This article is a long-winded way of saying the housing measure in the Feds inflation index calculation is a lagging indicator (because it is more of a long-term average instead of short-term snap shot). You do a good job explaining that and suggest that a different number might be better (though possibly “noisier”). I don’t know enough to argue for or against a change, but I do have a question that the article does not address.
The Fed managers surely know that it is a lagging indicator, is it your position that they are too stupid or too rigid to let more time sensitive data affect their decision? Or is it obvious (from past moves) that they have not considered it? In other words, the measurement has its flaws (like every other aggregate index) but would a change make a difference in the decisions that are made?
I’m not trying to me argumentative or dismissive of your point. It is an honest question because I don’t know anything about what drives the Fed.
Focused on housing sales and related items, while also making clear that housing markets vary wildly across geographic areas even though there is some coupling to broader economic trends.
The Federal Open Market Committee—the group that determines monetary policy—does forecast inflation among other economic predictions to guide their policy. That includes estimating future rent contributions to PCE using data such as market rents. For example, here’s the conclusions from a Feb 2022 letter from the SF Fed predicting higher inflation through sustained increases in the rental component due to higher market rents. [1]
> As the U.S. economy recovers from the effects of the COVID-19 pandemic, some increase in rent inflation should be expected, given that landlords can ask for higher rents when prospective tenants are employed and earning higher incomes. However, the extraordinarily large increases in two leading indicators of future rent inflation—asking rent inflation and house price inflation—point to significant upside risks to the overall inflation outlook. The potential increases are particularly significant for CPI inflation, which places a larger weight on shelter costs. Still, the potential additions to PCE inflation of about 0.5pp for both 2022 and 2023 are important to consider in light of the Federal Reserve’s 2% inflation target.
So the Fed is to some extent considering how current market rents will impact future inflation and using that consideration in developing monetary policy. It could be argued that the FOMC members are not putting sufficient weight on this specific consideration. Additionally, their forecasts may incorrectly estimate the future inflation contribution of rents and misguide policy. For example, in May 2022 the Richmond Fed revisited Aug 2021 predictions about future inflationary contributions of rents made by the Dallas Fed and found the predictions significantly underestimated inflation. [2]
> Growth in the PCE rent index exceeded 4 percent year over year in March, a threshold that the model predicted would not be crossed until the second quarter of 2023. Similarly, imputed rent for homeowners rose 4.4 percent in March, which the model predicted would have been attained in the first quarter of 2023. Year-over-year growth in both rent and imputed rent was over 1.5 percentage points higher than the upper confidence interval of the forecast in March.
It may simply be that during periods of high economic volatility it is hard to estimate the future trajectory of inflation. Any predictive model developed on historical data may simply fail to extrapolate to the specific novel economic forces that cause the volatility.
In retrospect it’s easy to say that the Fed should’ve better extrapolated how surging market rents in 2021 would lead to sustained inflation as leases rolled over. But had the sudden surge been followed by a comparable collapse in rents then such predictions would also have looked foolish. The core problem may be the inability to predict the evolution of our economy during periods of high volatility.
Proof of average-rent housing disinflation: Yesterday my landlord sent me the renewal/rent increase notification for 2022-23. It was only a hair over 4%. (This is about in line with previous increases since I moved in in 2017.)
The question dominating US monetary policy is who will win the 2024 presidential election. One of the leading candidates attempted a putsch last year and failed. The governors of the fed are highly educated elites from coastal bubbles. They care more about preserving our form of government than technical measures of inflation. They’ll try to force a quick correction in order to get a boom going by 2024.
One of the issues with using mortgage costs as part of the index is that the higher interest rates used to combat inflation lead to inflation in mortgage costs.
I feel badly for Jerome Powell, who seems like a perfectly pleasant guy who essentially did what ‘everyone’ thought was right but now likely will be remembered as one of the worst Fed Chairs in history.
Tough to see your legacy disintegrate in real time, I would think.
Good points. But it ignores what the Fed COULD pay more attention to: the Treasury Inflation Protected Securities -- TIPS -- expectation rate. It DOES include expectations concerning food and fuel and it enlists the crowdsourcing abilities of (potentially) million of participants in the bond markets. TIPs signaled that people had ceased believing that eh Fed would achieve it 2% target at least by September and the Fed did not move until November. TIPS would be even more useful if Treasury would issue them in more tenors, 1,2,3,7 15 and 30 instead of just 5 and 10.
"And by the same token, the insane surge in spot rents in the summer of 2021 was not a sudden and dramatic blow to Americans’ living standards because, again, very few people actually signed new leases during that moment. The aggregate ups and downs of housing costs are more gentle, as the BLS says."
The cratering of prices in 2020 decreased overall price appreciation by what looks to be 45% or so and then the 2021 run up in prices increases it by what looks like 150%. in BLS terms.
Given that we have the Apartment List percent numbers, I think you can calculate what percentage of the rental population signed new leases during this time, and it seems like it should be a significant number if the overall rate of price appreciation for all apartment rents is oscillating over a 200% spread.
Edit:
In fact: Isn't it as simple that a 2% runup in BLS price during a 10% increase in Apartment List price means that 20% of the rental population has signed to pay the newly increased price during the time period being measured?
I've never liked the concept of OER, beginning when I learned about it in college econ. It just seems too homo economicus. Plus it seems to involve a lot more guessing than measuring. And if the majority of people live in homes they own, shouldn't we assume that if everyone rented instead that rental prices would dramatically fall, and therefore OER is not representative to begin with?
The problem is potentially even worse. This paper finds that raising interest rates causes rents to temporarily go UP as raising the cost of borrowing pushes some people out of buying and into renting. The trend likely doesn’t reverse until you get true recessionary conditions.
“ in contrast to house prices, housing rents increase in response to contractionary monetary policy shocks. We also find that, after a contractionary monetary policy shock, rental vacancies and the homeownership rate decline. This combination of results suggests that monetary policy may affect housing tenure decisions (own versus rent).”
This article is a long-winded way of saying the housing measure in the Feds inflation index calculation is a lagging indicator (because it is more of a long-term average instead of short-term snap shot). You do a good job explaining that and suggest that a different number might be better (though possibly “noisier”). I don’t know enough to argue for or against a change, but I do have a question that the article does not address.
The Fed managers surely know that it is a lagging indicator, is it your position that they are too stupid or too rigid to let more time sensitive data affect their decision? Or is it obvious (from past moves) that they have not considered it? In other words, the measurement has its flaws (like every other aggregate index) but would a change make a difference in the decisions that are made?
I’m not trying to me argumentative or dismissive of your point. It is an honest question because I don’t know anything about what drives the Fed.
Just as a data source for those curious: AEI's housing center sucks up a big pile of data and has it in some (potentially) useful formats -
https://www.aei.org/housing/housing-market-indicators/
Focused on housing sales and related items, while also making clear that housing markets vary wildly across geographic areas even though there is some coupling to broader economic trends.
The Federal Open Market Committee—the group that determines monetary policy—does forecast inflation among other economic predictions to guide their policy. That includes estimating future rent contributions to PCE using data such as market rents. For example, here’s the conclusions from a Feb 2022 letter from the SF Fed predicting higher inflation through sustained increases in the rental component due to higher market rents. [1]
> As the U.S. economy recovers from the effects of the COVID-19 pandemic, some increase in rent inflation should be expected, given that landlords can ask for higher rents when prospective tenants are employed and earning higher incomes. However, the extraordinarily large increases in two leading indicators of future rent inflation—asking rent inflation and house price inflation—point to significant upside risks to the overall inflation outlook. The potential increases are particularly significant for CPI inflation, which places a larger weight on shelter costs. Still, the potential additions to PCE inflation of about 0.5pp for both 2022 and 2023 are important to consider in light of the Federal Reserve’s 2% inflation target.
So the Fed is to some extent considering how current market rents will impact future inflation and using that consideration in developing monetary policy. It could be argued that the FOMC members are not putting sufficient weight on this specific consideration. Additionally, their forecasts may incorrectly estimate the future inflation contribution of rents and misguide policy. For example, in May 2022 the Richmond Fed revisited Aug 2021 predictions about future inflationary contributions of rents made by the Dallas Fed and found the predictions significantly underestimated inflation. [2]
> Growth in the PCE rent index exceeded 4 percent year over year in March, a threshold that the model predicted would not be crossed until the second quarter of 2023. Similarly, imputed rent for homeowners rose 4.4 percent in March, which the model predicted would have been attained in the first quarter of 2023. Year-over-year growth in both rent and imputed rent was over 1.5 percentage points higher than the upper confidence interval of the forecast in March.
It may simply be that during periods of high economic volatility it is hard to estimate the future trajectory of inflation. Any predictive model developed on historical data may simply fail to extrapolate to the specific novel economic forces that cause the volatility.
In retrospect it’s easy to say that the Fed should’ve better extrapolated how surging market rents in 2021 would lead to sustained inflation as leases rolled over. But had the sudden surge been followed by a comparable collapse in rents then such predictions would also have looked foolish. The core problem may be the inability to predict the evolution of our economy during periods of high volatility.
[1] “Will Rising Rents Push Up Future Inflation?”, https://www.frbsf.org/economic-research/publications/economic-letter/2022/february/will-rising-rents-push-up-future-inflation/
[2] “Revisiting Forecasts for Rent Inflation”, https://www.richmondfed.org/research/national_economy/macro_minute/2022/mm_05_24_22
Proof of average-rent housing disinflation: Yesterday my landlord sent me the renewal/rent increase notification for 2022-23. It was only a hair over 4%. (This is about in line with previous increases since I moved in in 2017.)
The question dominating US monetary policy is who will win the 2024 presidential election. One of the leading candidates attempted a putsch last year and failed. The governors of the fed are highly educated elites from coastal bubbles. They care more about preserving our form of government than technical measures of inflation. They’ll try to force a quick correction in order to get a boom going by 2024.
From an Electrical Engineer's point of view, this seems like a pretty simple application of control theory.
Can we just replace the Fed with a PID controller?
One of the issues with using mortgage costs as part of the index is that the higher interest rates used to combat inflation lead to inflation in mortgage costs.
There’s an interesting deep-dive in the context of the Canadian economy here: https://jbconsulting.substack.com/p/on-shelter-futures-part-1
Adjusted for quality, movie ticket prices these days look even worse when compared to the 1970's
I feel badly for Jerome Powell, who seems like a perfectly pleasant guy who essentially did what ‘everyone’ thought was right but now likely will be remembered as one of the worst Fed Chairs in history.
Tough to see your legacy disintegrate in real time, I would think.
But wait, why don’t central banks look at home
prices? I what would be the Downside to measuring home prices and rent and then sorting them out according to the proportion of each in the economy?
Good points. But it ignores what the Fed COULD pay more attention to: the Treasury Inflation Protected Securities -- TIPS -- expectation rate. It DOES include expectations concerning food and fuel and it enlists the crowdsourcing abilities of (potentially) million of participants in the bond markets. TIPs signaled that people had ceased believing that eh Fed would achieve it 2% target at least by September and the Fed did not move until November. TIPS would be even more useful if Treasury would issue them in more tenors, 1,2,3,7 15 and 30 instead of just 5 and 10.
When I'm president you will be my first nominee for fed chair and then you can fix it.
"And by the same token, the insane surge in spot rents in the summer of 2021 was not a sudden and dramatic blow to Americans’ living standards because, again, very few people actually signed new leases during that moment. The aggregate ups and downs of housing costs are more gentle, as the BLS says."
The cratering of prices in 2020 decreased overall price appreciation by what looks to be 45% or so and then the 2021 run up in prices increases it by what looks like 150%. in BLS terms.
Given that we have the Apartment List percent numbers, I think you can calculate what percentage of the rental population signed new leases during this time, and it seems like it should be a significant number if the overall rate of price appreciation for all apartment rents is oscillating over a 200% spread.
Edit:
In fact: Isn't it as simple that a 2% runup in BLS price during a 10% increase in Apartment List price means that 20% of the rental population has signed to pay the newly increased price during the time period being measured?
I've never liked the concept of OER, beginning when I learned about it in college econ. It just seems too homo economicus. Plus it seems to involve a lot more guessing than measuring. And if the majority of people live in homes they own, shouldn't we assume that if everyone rented instead that rental prices would dramatically fall, and therefore OER is not representative to begin with?
The problem is potentially even worse. This paper finds that raising interest rates causes rents to temporarily go UP as raising the cost of borrowing pushes some people out of buying and into renting. The trend likely doesn’t reverse until you get true recessionary conditions.
https://www.federalreserve.gov/econres/ifdp/files/ifdp1248.pdf
“ in contrast to house prices, housing rents increase in response to contractionary monetary policy shocks. We also find that, after a contractionary monetary policy shock, rental vacancies and the homeownership rate decline. This combination of results suggests that monetary policy may affect housing tenure decisions (own versus rent).”
The article seems to assume that the Fed doesn't know what Matt knows. I have to believe that they do know. (Fingers crossed.)