Joe wants to buy a new car. He has bitcoin, but of course, he does not want to give up his $50,000 bitcoins expecting them to be worth $100,000 next year. Spending deflationary currency comes with it's own expensive repercussions, as anyone who as ever spent cryptocurrencies can attest. So Joe pawns (gets a loan on) some bitcoins. The bitcoins are locked into a smart contract. The bank advances Joe $50,000 on two $50,000 bitcoins. Joe pays back $1000 a month worth of stablecoins, and when he pays back the $50,000 his bitcoins come back to him, probably worth $250,000 by then. If he fails to make a payment, the bank gets his coins. There is a great incentive for Joe to pay, and essentially no risk to the bank. The bank can charge some nominal interest for an easy profit that in the end costs them essentially nothing.
Banks will not enter these deals, so what will happen if I locked my currencies in the bank and the price continues to drop, then the bank will lose.
Building on the assumption that the price will continue to rise is unrealistic.
Since we are talking about new technology, we will not go back to using traditional models such as banks. Rather, there may be portals that decentralize currencies and use smart contracts between different blockchains. Then the need for cash may be less than now.