Dean Baker – Mysteries of the AI Bubble: Where’s the Consumption?

What would happen if AI stocks continue to fall?

Dean Baker is the senior economist at the Center for Economic and Policy Research in Washington, DC. 

Cross-posted from CEPR

A digital bubble with tech icons and circuitry is about to be popped by a needle. Red line graph in the background indicates volatility.

Many of us who have been warning of an AI bubble have raised the prospect of a big hit to consumption if the bubble bursts. We base this concern on the conventional view that there is a stock wealth effect on consumption of roughly 3 cents on a dollar.

This means that if the market were to lose 30 percent of its value (it lost 50 percent in the collapse of the tech bubble), it would eliminate roughly $24 trillion in stock wealth. With a wealth effect of 3 percent, this would translate into a hit to annual consumption of $720 billion, a bit more than 2 percent of GDP. A hit of this magnitude, coupled with the collapse of the AI investment boom, would almost certainly be large enough to throw the economy into a recession.

Jason Furman has challenged this view, arguing that it seems that the wealth effect on consumption has been much smaller in the years since the pandemic, just 1 cent on the dollar. He points to the fact that the consumption share of GDP does not seem to be especially high at present, meaning that it doesn’t look like the huge increase in stock wealth in the last five years has had much impact on consumption.

This leaves us with a story where we either believe something has changed in the longstanding relationship between stock wealth and consumption or alternatively, that we are missing something. I’m going to explore the “we are missing something” route.

Specifically, I am not sure that people always see savings and consumption the way economists measure them. In this case, the issue would be how they treat the money they spend on buying a home and related expenses like property tax and insurance. In the GDP accounts, the purchase of a home is treated as investment, a form of savings. Similarly, the monthly mortgage payment, as well as property taxes and insurance, are not counted directly at all. They are supposed to be picked up through owners’ equivalent rent, since these expenses should be among the costs driving up rent.

The definitional issues don’t really matter in terms of accounting for consumption and saving, as long as housing costs track rents more or less closely, as has generally been the case. However, this has not been true since the pandemic. (It also wasn’t the case in the housing bubble.)

As can be seen, adjusted for overall inflation, the cost of buying a house has outpaced the cost of renting by around 20 percent since the pandemic. This means, other things equal, homeowners would be paying about 20 percent more on their monthly mortgage payments, taxes, and insurance than if house prices had just risen in step with rents.

Suppose that people acted as though this additional 20 percent they are paying on their mortgage and other related ownership costs were consumption and not investment. In the third quarter of this year, the imputed value of owners’ equivalent rent (OER) was $2,652 billion. If this were increased by 20 percent, in line with the gap between house prices and rent, that would increase spending on OER by $530 billion, or 1.6 percent of GDP. It would imply a corresponding reduction in savings. This adjustment would get us considerably closer to the 3 percent estimate of the wealth effect on consumption.

The relevant issue for a collapse of an AI-driven stock bubble is how a plunge in stock wealth would affect this picture. My speculation is that if we continue to see the stock prices of the big AI companies plunge, eager buyers for expensive houses would vanish, and some current owners look to sell to make up for their reduced stock portfolios. That should lower house prices, a good story in my book. Also, we may see some house owners look to cut spending in other areas if they don’t see their housing wealth as part of their targeted savings. (If house prices fall, then there would be a negative wealth effect on consumption from this as well.)

This story is clearly speculative, and I’m generally inclined to say that we should accept the accounting categories that God gave us. But when we have a seeming anomaly, like a sudden shrinking of the size of the stock wealth effect, it is worth trying to beat up our categories a bit to see if we might be missing something.

____________________________________

BRAVE NEW EUROPE has begun its yearly fundraiser for 2027. It is one of the very few Resistance Media in Europe. We publish expert analyses and reports by some of the leading thinkers from across the world who you will not find in state and corporate mainstream media. Support us in our work.

To donate please go HERE.

Be the first to comment

Leave a Reply

Your email address will not be published.


*