The market has caught up with our misgivings about inflation in recent months (and then some), and recent inflation readings have consistently surprised to the upside. In the year to June the CPI increased by 5.4%, with the ‘core’ measure (excluding food and energy—because who needs those?) rising 4.5%. These were startling numbers, but policymakers and the mainstream financial press continue to repeat the mantra that this is all ‘transitory’, as they try to keep the general public from paying too close attention. They put these moves down to low base effects from last year. That appears to ignore the far more concerning fact that the core CPI measure increased 0.9% versus the prior month – yes, that’s an annualized rate of over 10% for those not used to seeing the math work out that high. Again, policymakers are claiming this is the result of transitory factors due to the pandemic, with massive jumps in used car and truck prices in particular driving much of that move.
That is true, but what is also true is that far more non-transitory drivers of inflation are now flashing red.
The largest component of the CPI is shelter, at nearly one-third of the total inflation measure. Shelter is comprised of both rent and owner’s equivalent rent (‘OER’), the latter being a measure of what it would cost a homeowner to rent an equivalent dwelling. OER is estimated from observed market rents on equivalent types of property.
The shelter component of CPI tends to move slowly as the data captures rents actually paid, not asking rents on vacant properties, so it takes a while for increases to show up in the inflation data given leases extend for one to two years. But changes that do occur tend to be sticky.
So far shelter hasn’t played a significant role in the high headline inflation readings. In the June CPI reading rent and OER increased only 0.2% m/m and 0.3% m/m respectively. There is substantial evidence, however, that rents are rising quite fast. According to Apartment List, the median national rent in the U.S. increased 9.2% in the first half of 2021. Much of that was a recovery from the pandemic, but rents are now back above pre-COVID levels and moving higher. Surveys indicate consumers are bracing for high single digit rent increases over the coming twelve months.
The bigger of the two components that make up shelter is the OER, which is nearly a quarter of the total CPI by itself. Unlike some other countries the purchase price or construction cost of a property does not drive OER so the fear that rising lumber and other building material costs will impact the CPI through OER are misguided (unlike Australia where they will, once the government’s A$25k renovation handouts to homeowners disappear, which are applied as an offset against rising prices in the calculation of CPI… yes, seriously).
But the fact that house prices are rising does influence OER eventually, and house prices are absolutely on fire. The S&P Case-Shiller national home price index rose 14.6% y/y in April, one of the fastest increases on record. In late June the National Association of Realtors said that the median existing-home sale price in May reached more than $350,000 for the first time, and was up nearly 24% from a year ago. That was the largest annual increase in the data going back to 1999.
The government-sponsored Federal National Mortgage Association (‘Fannie Mae’) estimates that house price gains historically lead changes in the CPI shelter cost measure by around 5 quarters[1]. The below chart illustrates this lagged effect quite well.
That means the rapid gains in house prices will only really start to impact shelter inflation in the 3rd and 4th quarters of next year.
Fannie Mae’s research department put together a simple three factor linear regression model that seeks to explain changes in OER based on 1) annual house price growth, 2) the single family rental vacancy rate and 3) 1-year inflation expectations (as measured by the Cleveland Fed). The model predicted 85% of the variation in OER growth going back to 1984.
This is what it shows.
By mid-2022 the shelter component of the CPI will be running at 4.5% according to Fannie Mae’s model, which would contribute between 1.5 to 1.9 percentage points to the CPI and core CPI respectively. If housing prices continue to move higher in the second half at the current rate, according to the report, “housing could contribute more than 2 percentage points to core CPI inflation by the end of 2022”.
The shelter contribution for the Fed’s preferred inflation measure, the core Personal Consumption Expenditures (‘PCE’) index, is somewhat lower at around 16%, so the impact on the Fed’s preferred measure is not as pronounced. Nevertheless this would drive around 1 percentage point of core PCE inflation.
The fact that house prices and rents are both rising points to an underlying imbalance between supply and demand for housing. There are some estimates that the US faces a shortfall of approximately 5.5 million houses due to the lack of new home construction after the financial crisis of 2008. This shortfall is occurring just as the Millennial generation, the largest demographic in U.S. history, enters peak family formation years.
Recall the survey of consumer inflation expectations from the New York Fed highlighted in previous letters. Earlier it was the level of uncertainty of future inflation that piqued our interest, but now inflation expectations are clearly on the rise. These surveys tend to run hotter than the reported inflation, but June’s reading of 4.8% for the one-year ahead and 3.5% for the three-years ahead are the highest in the history of the series back to 2013.
Higher inflation can be good for stocks up to a point. The academic studies show, however, that higher inflation is generally associated with poor equity returns. Then again, the alternatives to stocks are generally even worse if, or when, inflation finally does take hold.
[1] Eric Brescia, Economic & Strategic Research (ESR) Group, Fannie Mae, Housing Insights: Housing Poised to Become Strong Driver of Inflation, 9 June 2021