Safety is the major problem of all the payment systems seen above.
What if the chosen cryptographic algorithm is not safe? Most algorithms used in these monetary systems have been around for many years already (DES [DP92], RSA [RSA78]), and some of the best cryptology experts tried to break them using sophisticated methods without success. Although one can never exclude the possibility that someone invents a new brilliant way to attack a given algorithm, chances that this might happen are still very small.
Another weak spot is the users personal hardware (e.g. the electronic wallet) and his copy of the software. Only complete physical security can guarantee that nobody can ease a user of some money.
The installation in the bank must be perfectly secure too. Just imagine the following:
In the system that DigiCash uses, digital notes can be identified by verifying the bank's signature on the note. The bank's signature can not be forged, since the bank is the only one that knows the secret key. Imagine what would happen if someone outside the bank would get hold of that secret key. He then would be able to create an unlimited amount of money, thus threatening to collapse the market. Or what if terrorists drop a bomb on the one central bank building that holds the secret key? A new key would have to be generated and all digital ``coins'' using the old key would have to be exchanged. A solution to that problem would be to make copies of the secret key and store them in different locations, but if several copies of the key exist, the risk that someone can get hold of one gets higher.
Another disadvantage is a possible uncontrolled growth of E-cash systems. Such a monetary explosion could undermine bank- and government-controlled money systems, giving rise to a confusing and inefficient Babel of competing systems.