Taking the Korean experience as a laboratory experiment in systemic financial crises, the authors analyze distress in individual institutions among two groups of financial intermediaries. They pool together a group of large financial intermediaries (commercial banks, merchant banking corporations) and another group of tiny mutual savings and finance companies. Both the too-big-to-fail doctrine and the credit channel approach suggest that the probability of distress would be greater for small intermediaries. But the authors find that proportionately fewer small intermediaries were distressed than were large intermediaries. They offer two hypothetical explanations for this unexpected result: 1) Exchange rate exposure - a major shock to Korean intermediaries - was presumably negligible for the small financial intermediaries. 2) Small financial intermediaries allocated loans better, because of the "peer monitoring" natural to their mutual nature and deep local roots. Available data did not allow the authors to test the first hypothesis, but they did find support for the second one. Estimating a logit model, they find that the probability of distress was systematically smaller for the mutual savings and finance companies that stayed closer to their origins (for example, collecting many deposits as "credit mutual installment savings") and for those with a longer history of doing business in their local community.