Fractional-reserve banking consists in banks lending out more money than they actually possess. This can also be seen as "money creation", or at least "money substitute creation"-- the claim on the bank (e.g. checking account deposit) is practically equivalent to cash-in-hand.
A short-seller borrows shares in a stock, sells them on the market, buys them back later, and returns them to the lender. This becomes a "naked short" if the borrowed shares "do not exist", i.e. are "share substitutes" credited to a short-seller's account without being backed by specific shares. Thus naked-shorting relies on a practice of "fractional-reserve brokerage".
Fractional-reserve banking is widely accepted while naked-shorting is widely considered unethical and is in most cases illegal. This confuses me because, per the above logic, they seem to be the same practice, just with a different asset class. In both cases, an abstract financial asset in the form of a claim on a backing asset (bank cash reserves or broker-held equities) is "created from thin air". In both cases, the existing stock of the asset in devalued by the creation of extra units, i.e. inflation.
Am I missing something? Are fractional-reserve banking and fractional-reserve brokerage, for lack of a better term, "structurally equivalent"? If not, what are the important differences in their economic effects?