Kevin Erdmann on House Prices and Rents, 8/29
plus my thoughts
If you are interested in the U.S. housing market, then I strongly advise you to follow Erdmann’s substack. On the market’s current state, he writes
Homes have taken 6-7 months to construct for decades, until 2021. It currently stands at 9.7 months.
As he points out, this is symptomatic of supply bottlenecks. With fears of rising interest rates, you would think that if anything builders would be in a hurry to complete the homes that they have under construction.
Rents are rising at the low end (because of systematic, persistent lack of housing production). That is driving prices up. And, remember, when rents are higher, price/rents rise. For each 1% increase in rents, you always see more than a 1% increase in prices. So, as low end rent affordability gets worse, low end price affordability is even worser
I think he is basically right to point to rising rents as an indicator that the supply of housing has not been keeping up with the population. But I want to delve into the sentence that I bolded in the quote above. I agree that this seems to be the pattern, but it is not obvious to me that it should necessarily happen that way.
Suppose you want to buy a housing unit as an investment. You calculate
Profit = (rent) plus (increase in price) minus (interest cost)
If you can get $12,000 a year in rent, and if the price goes up by $8000 in the year after you bought it, and if the interest cost is $15,000, then you have a $5000 profit after one year. Note that the interest cost is there regardless of whether you borrow the money to buy the housing unit. Instead of buying the housing unit, you could have bought an interest-bearing security, and foregoing the earnings on that security is an opportunity cost. Note that rent should be net of management expenses and other costs, such as taxes or HOA fees. Note that the price increase should be net of depreciation and maintenance.
Putting everything in percentage terms, we have
Expected profit as a percent of price = (rent as a percent of price) plus (expected rate of price appreciation) minus (the interest rate)
Note that the formula uses the expected rate of price appreciation, because we do not know in advance what the appreciation rate will be.
In a competitive market, the expected profit should be just enough to compensate for risk. Let us simplify and say that expected profit should be zero. In that case, we can move the rent/price ratio over the left side of the equation, giving
rent as a percent of price = interest rate minus expected rate of appreciation
As an example, suppose that we have a $100,000 house with $10,000 rent, so that the rent/price ratio is ten percent. Suppose that the interest rate is ten percent and the expected rate of appreciation is zero.
Now keep the interest rate the same but let rent be $10,100, or one percent higher. How can the price go up by, say, two percent, to $102,000? Now the rent/price ratio is lower, at 9.9 percent. So to get the equation to balance, expected appreciation has to be 0.1 percent. 9.9 = 10 - 0.1
My point is that you get prices going up faster than rents when market participants are relatively optimistic about home price appreciation. So if rising rents always raise the ratio of prices to rents (the same as lowering the rent/price ratio), then rising rents must always cause market participants to be more optimistic about price appreciation.
I do not see why this should automatically be the case. It strikes me as something that would happen if people hold extrapolative expectations. That is, people expect rents and prices to increase more where they have gone up recently.
Extrapolative expectations are not fully rational. I would expect such expectations to lead to local housing bubbles followed by periodic reversals (the reversals might not show up in nominal terms, but only in inflation-adjusted terms).