At the individual level, nominal purchasing power is determined by nominal wealth. At the aggregate level, nominal purchasing power is determined by the supply and demand for base money.
People need to pay attention to the distinction between the micro level and the macro level. Bill Gates does not refrain by buying a Ferrari today because he doesn’t happen to be holding any base money. His nominal demand for goods and services is based on his nominal wealth. If the Fed buys $100 million in bonds from Gates and pays him cash, he doesn’t feel energized to go out and buy goods and services (his nominal wealth is unchanged), rather he puts the money in the bank and then invests it elsewhere. Of course the aggregate NGDP will rise a tiny bit, and for that reason everyone, including Gates, will spend a tiny bit more on goods and services.
If the Fed spends $100 trillion on financial assets, then nominal expenditure in the aggregate will soar (mostly due to inflation, but output will rise slightly in the short run.) Monetary policy operates at the aggregate level, it cannot be explained at the individual level, except by using the concept of the supply and demand for base money. The hot potato effect. But here’s the problem. In the short run prices are sticky, and individual people’s nominal purchases of goods and services depends on their nominal wealth. So we have to get from here to there via some sort of “transmission mechanism.” In the apple market we often assume a Walrasian auctioneer. But in macro that assumption assumes away all that is interesting. Hence people flounder, unable to conceive that monetary policy is all about debasing the medium of account by increasing its supply relative to demand. They want some sort of understandable mechanism that they can visualize at the individual level–the Keynesian interest rate, the Austrian Cantillon effect, but there just isn’t any. Or perhaps I should say there are far too many, each of which plays only a tiny role.
Imagine trying to explain to a wheat farmer why increased wheat production will lower the price. He’d say “I don’t notice any price drop when I sell more wheat.”
PS. I see people are still confused by Cantillon effects. It is not true that the Fed buying $100 billion in T-securities raises the price of T-securities, holding fiscal policy constant. That’s because if you hold fiscal policy constant the alternative would be to inject the money via already existing government spending programs. That would mean less borrowing. So the Fed’s decision to buy $100 billion in T-securities increases both the supply and demand for T-securities by $100 billion, leaving the price unaffected (relative to the alternative injection method.) Now obviously if you bought $100 billion in cars instead of T-bonds the car market would be impacted. But that would be due to fiscal policy, not monetary policy. The Richman quote that started all this specified that it mattered who got the new money (banks, bond holders, etc) because they would have more purchasing power. That’s flat out wrong.