Minhyuk Nam

The foundational literature on cap-and-trade — Montgomery (1972), Rubin (1996) — establishes that emission permit markets achieve cost-effective allocations under competitive equilibrium. Competitive equilibrium, however, requires a Walrasian price-clearing mechanism that does not exist in practice: in the EU Emissions Trading System, firms must find counterparties through costly search, pay to access centralized platforms, and negotiate bilaterally in decentralized markets. Using the complete record of allowance transfers from the Union Transaction Log, we compile a new transaction-level dataset covering all three trading venues — government auctions, the exchange limit-order market, and over-the-counter bilateral trades — and document patterns inconsistent with the competitive benchmark: persistent cross-venue price dispersion, inefficient inventory holdings, and active financial intermediation. We develop a continuous-time search and matching model that captures these institutional features. Heterogeneous emitters choose which venues to access, how intensely to search for OTC counterparties, and how much to trade; symmetric financial intermediaries (banks) buy from surplus firms and sell to deficit firms, earning an intermediation spread that exists only because of frictions. The model nests competitive equilibrium — where banks are redundant — as a benchmark, and the gap between the two equilibria provides the metric for measuring the cost of frictions and the value of intermediation. We specify a parametric version of the model and establish identification of all structural parameters. Estimation by full-solution methods is in progress. The estimated model will enable counterfactual analysis of market design, including the welfare consequences of financial participation.

International climate agreements must delegate emission decisions to sovereign countries that hold private information about their abatement costs while managing a global externality whose damages depend on the cumulative stock of atmospheric carbon. We formulate this problem using dynamic mechanism design and characterize the structure of optimal agreements. Our two main results are that optimal treaties exhibit a cutoff structure in which countries with sufficiently low abatement costs are induced to reduce emissions, and we derive conditions under which simple “total carbon budget” mechanisms implement the social optimum.