observability. Along with novel algorithms for computing equilibrium payoffs, we show that the decentralized platform can achieve an efficient and stable market outcome. The model also highlights that the proposed mechanism can endogenously offer a base fee per gas without any restructuration of the initial blockchain transaction fee mechanism. The theoretical results that underlie the algorithms also imply bounds on the computational complexity of equilibrium payoffs. 

Artificial Intelligence and Dual Contract, 

Presentations: 2024 American Economic Association (AEA) Annual Meeting

Abstract: With the dramatic progress of artificial intelligence algorithms in recent times, it is hoped that algorithms will soon supplant human decision-makers in various fields, such as contract design. We analyze the possible consequences by experimentally studying the behavior of algorithms powered by Artificial Intelligence (Multi-agent Q-learning) in a workhorse dual contract model for dual-principal-agent problems. We find that the AI algorithms autonomously learn to design incentive-compatible contracts without external guidance or communication among themselves. We emphasize that the principal, powered by distinct AI algorithms, can play mixed-sum behavior such as collusion and competition. We find that the more intelligent principals tend to become cooperative, and the less intelligent principals are endogenizing myopia and tend to become competitive. Under the optimal contract, the lower contract incentive to the agent is sustained by collusive strategies between the principals. This finding is robust to principal heterogeneity, changes in the number of players involved in the contract, and various forms of uncertainty.

Dynamic Contracting with AI, 

Abstract: This paper investigates a dynamic agency problem that includes AI-style agent and principal. We develop a quantitative bionic approach to dynamic contracting based on calibrating the incentive properties of a workhorse contracts model in which self-awareness and memory are endowed. We further solve the bionic model quantitatively via specific algorithms powered by Artificial Intelligence. In different scenarios, the bionic framework can endogenize various stylized preferences inspired by psychology studies and neuroscience. Note that the model theoretically rationalizes the neuro-foundations of a strand of conventional contract models in microeconomics, shedding new light on widely-debated issues surrounding the dynamic contracting problems.

Polynomial Factor Models, 

Abstract: Polynomial factor models (henceforth, PFM) represent a new class of factor models for high-dimensional panel data. We develop several econometric theories for factor models of latent factor interactions. Unlike approximate factor models (AFM), which are based on linear combinations of observed variables, PFM is based on polynomial combinations of latent factors. We show that PFM can capture nonlinear relationships between latent factors and observed variables and can be used to estimate latent factor interactions. We propose a general estimation approach for the PFM, labeled Polynomial Factor Analysis (PFA), to consistently estimate all the factor interaction-dependent factors and loadings. Their asymptotic distributions are established, and the proposed estimator is shown to be consistent and asymptotically regular. Simulation results demonstrate the effectiveness of the proposed method, and the empirical results show that PFA can provide more accurate estimations than the existing methods.

Factorization Asset Pricing, 

Best Paper Award of the 4th Prospective Economists Forum.

Abstract: We propose a new modeling approach for the cross-section of returns. Our model, Factorization Asset Pricing Model (FAPM), allows for predictor interactions by introducing second-order observable characteristics interactions regarding the unobservable high-order loadings. If the characteristics and expected return relationship are driven by compensation for exposure to the interactions between latent risk factors, FAPM will identify the corresponding high-order interactions between latent factors. If no such interactions exist, FAPM infers that the characteristics interaction effect is compensation without risk and allocates it to an "anomaly" intercept. Along with the investigation of returns and characteristics at the stock-level, we show that our factorization approach can be identified as a more accurate and logically transparent method among current existing methodologies, which is able to account for all interactions between predictors using factorized parameters. We also highlight that few predictors input can predict well, while numerous predictors set may generate negative effects due to adverse predictor interactions. Remarkably, irrelevant predictors in standard models may play essential roles in FAPM because their interactions with other predictors can be significant.

Robust q Theory,

Presentations: SFS Cavalcade North America Conference, Society for Economic Dynamics (SED) Annual Meeting, Econometric Society Winter Meetings, Midwest Finance Association (MFA) annual meeting, RES Conference, China Meeting of the Econometric Society.

Abstract: Models of the q theory typically assume that investments are determined by a specific approximating structured q model, hence ruling out perturbations due to a set of statistically nearby unstructured alternatives. This paper formulates a generalized framework, where concern to unstructured q models is admissible. By adopting relative entropy restriction for a set of unstructured alternative models, the model delivers generalized approximation to the ”truth” of the q theory, thereby resolving two main puzzles in conventional q models: (1) time-varying investment-q relationship, and (2) low investment despite high q. By exploiting polynomial specifications and novel measures of perturbations, I empirically demonstrate the critical importance of the approximation error driven by unstructured models.

Dynamic Patent Portfolio Management, 

Presentations: American Economic Association (AEA) Annual Meeting, Asian Meeting of the Econometric Society, North American Summer Meeting of the Econometric Society. 

Abstract:  I propose a tractable model integrating dynamic internal capital allocation with imperfect patent protection, thereby endogenizing patent war as well as financing constraints. I emphasize the central importance of capital intangibility for corporate decisions when intangibles are insecure. The main results are: (1) relative to the first- best, imperfect patent protection introduces the non-diversifiable patent risk as a new factor for internal capital misallocation; (2) the firm manages patent risk via patent portfolio management, litigation, and patent insurance; (3) the endogenous capital reallocation decreases the impact of the imperfect patent protection; (4) firms on the verge of liquidation tend to use patent litigation to alleviate financial distress; (5) dynamic technology choice play an important rule in patent portfolio management; and (6) imperfect patent protection creates the wedge between average q and marginal q, yet this wedge will be diminished in the scenario of high intangibility. Remarkably, this paper also extends the Modern Portfolio Theory to real investment sense.

Financial Flexibility Robustness, 

Abstract: I examine how the robustness of investment opportunities influence firm payout policy and cash holdings. By exploiting new measures, the perturbations of q, a novel counterintuitive yet reasonable fact emerge: low robustness of investment opportunities is able to spur firm propensity to pay dividends, lower repurchase shares, and decrease the cash a firm holds simultaneously. Specifically, firms that are likely to hold fewer amounts of cash when the robustness of investment opportunities is low, which is distinct from the standard channel of uncertainty. These results are consistent with firms’ liquidity management policies being significantly shaped by robustness concerns.

Fragile Robustness, 

Abstract: This paper examines the effect of robustness concerns on the formation of bubbles and crashes in the U.S. stock market. Although bubbles form in both the ambiguous and the risky environments, I find that stock prices do not crash in the ambiguous (low robustness) case, whereas they do so in the risky (high robustness) one. Namely, firms facing more knowable uncertainties, or risks, are more prone to stock crashes. These findings are consistent with recent theoretical work that highlights the critical role of robustness concerns in explaining surprisingly low stock prices and the high equity premium. 

Capital Allocation Based Asset Pricing, 

Abstract: I explore an intricate interaction between a firm’s risk exposure, intangible capital accumulation, and physical capital accumulation by using a unified dynamic investment model of capital allocation. The model emphasizes both the importance of the marginal value of the intangible capital and the idiosyncratic risk for corporate decisions, then implicates the firm value and expected returns. A model feature that the good idiosyncratic volatility is endogenously generated, but bad idiosyncratic volatility does not. The model provides several implications: (1) high bad idiosyncratic volatility lowers the firm’s profitability, expected returns, and Tobin’s q due to risk management and capital misallocation. Typically, this effect is much stronger in value firms; (2) there is a positive interaction between the value effect, profitability effect, and momentum. The idiosyncratic volatility can also affect the other anomalies indirectly via a complicated interaction between the corporate investment, internal capital allocation, and risk management, which implies that the idiosyncratic risk is priced indirectly as well; (3) high-tech firm invests more in intangible capital, holds more cash, and finds it much harder to achieve its growth firm region than low-tech firm. This result applies to firms with high idiosyncratic volatility as well.

Tenuous Information and Asset Returns, with Yunhong Yang

Abstract: This paper fills the gap by developing an ambiguity-averse investor model in which information asymmetry among investors discourages uninformed investors’ market participation, and market non-participation induces extra cost of capital for the firm. We find that when informed investors observe more private information, uninformed investors participate in the market less and the total cost of capital may take on an inverse U-shape.

Dynamic Feedback and Corporate Investment, with Liyan Yang