Wonder why the Fed and the banks are so desperate to reflate the second housing bubble, to the delight of flippers and taxpayer consequences (deja vu) be damned? Simple: as Goldman points out in a note released last night, "without the boost from housing, real GDP growth would fall below 1% this year." That's the revised GDP by the way, the one that now includes iTunes song sales and underfunded pension plans in the sumtotal. Which in reality means that ex housing, GDP would almost certainly be negative. So the bigger question is what happens to housing which has already seen a shock to the system following the surge in interest rates in the past month and which hobbled both homebuilders and mortgage applications? This is what Goldman sees there: "On house prices, we have started to see the first signs of deceleration and expect a slowdown from the 10%+ pace observed over the past year. Our bottom-up house price model projects 4-5% annual growth rate in the next two years." Alas, since prices moves from top and bottom inflection point never happen in a straight line as everyone rushes to buy, or sell as the case may be, resulting in a skewed and pronounced move, once the reality seeps in that the artificial housing 'recovery' is over, watch what happens when everyone rushes for the door. That goes for GDP as well.
More from Goldman on housing's contribution to GDP:
How much will housing contribute to GDP growth?
Housing contributes to GDP growth in three ways: (1) the direct effect through residential investment; (2) the consumption impact through housing wealth and active mortgage equity withdrawal (MEW); and (3) the multiplier effect through increases in housing-related employment and easing of bank lending standards in a stronger home price environment. Using the same framework outlined in prior research, we project the contribution of housing to real GDP growth through late 2014.
As described before, we expect housing starts, the most important component of residential investment, to continue to recover despite the near-term challenges. Under this forecast, residential investment is projected to grow at an annual rate of 13-14% in 2013 and 2014, adding 0.4 percentage points per year to real GDP growth. This direct impact of residential investment continues to be the largest component of the housing contribution to GDP growth.
Our own estimate suggests that the housing wealth effect generates $0.04 of consumption growth for every $1 increase in housing wealth, which is broadly in line with the findings of the academic literature. Despite substantial uncertainty associated with the magnitude and timing of the effect, we believe that $0.04 spread over one year is a reasonable estimate at present. Combining this estimate with our view of house price deceleration, we project the housing wealth effect to contribute about 0.2 percentage points to GDP growth in 2013 and 2014, with the effect peaking at the end of this year.
The MEW impact, reflecting the effect of cash-out refinancing and home equity borrowing on consumption, continues to be small. We estimate that it adds less than a tenth of a point to GDP growth annually over the next six quarters. This estimate reflects the ongoing household deleveraging and tight mortgage lending standards. For example, according to data published by Freddie Mac, while nearly 90% of mortgage refinances in 2006 involved cash-outs, less than 15% of refinances do so today. Our own calculation using loan-level data shows that the average credit score on cash-out refinances increased after the housing crisis by even more than the average credit score on mortgage taken out to purchase homes. This suggests that MEW has been playing a minor role in the economy over the past few years.
The spillover effect via the labor market and the financial sector is more difficult to quantify but likely to be small as well. For example, we construct a measure of housing-related employment by adding payroll employment in housing relevant sectors (e.g. construction workers, furniture manufacturing and sales, mortgage brokers, and building services). During the year ending in June 2013, housing-related employment rose 2.9% and employment in other sectors rose 1.6%, reflecting the outperformance of housing relative to the overall economy. The higher growth rate of housing-related employment translates into 0.1 percentage point employment growth, or five basis points of income growth and less than three basis points of GDP growth. Although we think housing-related employment will continue to outperform other sectors and bank balance sheets will continue to improve as fewer mortgages default, the total multiplier effect on GDP growth is likely to be well below a tenth of a percentage point going forward.
Summing up, we expect housing to contribute 0.75 percentage points to real GDP growth annually over the next six quarters. This is significant, especially considering that real GDP is projected to grow at a sluggish rate of 1.5% in 2013. Without the boost from housing, real GDP growth would fall below 1% this year. That said, the 0.75 percentage point boost is still well below the 2-3 percentage point impact we saw on the downside during the Great Recession.
And why the administration, and those still long housing, should be worried: next stop lower prices.
The acceleration in house price growth during the first half of 2013 is quite striking. According to the CoreLogic National House Price Index, house prices increased 12% year-over-year in June 2013. The S&P Case-Shiller 20-city home price indices show that some of the hardest hit cities such as Phoenix, Las Vegas, and Atlanta registered growth rates of over 20%. In our view, some of the recent strong rebound in house prices is probably due to temporary factors, including demand from investors in the most depressed parts of the country. In fact, house price appreciation in Phoenix, a city particularly affected by the housing bubble where house prices bottomed five months before the nation as a whole, has started decelerating in recent months (Exhibit 2).
More broadly, there are three reasons why we think house prices should slow. First, national house prices are no longer cheap. Back in late 2011, house prices were below the fair value implied by fundamentals such as income and population, especially in areas that experienced the most severe housing downturn. With the growth over the past year, our model suggests that national house prices are now at fair value. Although the positive momentum in the housing market points to continued appreciation, the pace of price increases should decelerate in our view, or house prices will become notably overvalued soon.
Second, higher mortgage interest rates and increasing supply on the market are likely to restrain house price growth. After the recent rates sell-off, 30-year mortgage interest rates are 100bp higher than the lows seen in late 2012. Although we do not expect the 4.4% mortgage rate to end the housing recovery, higher interest rates do imply a reduction in housing demand, all else equal. In addition, housing inventory listed for sale was extremely low in 2012 due to a slow economic recovery and the large number of "underwater" borrowers who owe more on their mortgages than the house is worth. As the labor market improves and house prices rise, the inventory on the market has begun to increase.
Lastly, the double-digit house price appreciation over the past year was partially driven by a declining share of distressed sales and investor activity. Since distressed properties are typically sold at discount, a smaller share of distressed transactions means higher measured house price appreciation, even when we hold everything else constant. As the fraction of distressed sales normalizes and REO-to-rental yields compress on the back of higher house prices, growth in house prices is likely to decelerate. Our bottom-up house price model currently forecasts national house prices to increase at a pace of 4-5% per year over the next two years.
So... is it untaper time yet?