As I understand Chaum's Digi-Cash scheme the bank customer is protected both against bank fraud and merchant fraud but not against collusion between the two. The customer need not be unknown to the bank in order that her purchases with bank notes be unknown to the bank. This is where blind bank signatures come in. Neither need the merchant be unknown to the bank. Remailers are unnecessary to hide connection between the merchant and his customers. The bank knows both but has no way to associate them. This being the case, the bank customer is protected against a false denial of deposit. When one deposits Digi-Cash in a bank he presumably waits for an indication from the bank that the cash was still valid (not previously deposited). The customer retains this indication as a receipt which is signed by a private bank key. A bank's positive reputation is useful but not necessary at this point. Another concern is the false denial of payment by the merchant. If the merchant's customer is willing to reveal to a judge that she paid the merchant, and she retained the note with which she paid, and the bank retains a record of the merchant's deposits then she can argue that it was she that caused the bank to mint the note since only the person for whom the note was minted and the merchant who had cashed it (and the bank) should know its bits. This would require that the bank respond to the judicial query "has bank customer M deposited note x?".