Underpricing of new issues relates negatively to underwriter reputation in studies covering the US during the early 1970s until 1997 but positively in one study of IPOs from 1992-4. This paper investigates whether IPO underpricing depends on the organization of the financial system, whether underwriter reputation is a consistent indicator of firm quality and therefore (negatively) of underpricing, and whether this reputation effect also appears in completely different contexts. The study also looks for truncation in the observed returns distributions that may hint at price support activities on the part of underwriters. To answer these questions, the study presents evidence on new issues of stocks and their one-day returns in the Berlin market of 1882-1892 along with parallel data from the New York markets of 1998-2000. Despite what appear to be major institutional differences between the US and Germany, underpricing and its correlates are remarkably similar in the two cases: median underpricing is nearly the same in the two samples. Strikingly, the link between underwriter reputation and underpricing is positive both in the U.S. of recent years and in Berlin of the 1880s. This finding is in stark contrast with those for the US in the 1970s and 80s. The trade off between prices and rationing faced by underwriters might result in this instability in the reputation-underpricing link. Finally, the observed distributions of first-day returns for both markets display marked skewness toward positive values - a pattern that is consistent with left censoring and quite possibly with underwriter price supports. These results support a number of conclusions: first, either underwriter reputation is a poor signal of firm quality or firm quality is positively related to underpricing in certain circumstances; second, the largest and most prestigious underwriters may exert market power or at least provide more or better service in return for their higher indirect costs; third, the relationship between underwriter reputation or market share and underpricing clearly varies dramatically over time and across countries; and finally, financial system design-in particular, universal and relationship banking-may have little impact on the performance of new issues markets. Given the German results, one should conclude either that significant information asymmetries exist despite universality and formal relationships in the banking system or that information problems are unnecessary for the emergence of underpricing.