Recent analysis of policy-induced income changes along the agro-food chain in Bulgaria uses an extension of the PSE/CSE methodology. It shows that measures taken by the government with the intention of supporting consumers have led to a substantial negative impact on producers, without having a positive impact on consumers; that throughout the period 1990-1993 farmers have been net losers and retailers have been net gainers from agro-food policies; and that by the end of the period, both farmers and consumers were losing heavily to the benefit of processors and, overwhelmingly, retailers. The paper presents hypotheses to explain some of these counterintuitive results, using a more detailed study of the wheat-flour-bread chain. The first conclusion is that transitional excess profits and product quality improvements may explain part of the calculated transfers to retailers, but that this explanation is insufficient. A second conclusion is that the large net transfers to "retailers" are mostly rents collected by flour wholesale trading enterprises. A third conclusion is that "transition business practices", including monopolistic and collusive behavior in the agro-food chain, induce rents which are not included in the calculated transfers. Also, critical assumptions on conversion factors and exchange rates affect the results. While our analysis suggests that effective consumer taxation has been overestimated by the calculations, we support general conclusions on the inefficiency of the Bulgarian policies.