by Calculated Risk on 12/14/2015 11:59:00 AM

A few excerpts from a piece by Michelle Meyer at Merrill Lynch: Home sweet home

We expect continued improvement in homebuilding and sales in 2016 and 2017, but still far from a V-shaped trajectory. Here are our baseline forecasts:

• Housing starts to average 1.275 million in 2016 and 1.4 million in 2017 on the way to a return to the historical average of 1.5 million by the end of 2017.

[CR NOTE: For reference, housing starts will probably be just over 1.1 million in 2015].

• Existing home sales to increase 5% in 2016 and 3% in 2017. We look for more robust growth in new home sales with a gain of 10% and about 14% over the next two years, respectively.

[CR Note: New Home sales will probably be just over 500 thousand in 2015]

• Home price appreciation should slow with prices up only 1.8% in 2016. The forecast for 2017 becomes more controversial as our baseline forecast is for a decline of 1.5%, as our model looks for home prices to converge to income.

[CR Note: A decline in nominal prices seems unlikely in 2017. However a decline in real prices (nominal price increases less than inflation) is possible].

When considering the forecasts for the next two years, we have to ask two critical questions: what is a “normal” rate of activity, and, after years of below-normal activity, could the sector overshoot?

In our view, a reasonable estimate of “normal” is the pre-crisis average of about 1.5 million. The math is simple: household formation of 1.2 million + demolitions of 300K + some number of second home purchases. There is a risk that household formation is on a slightly slower pace given persistently high rates of doubling up among young adults.

Starting in the early 2000s, and accelerating post crisis, there is a rising share of “grown-up” children who live with their parents. While this is partly due to the Great Recession and slow recovery, there could also be some secular changes related to delayed marriage. This could imply a somewhat lower equilibrium pace of housing starts.

That said, in the medium term, the risk is that we overshoot this new-normal level given pent-up household formation. Although we are skeptical about the quality of the data, the recent statistics from the Housing Vacancy Survey show a surge in household formation, implying formation of about 1.5 million this year. While the data will probably be revised lower, there is an improving trend. Provided the overall economy continues to heal as we are expecting, with 2.5% growth in 2016 and an unemployment rate of 4.5% by year-end, we will see further growth in households.

Indeed, there is a risk that housing demand continues to overshoot supply, as we believe it has this year. Homebuilders have been complaining that a shortage of labor has been a major impediment to production. According to the NAHB, 52% of homebuilders report labor shortages, back to 2000 levels and up notably from 21% just three years ago. The challenge is that many skilled workers left the construction industry after the housing bubble burst and have not come back This could continue to extend delivery times and slow production.