In the last couple of years, I have concentrated on two research areas, Trade Credit and Sustainable Finance, and am currently revising several single-authored articles. You can contact me for the latest version of each article, but I have chosen not to offer them publicly until the review processes are finalized. I have also just sent a manuscript for “The Future of the Chinese Economy” to its publisher, and the book should see the light of day before 2027.

On Sustainable Finance:

# The Dynamic Effects of Greenwashing

Sustainability can pay, yet benefit nobody. A dynamic discrete-time CARA-normal Bayesian-learning equilibrium with endogenous signal quality, in which genuine sustainable companies have higher cash flows, standard firms can pay a lifetime cost (certification, maintenance, expected penalties) to present as green when they are not but risk detection by a regulator, investors observe pool composition and detection events but learn the cash-flow effect only from the pool-averaged cash flow, and demand for green assets attracts greenwashing, indicates that any greenium is dissipated into certification and regulatory costs and committed firms subsidize greenwashers. A 1,835,008-draw Sobol-Saltelli polytope across thirteen dimensions identifies one binding parameter, the share of green investors. Greenwashing also impacts learning. The green-standard differential signal is attenuated by the actual fraction of sustainable firms, with a slowing posterior-variance decay biasing it toward zero.

# The Conditional Exclusion Intertemporal Capital Asset Pricing Model (CE-ICAPM)

In the CE-ICAPM, sustainable-mandate investors exclude certain assets, with their wealth share as an endogenous variable. Excess returns decompose into market risk, taste premium, exclusion premium governed by the Schur complement of the covariance matrix, and hedging premia. The central results are an identification asymmetry, correlation clustering that amplifies the 0.58% annual premium during crises, and possible welfare gains for unconstrained investors. Direct cross-sectional tests would require 143 years for 80% power, consistent with inconclusive carbon-premium estimates. Changes in wealth shares forecast excluded-minus-investable return differentials, with an exclusion-pressure test on simulated data showing 33–37% power and >90% sign accuracy.

#'”The ESG Premium is Dead; Long Live the ESG Premium”: The Dynamic Equilibrium Model of ESG Asset Pricing (DEAP)’

I develop a dynamic equilibrium model of ESG asset pricing (DEAP) integrating belief heterogeneity, preference heterogeneity, and endogenous wealth dynamics. Conventional and Sustainable investors trade market and ESG exposures under a Restricted-Beliefs Conditional Equilibrium (RBCE) and equivalent Heterogeneous-Bayesian Conditional Equilibrium (HBCE). Two distinct
wealth-share thresholds allow for perceived negative “greenium”, ESG funds outperformance over conventional ones and negative ESG factor alpha happening simultaneously. The DEAP implies that short-term ESG-return regressions are contaminated by Stambaugh bias. Wealth dynamics exhibit endogenous mean reversion. Empirical estimates pooling across wealth-share regimes depend on sample composition instead of an underlying “ESG premium”.

On Trade Credit

# The Theory of Trade Credit

Three axioms, firm optimization, bilateral bundling, diminishing returns, decompose bilateral surplus into real gains from exchange and a financial wedge times credit volume. The wedge splits into an SDF differential, an expected credit loss, and a covariance risk premium. Trade credit is self-limiting: extending credit narrows the wedge from both sides, distinguishing it from unilateral bank lending. A Modigliani–Miller benchmark follows when both firms are unconstrained and the payment rate is independent of volume. Four restrictions follow: cross-equation pricing, enforcement-recovery substitutability, dynamics separating extension from extraction, and an approximate Meltzer condition. The axioms classify the literature into three zones.

#A non-parsimonious integration of trade credit with moral hazard, supply chain networks, and financial intermediation

This paper develops a comprehensive yet non-exhaustive dynamic theory of trade credit within a fully axiomatic framework that generates empirical results consistent with the extensive empirical literature on trade credit. We establish a continuous-time model in which all results, from upstreamness and centrality effects to monetary policy transmission, are systematically derived from a non-parsimonious model that refers to a firm’s operations, market power, financing decisions, position in the supply chain, and currency arbitrage opportunities. Our framework endogenizes the components of working capital while accounting for input illiquidity, asymmetric adjustment costs, and strategic interactions within production networks. The model’s non-parsimonious nature is compensated by allowing us to integrate stylized facts and empirical results from the literature on trade credit under one umbrella: (1) upstream and central firms provide more net trade credit even when facing strong financial constraints; (2) trade credit serves as a buffer during currency crises and monetary policy shocks; (3) firms engage in financial intermediation by borrowing in foreign currency to fund trade credit in local currency; and (4) operating shocks trigger increased trade credit flows to preserve supply chain stability. It also allows us to develop new predictions not yet been tested in the scientific literature.

# When Does Trade Reputation Matter?

I develop a dynamic model of firm-level working-capital financing in which trade reputation is a firm-chosen, voluntary, depreciating intangible asset with growth-conditional payoff. Firms invest in costly reputation to expand supplier-credit capacity. In periods of slow growth, the carrying cost exceeds the financing benefit and the optimal policy is to let reputation depreciate. A closed-form growth threshold in three observable parameters, the trade-credit carrying cost, the reputation depreciation rate, and the capital share, partitions the firm distribution into a narrow active region where the mechanism relaxes the investment constraint. Using simulated data, trade reputation has a hump-shaped effect on aggregate investment. A counterfactual reform that raises bank-credit access indicates that firms that maintain reputation temporarily invest up to 22 basis points more in aggregate than firms that let it depreciate. The model generates null results on pooled trade-credit and investment regressions, alongside a firm-level internal-funds multiplier, concentrated where bank access is scarce, consistent with the empirical literature’s pattern of weak unconditional and stronger conditional effects. Trade reputation is usually an out-of-the-money option that pays off when investment growth is fast enough.