In a recent guest contribution to HousingWire, Romi Mahajan suggested that the  National Association of Home Builders Housing Market Index is “inefficient, unscientific” and not predictive of future sales – In other words, useless or even “damaging” to the economy. 

I disagree.

The NAHB HMI is a builder confidence survey that provides a gauge of builder’s opinions on whether they feel confident enough to build or not to build single-family homes. The April HMI was 30, the lowest since June 2012.

If you follow the index you know that number below 50 is considered “negative” or whereas a number above 50 indicates a favorable outlook on the future of housing starts. The index rose slightly in May to 37, which was better, but still considered negative.

As part of my work, I track all kinds of confidence indices, and while I understand they are not the equivalent of a “real economic model” in terms of predictive ability, many of them are useful and can give clues early on with shifting positive data. 

From the chart below, we can see that the HMI had a massive collapse during the housing bubble crash and a slow but steady recovery up until 2018, after which there was a noticeable drop.

In 2018, mortgage rates rose to 5%, and the builders had a terrible fourth quarter. The monthly supply of housing spiked to 1994 levels because higher interest rates scared off buyers and they simply weren’t buying as many new homes – especially on the larger, more expensive new homes.  This wasn’t a good time for the builders.

More recently, even with the coronavirus-induced recession on-going, monthly supply never got back to 2018 levels. The recession definitely threw a wrench into the overheating gears of the 2020 housing market, but it didn’t kill it.

Housing starts, for example, are still a positive, year to date, and in the last new home sales report, sales beat estimates by one of the largest margins ever recorded in U.S. history. 

This recession has created negative waterfall economic charts and parabolic jobless claims charts, and the U.S. housing market also took a hit. But the HMI index did show a bounce recently, and this bounce was before the better new home sales data report. A similar heads up was given after rates move up in 2018. Economist Robert Dietz from the NAHB shows a good correlation here on the builder’s confidence moving along with housing starts.

Typically housing starts get weaker going into a recession. Historically, this is a noticeable trend. However, going into the COVID-19 recession, housing starts grew almost 40% year over year in the month of February in 2020, creating one of the best housing month data lines we had in over a decade.

Of course, this was prior to the COVID-19 shutdowns. 

Then on March 12, the government started its process of shutting down the economy and for the short term, there goes the strongest start to housing in over a decade. COVID-19 is creating special circumstances for all kinds of sectors in the economy.

In this case, the reason why a sector is responding as it does matter.

This is very different than your standard, bog-basic recession. Why something is happening matters because not all sectors of the economy are affected the same or react the same to the strange circumstances COVID-19 has forced us to endure. Engaging with the data, even soft data like surveys of opinion gives us reliable hints into what to expect in specific segments of the economy and what parts are going to recover faster.

The HMI tells us that the housing sector is recovering, slowly but surely. That increase we saw in the HMI data also goes with the recent pickup in purchase application data which finally went positive year over a year recently after being down as much as 35% year over year.

Yes, the reasons always matter.