March saw a continuation of a year-long trend in which YoY shelter inflation accelerated by almost exactly an additional 0.3% each month. March came in at 5%, 0.3% higher than February's 4.7%, and so on.
Shelter is the largest category in the CPI basket, and it is critical for a reason. It isn't coming from Ukraine. It mostly isn't coming from supply chain bottlenecks and shipping container shortages. It isn't discretionary. And most importantly, it's a deeply lagging indicator with a long way still to go to catch up to the current activity in the housing and rental markets.
My expectation is to see this trend continue. Base year effects (in the second derivative) might slow it down a bit, but by June we could be looking at almost 6% inflation in shelter alone.
And remember -- a linearly increasing growth rate is faster than exponential price growth.
Interest rate hikes may help, but the dynamics are such that they'll need to make things worse before they make things better.
Nice observation and chart. Also really unsettling, do you think that housing prices running away like this might instigate some policy shifts?
I can't decide if the U.S. is headed towards a feudalistic society or not. If shelter cost inflation is accelerating it gives me hope, in a sort of grim accelerationist way, that housing will have to be addressed soon. I want to believe that we will see a housing campaign to "promote the American dream" or something, but everyone I talk to seems to think it's not gonna happen and democracy isn't alive enough to fix the corporate rent seeking.
> do you think that housing prices running away like this might instigate some policy shifts?
Probably, yes. It was easy to see this coming, so I think policymakers should have shifted policy earlier, but it will be hard not to do so at this point, because now you don't need to look ahead anymore to see it.
The last time shelter inflation rose in this fashion was around 2010 to 2013 -- but that was shelter bouncing back from -0.6% to 2.3% after the GFC, rather than running ahead of the inflation target. The time before that was 2005-2007, when shelter inflation ramped up to 4.3% before the GFC. And the time before that was the '80s Volcker shock.
That's not just pattern-matching; that's a causal relationship. When interest rates gradually rise or fall, housing prices fall or rise in tandem such that shelter inflation -- rent and mortage payments -- is fairly constant at around 2%. But when interest rates step-change suddenly, shelter reacts violently. When that change is a rate cut, shelter initially falls (as mortgages get cheaper) and then climbs back (as housing prices adjust for the lower rates). When that is a rate hike, shelter initially spikes (as mortgages get expensive) and then falls (after the real estate market corrects).
But real estate market corrections are more broadly devastating than stock market corrections (although they often go in pairs). Housing is a highly-leveraged asset, is most peoples' largest (or only) investment, is used as collateral for other loans, and so on.
Shelter is a good approximator for overall long-term inflation expectations, labor prices, and wage inflation. So the Fed doesn't have much choice but to either address it or allow an inflationary wage-price spiral.
I agree that shelter(housing) prices increasing at 5% is not good, but when we're seeing inflation of 8.5% there's something else that's driving things up.
When you look at just the energy component of the CPI you'll notice that it began to outpace inflation in March of 2021 at 13.2% and has fluctuated between 25% and 32% since them. The previous chart on that page that breaks down the most recent month's inflation between food, energy, and all other items you see that food is in-line with inflation, energy is crazy, and other items are at 6.5%.
We have a lot of problems going on right now, but why the energy sector has been experiencing crazy inflation for the last year might be more of a factor that the other things grabbing the headlines right now.
Shelter isn't housing, it's rent. Housing is up much more -- 21.7% YoY (as of January 2022, on top of 9% the previous year [1]) -- but that isn't directly captured in the CPI. Shelter CPI goes approximately as a massive low-pass filter on the product of (housing prices x mortgage rates), and so rising interest rates can carry shelter up higher, even after housing prices plateau.
Not seeing how that graph suggests shelter is the biggest story. Shelter looks linear since Feb 2021, but only caught up to pre-pandemic levels around Nov/Dec 2021, and while linear since then falls FAR behind energy, food, basically everything else incl the aggregate. A few months of linear shelter prices doesn't register as a meaningfully different trend than the rest of the basket that's been booming this winter, just that shelter would be more time delayed and smoothed0out due to leases coming up.
The reason shelter is the biggest story is not because shelter inflation is leading the pack (of course it's not, so energy is grabbing all the attention!), but because shelter has so much inertia behind it, and couples so much more closely to interest rates than to supply chains and wars.
> It mostly isn't coming from supply chain bottlenecks and shipping container shortages.
Inventory is still bumping along record lows. Housing starts were up in the new year, but it will take time for new units to come online after being side lined with labor and material shortages. ~5% mortgage rates are also likely to begin to slow things down.
Absolutely, but that's part of the mechanism through which price rises happen, more than a causal explanation of it. It further suggests that shelter has a lot of momentum.
> Housing starts were... side lined with labor and material shortages.
That's a bigger part, but easily offset by historically low population growth in the US. Low pandemic interest rates were almost certainly a much bigger factor in driving the housing market than the cost of construction.
> ~5% mortgage rates are also likely to begin to slow things down.
Again, absolutely (although if inflation also levels out at 5%, that may be neutral). But, as I mentioned in my comment, there's a "right half-plane zero" in the transfer function between interest rates, housing prices, homeownership costs, and shelter. Rising mortgage rates may level off housing prices, but meanwhile drive up the cost of homeownership. Even as the housing market plateaus, those rising homeownership costs can carry shelter higher, both in the form of real rents and "owner's-equivalent" rents. (In Canada the relationship is more direct, with mortgage interest costs directly included in the shelter basket, but the relationship exists indirectly in the US CPI metric too).
That shelter CPI can trend higher as housing prices plateau is the price of reciprocity that we pay for shelter CPI appearing low after interest rate cuts, even as housing prices grew rapidly.
https://www.bls.gov/charts/consumer-price-index/consumer-pri...
March saw a continuation of a year-long trend in which YoY shelter inflation accelerated by almost exactly an additional 0.3% each month. March came in at 5%, 0.3% higher than February's 4.7%, and so on.
Shelter is the largest category in the CPI basket, and it is critical for a reason. It isn't coming from Ukraine. It mostly isn't coming from supply chain bottlenecks and shipping container shortages. It isn't discretionary. And most importantly, it's a deeply lagging indicator with a long way still to go to catch up to the current activity in the housing and rental markets.
My expectation is to see this trend continue. Base year effects (in the second derivative) might slow it down a bit, but by June we could be looking at almost 6% inflation in shelter alone.
And remember -- a linearly increasing growth rate is faster than exponential price growth.
Interest rate hikes may help, but the dynamics are such that they'll need to make things worse before they make things better.