We build a theory of financial intermediation based on the premise that some investors are better able to figure out the trade motives of their counterparties in bilateral meetings—screening experts. We solve for the equilibrium market structure and study how information asymmetries stemming from heterogeneity in screening expertise shape up the core–periphery trade structure. In particular, the core of the market is populated by screening experts: they have the largest share of trade volume, they are actively engaged in middleman activity, and trade with the most counterparties. Using transaction-level micro-data and information disclosure requirements, we provide extensive evidence consistent only with our theory of financial intermediation.

Monetary exchange is called essential when better outcomes become incentive compatible when money is introduced. We study essentiality theoretically and experimentally using finite-horizon monetary models that are naturally suited to the lab. Following mechanism design, we also study the effects of strategy recommendations both when they are incentive compatible and when they are not. Results show that output and welfare are significantly enhanced by fiat currency if monetary equilibrium exists but not otherwise. Also, recommendations help if incentive compatible but not otherwise. Sometimes money gets used when it should not, and we investigate why, using surveys and measures of social preferences.