The Bank of England staff recently published an article supporting the heterodox view that loans by commercial banks create deposits. That sparked a HUGE reaction round the world. There is however a sense in which loans do NOT CREATE deposits. It all hinges on what you mean by the word “loan”: a word that actually has two meanings.
First, there is something like “the transfer of real wealth by a lender to a borrower for a significant length of time with the borrower normally paying interest to the lender”.
Second, there is something like “a bank credits the account of a customer with $X so as to provide the customer with liquid working capital, or “day to day transaction money”, with the AVERAGE balance on the account remaining around $X.”
That distinction can be illustrated by reference to a barter economy in which a bank sets up in business. Citizens in that economy would deposit collateral at the bank and have their accounts credited so as to enable them to trade with each other using money instead of barter. But if that was their ONLY INTENTION, then the AVERAGE BALANCE on each person's account would remain at its initial level. That is, no one would have obtained a loan from anyone else in the transfer of real wealth sense of the word.
So in the latter scenario, the bank creates money, but not as a result of making a loan in the “transfer of real wealth” sense.
In contrast, where a citizen in the formerly barter economy wanted a loan in the transfer of real wealth sense, that wealth cannot come from the bank, because a bank is just a collection of office blocks full of employees, computers etc. The wealth can only come from other citizens who have decided to forgo the consumption of REAL WEALTH for an extended period and deposit the money they've earned at the bank. So in the latter scenario, deposits come before loans (contrary to the claims of the more naïve heterodox economists).
Moreover, in the latter scenario, no money has been created: that is, as long as those long term depositors have no intention of spending their money, what they've done in effect is to transfer their money to the borrower. That is, the bank acts as an intermediary between borrower and lender.
Indeed, the latter point is reflected in the way the money supply is measured in most countries: that is, so called money in term or deposit accounts to which the account holder does not have access for several months is often not counted as money.
Where a bank makes a loan in the “transfer of real wealth” sense, it can well be argued that no money is created. In contrast, where it makes a so called loan that simply consists of providing customers with day to day transaction money or liquid working capital, then money IS CREATED. But that's not a loan in the transfer of wealth sense.