Then there are not, by definition, "more projects chasing a shrinking supply of capital."
It can be true if it is a subset of “more projects”, if the decision mechanism works outside a market, or if that capital is accessible via mechanisms within the market which are not directly linked to borrowing rates, such as where the cost of capital was locked in years ago. Rates are not a panacea.
> This is supply-side inflation. You can still bring down inflation with rates by cooling demand. (Less free cash in the economy means less spending on AI means less demand for GPUs.) The downside is it costs you more output than in a balanced economy.
You said the same thing I said, just euphemistically. This is like saying you can rid your house of insects with a flamethrower.
Sure. Now do the next project that's going to lock itself in today.
> Rates are not a panacea
Nobody argued as much. Just that the effect has size and meaningful magnitude.
> You said the same thing I said, just euphemistically
No, I didn't. In an economy where private investment dominates consumption, you can reduce output and thus inflation without touching consumption, i.e. without impacting household budgets. You're giving up future jobs. But barring a large influx of new workers, something America doesn't have a problem with right now, unemployment will stay steady.
This is only true if the cost of capital is a functional constraint. If the constraint is elsewhere (in this case: sourcing GPUs, bringing online energy resources) it will not have any significant effect.
When the expected return is very high because the captured resource is perceived to be monopolistic, this lever does practically nothing. Unless they raise it so much they destroy everything else.