Housing is Back, UCLA Anderson Forecast, September 2015

After a long, hard slog, housing starts (both single and
multi-family) are poised to approach the long-term
average (1959-2014) of just under 1.5 million units in
2016. (See Figure 1) Specifically we are forecasting housing
starts of 1.14 million units this year and 1.42 million units
and 1.44 million units in 2016 and 2017, respectively. This
level of activity is well above 1.00 million units recorded in
2014 and the 2009 low of 0.55 million units. Remember that
the level of activity we forecast is far from the mid-2000s
boom level of above two million units a year. We would also
note that with the shift to multi-family starts, the per-unit
GDP “bang for the buck” has declined, but that factor has
been partially offset by increased emphasis on higher-end
housing in the new construction market.

Our forecast is underpinned by continued growth in
real GDP that will likely run at a 3% rate in 2016, continued
jobs gains in excess of 200,000 a month for most of the forecast
period, relatively low mortgage rates--at least through
2016 and household formations in excess of one million a
year in 2016 and 2017. (See Figures 2, 3, 4 and 5) To dig into
the weeds, our estimates for household formation is derived
from the Current Population Survey which when compared
to the Housing Vacancy Survey seem conservative. Further,
the improving labor market will act as an ongoing stimulus
to household formations.

Although low mortgage rates have been with us for
years, what is important is that credit standards have eased
with respect to FICO scores and down payment requirements
have been reduced. To be sure we are not going back to the
“wild west” lending standards of 2005, but compared to
2010, and yes early 2014, mortgage credit conditions have
decidedly eased. Moreover, we do not believe that higher
mortgage rates will meaningfully cut into housing activity
until 2017 as a rise in rates will initially hasten buyers into
the market out of fear that rates will go much higher. Time
will tell whether or not this assumption is too heroic


Figure 4 30-Year Conventional Mortgage Rate,
2005Q1 – 2017Q4

Figure 5 Household Formations, 2010-2017F

Figure 2 Real GDP Growth, 2005Q1 – 2017Q4F
Sources: U.S. Department of Commerce and UCLA Anderson Forecast

Figure 6 S&P/Case-Shiller Home Price Index, 20 City Composite, 2000 - June 2015, 2000=100

Figure 7 Existing Home Sales, 2000 - 2017F

The rebound in housing construction is being confirmed
by rising home prices with the widely reported Case-
Shiller Index up 5% year-over-year and up 30% since the
low in 2012. (See Figure 6) Similarly, existing home sales
are forecast to be 5.3 million units this year up from the 4.1
million unit low in 2008. (See Figure 7) We forecast that
existing home sales will reach 5.5 million units in 2016 and
modestly decline to 5.3 million units in 2017.

Interestingly, the housing recovery is occurring
under the backdrop of an unprecedented decline in
home ownership. Specifically, the home ownership rate
has declined from 69% in 2005 to the current 63.5%,
which is roughly where it was in 1989. (See Figure 8)
The decline in the home ownership rate is attributable to the
after effects of the housing crash of 2006-2010 which scared
off would be homeowners, tighter mortgage requirements,

Figure 9 Student Loan Debt, 2006Q1-2015Q2F, $Billions

Figure 8 Homeownership Rate, 1965 - 2015Q2, NSA

sluggish income growth, a shift in consumer preferences to
urban versus suburban lifestyles, and the rapid growth in
student loans which now exceed $1.2 trillion (See Figure
9) In fact, the biggest drop in homeownership has taken
place in 25-34 year old cohort where the rate dropped 5 full
percentage points from 1993 -2014.1 We believe that this
declining trend has about run its course and will soon begin
reversing. In support of this notion we note that the recent
decline in life events associated with home ownership such
as marriage and childbirth have ebbed and are now in the
process of reversal.

The Boom in Multi-Family and Rentals

The flip-side of the decline in the homeownership
rate is a rise in renting which has triggered a boom in multifamily
housing starts (See Figure 10). Multi-family housing
starts which bottomed in 2009 at 112,000 units will exceed
400,000 units this year and average 460,000 units over the
next two years. The boom is underpinned by rents increasing
at a 3.5% a year rate in the official data, but according to the
publicly traded apartment real estate investment trusts, rents
are increasing on the order of 4.5-5.0%. (See Figure 11) As
we have noted before, the official data tends to lag the actual
market place because of the prevalence of rent controlled
jurisdictions in the official sample. Simply put, rents in con-

Figure 11 Consumer Price Index, Rent of Primary Residence, January 2000 - July 2015, Percent Change Year-Over-Year

Figure 10 Multi-Family Housing Starts, 2000Q1 – 2017Q4F

trolled jurisdictions aren’t typically marked to market until
a vacancy occurs. The primary reason that rental increases
have been sustainable is a very low 4% national (based on
79 cities) apartment vacancy rate, roughly half of what it
was a few years ago. (See Figure 12)

Moreover, this cycle has given rise to nationally oriented
single-family rental businesses funded by institutional
investors and public offerings of shares. This business is the
creature of the huge amount of bank foreclosed property that
came on the market in the aftermath of the financial crisis
enabling the bulk buying of single-family homes. Thus far,
single-family rentals have captured an unprecedented half
of the total rental market over the past few years and the
public companies have been reporting rental growth on the
order of 4% a year. 

In fact we are now witnessing the purchase
of new single-family homes for the rental market by
investment institutions and the development of homes for
rent by traditional home-builders. This consumer preference
for single-family rentals is one of the reasons we believe
that the American dream of at least living in a single-family
home is far from dead and ultimately many of those rental
units will turn into owner occupied housing.

The trends outlined above have not gone unnoticed
by the investment community as torrents of cash has flowed
into the sector driving up apartment values and spurring new
construction. In a yield constrained world, the cash flows
associated with apartment ownership have looked increasingly
attractive to institutional and retail investors alike
and that has driven initial yields down to below 5% and to
below 4% in the more favored markets. Just to note, initial
yields on apartment projects were close to 8% at the height
of the financial crisis.
Figure 12 Apartment Vacancy Rate, 1980 - March 2015

However, because we expect interest rates to rise
over the next few years, the decline in homeownership rate
to level off and high new construction levels to negatively
impact vacancy rates, the apartment boom is likely to show
real signs of strain by late next year.

More importantly, with rents rising faster than incomes,
affordability will soon become a binding constraint
on rents. For example, from 2004-2014, the percentage of
households paying more than 30% of their income rent
increased from 40% to 46%.2 With developers building for
the top of the market, meaning high income renters, they
may not yet to be cognizant of this trend, but they will soon
find out that the high-end apartment market might not be as
deep as they think.

Conclusion
Yes, housing is back. It will not be a rerun of the 2005
boom, but starts will soon approach 1.5 million units a year.
The multi-family apartment boom will continue throughout
2016 as developers race to keep up with demand for urban
infill housing. Nevertheless, housing activity will begin to
gradually fade in 2017 as mortgage rates rise and apartment
vacancies increase.