This paper explores how lender market structure affects the efficiency and equity of financial taxation, an important revenue source and policy tool for governments worldwide. Using a natural experiment—the unexpected introduction of a loan transaction tax in Ecuador—this research employs pass-through estimates, a quantitative model, and a comprehensive commercial loan dataset to investigate this issue. The model broadens the scope of traditional bank competition theories by allowing for a range of competitive behaviours, including joint profit maximization, credit rationing, and Bertrand-Nash competition. Contrary to the common assumption of fully competitive differentiated lending markets, this paper finds little evidence to support pure Bertrand-Nash competition or credit rationing. Instead, the results are more consistent with joint profit maximization among banks. While the researchers find that loan taxes are indeed greatly distortive, neglecting the possibility of uncompetitive lending inflates estimated tax deadweight loss by approximately 80 to 120%. This distortion occurs because non-competitive banks internalize a portion of the tax burden. Conversely, subsidies are less effective in non-competitive settings. Findings suggest policymakers consider the interplay between market structure and tax-and-subsidy strategies.