This paper analyses the pay-per-bid auctions which have appeared recently on the Internet andscored an immediate business success. In these auctions bidders pay a small, but irrevocablefee each time they want to increase the price. In this paper we test the model suggested by Platt, Price and Tappen (2010), which forecasts the distribution of closing prices depending on the item´s value, bid fee and price increment. The data from the Czech leading auction site Bonus.cz were chosen for the test.

more »

... Observed closing prices distribution of about 69 % of commonly auctioned items fits the model. However, we find some theoretical and practical flaws in the model. Contrary to the model predictions, we observed that auctions with smaller price increments generated significantly higher revenue than auctions with higher price increments. We suggest that bidders who favour skewness in payoff distribution cause auctions with lower price increment run longer and therefore explain this discrepancy.