100% money (1935): Fisher’s big insight on endogenous money

I am re-reading Irving Fisher’s 1935 “100% money” book for a workshop in Hamburg next month. The last time I read the book I thought that fractional reserve banking is a good description of reality. I don’t believe that anymore, but that is not the point I want to make. Re-reading Fisher (1935) I stumbled upon the following sentences, which are written under the title “‘Cash’ which is no cash” (p. 41):

Most deposits are created in the curious way just described – by lending. Sometimes a little actual cash passes through a teller’s window in one direction or the other – is borrowed and actually withdrawn, as for a payroll, or is deposited, as by a retail store which does a cash business. But typically and for the most part, checking deposits are manufactured out of loans, as in the imaginary example. In other words, nine-tenths of the depositor’s deposits can be made out of their own promises, with the help of the bank.

So, here is Irving Fisher in 1935 almost correctly describing financial alchemy. He supposes the bank start with a capital of $1 million, which in reality is not necessary, but still his view is one of endogenous money. Deposits are created by lending, he writes. (The missing 10% of deposits are created by lending the bank capital of $1 million, by the way.)

UPDATE 2013/10/01: I read on and on p. 129 Fisher clearly falls back into the usual loanable funds view: “But loans normally come out of savings, and the growth of the loans should depend on the growth of savings.” This, however, seems to be normative, whereas the quote from p. 41 seems to be descriptive. Is this the correct way to read Irving Fisher at his later stage? It seems to me he is arguing what we see can be described as endogenous money, but what we should have is loanable funds.