Can a case be made that anonymous digicash is less risky (to a bank) than NON-anonymous digicash? In certain circumstances, it might be. Where a bank is at risk of violence when it does not reveal transaction information, not possessing such information poses less risk. On the other hand, in the USA a bank is at risk of violence when it does not possess transaction information. Would a Chaum-style anonymous digital cash service be more profitable to a bank than a NON-anonymous digital cash service? Maybe. It depends on what the demand curve for transaction services of various kinds looks like and what the relative demand for privacy is. If there were already a fully identified digital money system, then creating an anonymous digital cash system would grab you most of the market which was willing to pay a premium for privacy services. That, by the way, is not everybody. There will be at least a local maximum at some large premium, simply because certain benefits of bank secrecy are so large. On the other hand, there is likely also a local maximum where the premium is fairly small. In this case you get not only all the people above, but a large percentage of the people who are willing to pay just a little more for privacy. As to where these local maxima actually are, and which yields the larger profits, I have no idea. Are the costs involved in offering and supporting anonymous digital cash more, or less, than the costs associated with NON-anonymous digital cash? The costs associated with anonymous digital cash may well be less that for identified digital money systems. There are additional services being sold in most identified systems, including statements of transaction logs, reversibility of transactions, delay in settlement, and availability of logs to government. This last service is sold to the government with each transaction, a hidden fourth party which taxes the bank with the requirement to offer this service, in order to permit the bank to operate. These additional service take resources to operate. Reversibility, I suspect, is the most expensive to operate, since it's all human labor that can't be easily handled by computer. Digital cash, on the other hand, needs a redeemed note database, but this is one of its only unique costs. Since settlement is immediate, reversibility is not an issue, and neither is any delay in settlement. There are far fewer long term records to keep. It is likely that digital cash is more efficient economically, since it unbundles a bunch of previously linked services and allows them to be purchased separately by those who actually need them. Eric