Housing always goes two ways at once (part 2)

Yesterday, I posted these two charts, of the cross-sectional trends in the Supply component and the Cyclical component of home prices across the 30 largest metro areas.

Today, I want to follow up on that with these 3 aligned charts. The top chart is key here. It tracks real growth in rents (via the CPI) and in home prices (via the Case-Shiller index). Rising rents from inadequate supply create a distinct signature in prices across each metropolitan area, which my model picks up on. But, even though CPI rent inflation isn’t an input in my model, it all lines up fairly well with the trends in my components.

Here, I have color coded the top Fred graph to note where Case-Shiller prices or rent inflation were ascendant versus declining or moderate. (Highlighted with colored bands that are high when ascendant and low when declining) Those periods roughly match trend shifts in my components. The sweet spot that the American housing market really needs is a period of moderate home price inflation and moderate or declining relative rents. That hasn’t really happened at all in the past 27 years.

Mostly that is because rent (and supply driven price inflation) has only temporarily moderated because of a foreclosure crisis and one-time credit shock and then because of a pandemic. These are probably not long-term solutions, though I wouldn’t want to give NIMBYs any ideas.

National aggregate price/income levels (approximated by the middle Fred graph) temporarily appeared to return to 20th century norms by 2012, which was a combination of moderating rents and declining prices. The bulk of the “return to normal” price/income levels was a spike in underwriting standards away from long term norms - a one-time shift of 40 points in the median approved mortgage from about 720 to about 760.

In “Building from the Ground Up”, I laid most of the early decline in the 2006-2008 period on the Federal Reserve, with tight lending regulations taking over from 2009 onward. If I were to take one more editing pass on the book today, I would probably push the lending issue back a bit in time, so that it shares some of the blame for the accumulating crisis in 2008. At the end of the day, the Fed is the last mover, and more stimulus in 2008 would have helped boost housing starts and counter the wealth shock that was happening. But, it would be difficult for anyone to react to a multi-sigma shift in lending norms that even its architects seem not to have appreciated. If there’s no shift in GSE and FHA lending standards in 2008, damage related to Fed decisions would likely have been much more benign.

The first of my 4 paper series at Mercatus was posted today. (I’ll be writing more about that.) The topic of that paper is that rising rents have been the key driver of the unusually high national price/income level. Basically, up through 2021, within each metro area, the high end homes where affordability isn’t as important have been on discount while low end rents have been rising sharply, and bringing up prices with them. That’s one reason why the notion that low interest rates are driving home prices higher is hard to maintain when you look more closely.

There may be a shift toward reversing recent cyclical boosts, especially in a few cities that have especially boomed. But, we are already seeing a resurgence of rent inflation. And, I don’t think we will see the median credit score on new mortgages going from 760 to 800. So, there is no short term mechanism to lower the aggregate value of American residential real estate as shown in the middle Fred graph. And, certainly no mechanism to get from 1.6x GDP back to 1.1x GDP or less. That is a project that will take years, if not decades, of higher rates of construction.

The relative effect of cyclical vs supply factors is quite different from place to place, and that’s why you should subscribe to the Erdmann Housing Tracker to receive up to date estimates of the relative components in the metro areas of interest to you.