The value of any financial asset, like money, is always and everywhere about expectations of the future. All financial assets are just promises, written on bits of paper. They are commitments about the future, and nothing more, and those commitments create expectations, and those expectations are what determine the demand for those financial assets. What happens right “now”, in the “current period” is always irrelevant, except insofar as it affects expectations of the future.
For monetary theorists who write down models, this creates another degree of freedom. You can say that a Fed action/statement of X will have an effect on expectations of Y, and this will have consequence Z for the economy. You can get just about any sort of result that way, and if you review the literature you will see that theorists have, indeed, gotten all sorts of results–one theorist can say that policy X will raise prices, and another theorist can say that policy X will lower prices.
Personally, I would use that degree of freedom to say that, to a first approximation, any Fed action/statement has zero effect on expectations. My view appears to be clearly false, in that investors hang on every word of the Fed, and Fed announcements often move markets–for a day or so. Then they go back to looking at other news. Whether there is any durable effect on markets is something you can believe or doubt, as you wish. I doubt.
Nick proposes an expected change that is beyond the range of his model, so once he allows that then anything goes, the model won’t work.
Does it add a degree of freedom, or just substitute a new one for an old one?
If the old one was a metric for ‘current period’ and the new replacement is a different metric for ‘expectations’, then there is equal constraint in terms of the space of answers a model can fit.
Of course, expectations cover a whole dimension of points in the future, so there could infinite possible pathways without appealing to some other constraint such as market data. Unfortunately, some of that futures market data is not as easily observed as current relative prices.
There are two meanings to durable, consistent or persistent. It isn’t hard to believe the effects may vary over time as does the meaning while having long lasting effects. Wealth itself is a long lasting effect even if subject to fluctuation.
When theorists disagree about whether X will raise or lower prices, do they only disagree about what effect X will have on expectations, or do they also disagree about what effect a given change in expectations will have on the economy? I mean, of course, a disagreement about the direction of change, not just the quantity.
It seems to me that PSST is compatible with an expectations-based view. And boom-bust cycles or other tumultuous economic resets are largely a function of ‘information shock’ in terms of an ‘expectations correction’.
Booms start for various reasons, and sometimes it’s political distortion and/or credit expansion. During a boom, people feel a little wealthier and more optimistic, and start engaging in new patterns of trade thinking they are sustainable. But eventually those beliefs prove to be incorrect, those patterns prove unsustainable, and a lot of economic activity suddenly comes to a halt and resources go idle. And it takes time to figure out how to put them to use in new sustainable patterns in a ‘new normal’, especially if people can’t calculate exactly what moves certain important actors will make because of political uncertainty.
This seems like a decent outline of the narrative of the recent crisis, especially as it relates to the housing bubble.
Honestly, I don’t see how one begins to explain a lot of economic behavior without resorting to some model expectations. A man doesn’t buy an insanely unaffordable house, or a bank extend a crazy loan to that man, because of the current price, but because of their expectations regarding the path of future prices.