Volume 14 Issue 3
This is an excerpt from the Fall 2014 issue of The Linneman Letter.
Cap Rates Have A Downward Bias
Looking Forward
NCREIF Total Quarterly Returns
8
6
4
2
Percent
As we mentioned in the last issue, we have recently been exploring
the statistical relationship between cap rates and changes in real
GDP, the unemployment rate, inflation, 10-year Treasury yields, and
the flow of funds into mortgages. In order to get a feel for the
magnitude of the impact of the mortgage flow of funds on cap rates
versus other factors, we examined historical correlations between
NCREIF cap rates and various macroeconomic factors. Based on
those correlations, we simulate the impact of the simultaneous
occurrence of a 5% increase in real GDP, a 200-bp rise in the
10-year Treasury yield, a 100-bp decline in the unemployment rate,
stable inflation, and a relatively modest (by historical standards) 25%
increase in real mortgage lending. The net simulation result is a
104-bp decline in cap rates, with apartments (which already have the
greatest flow of funds) declining the least and retail declining the
most. While not definitive, our empirical results are highly suggestive
that the cyclical flow of mortgage funds, which is just beginning,
will force cap rates downward – or at least hold them flat – even as
interest rates rise substantially.
By Dr. Peter Linneman, PhD
Chief Economist, NAI Global
Principal, Linneman Associates
0
-2
-4
-6
-8
-10
1978
1982
1986
1990
1994
1998
2002
2006
Table of Contents
The Great Unknown: Positive Net Jobs
Will The U.S. Economy Ever Get Back On Track?
The Fourth Branch of Government
Results of The Linneman Letter Survey
What’s Keeping You Up At Night?
Cap Rates Have A Downward Bias Looking Forward
Our analysis (which in view of data quality is more suggestive than
definitive) reveals that changes in real GDP and unemployment rates
impact cap rates quite modestly, results which we are convinced are
as expected. This is because once economic growth pushes rents to
levels that justify replacement, further economic growth simply
creates new supply rather than further boosting real cash flow
growth. Similarly, consistent with theory, we find little empirical impact
of inflation on cap rates, as higher inflation raises both alternative
asset returns and perpetuity cash flow growth by roughly equal
amounts, hence canceling out any significant impact on cap rates.
In terms of the impact of 10-year Treasury yields on cap rates, our
research also finds little notable impact. In fact, for both REIT implied
cap rates and NCREIF cap rates over the past 12 years, a decrease
in 10-year Treasury yields is generally associated with rising cap
Real Estate Capital Markets
Facts About Foreign-Born Immigrants
Education Matters
Housing Market Update
A New Modi For India
Market Close-up: Phoenix Office
Market Close-up: Seattle Industrial
Market Close-up: Detroit Multifamily
Market Close-up: Orlando Hotel
Office Market Outlook
Industrial Market Outlook
Multifamily Market Outlook
Retail Market Outlook
Hotel Market Outlook
Seniors Housing And Care Market Outlook
Pipeline Sensitivity Analysis
Vacancy/Occupancy And Absorption Projections
For more information about a subscription to
The Linneman Letter, contact Doug Linneman
at dlinneman@linnemanassociates.com.
2010
2014
Volume 14 Issue 3
Percent
9
8
7
Easy
5
1984
Tight
1989
Tight
6
Easy
1994
1999
2004
Easy
2009
2014
Source: NCREIF
NCREIF Transaction Cap Rates Over
10-Year Treasury Rates
600
400
200
0
-200
-400
-600
1984
Tight
Easy
1989
Tight
Most observers say, “But a higher Treasury yield means that real
estate must provide a higher return in order to be competitive, so
surely cap rates must rise with interest rates.” But the data does not
support this seemingly logical position. This is partially due to periods
of flight-to-safety (and subsequent reversal), creating periods of
negative correlation between cap rates and Treasury yields. But even
more so, the lack of a clear relationship between Treasury yields and
cap rates reflects the fact that the flow of mortgage funds swamps all
else in terms of impact on cap rates. That is, as mortgage funds
flow – even in the face of rising interest rates – mortgage spreads
narrow, loan-to-value ratios (LTVs) rise, loan covenants disappear,
and payment-in-kind (PIK) lending allows borrowing in excess of cash
NCREIF Transaction Cap Rates
10
Basis Points
rates. This is because large declines in Treasury yields are associated
with investor “flight-to-safety,” while subsequent increases in
Treasury yields reflect a return to normalcy. And over longer data
periods, NCREIF cap rate data reveal little correlation with 10-year
Treasury yields.
Easy
1994
1999
2004
2009
Easy
2014
Source: NCREIF
NCREIF Cap Rates (Basis Points), 1982Q4 - 2013Q4
NCREIF Cap Rates
Transactions
Apartment
Office
Industral
Retail
10-year Treasury Yeild
+200 bp
Linneman Index*
+25%
Real GDP
+5%
Unemployment Rate
-100 bp
Total
NS
NS
36
NS
NS
(70)
(69)
(133)
(102)
(125)
7
(17)
19
NS
18
(41)
NS
(59)
(25)
(37)
(104)
(85)
(137)
(127)
(145)
Unemployment Rate
-100 bp
NS
NS
NS
NS
(25)
(16)
NS
23
103
31
Total
183
248
116
195
134
244
298
(37)
441
(369)
* Linneman Real Estate Index measures multifamily and commerical mortgage as percent of GDP.
NS: Not satistically Significant
Source: Linneman Associates
Implied Cap Rates of REITs
Total REITs
Apartment
Office
Industral
Shopping Center
Regional Mall
Self Storage
Manu. Housing
Triple Net
Health Care
Impact on REIT-implied Cap Rates (Basis Points), 2002Q4 - 2013Q4
10-year Treasury Yeild
Linneman Index*
Real GDP
+200 bp
+25%
+5%
(76)
258
NS
(87)
335
NS
(112)
228
NS
(152)
347
NS
(88)
247
NS
NS
259
NS
NS
298
NS
NS
NS
(59)
NS
455
(117)
NS
(242)
(158)
* Linneman Real Estate Index is calculated based on multifamily and commerical mortgage as percent of GDP.
NS: Not satistically Significant
Source: Bank of America, Linneman Associates
Volume 14 Issue 3
coverage. In fact, as LTVs rise due to large mortgage flows, real
estate equity becomes more akin to the purchase of an option rather
than ownership per se. The result of massive mortgage lending flows
is to reduce the weighted cost of capital for real estate for highly
leveraged buyers, even at higher Treasury rates. This tends to drive
low leveraged buyers from the market, in turn lowering cap
rates. Hence, the statistical finding is that cap rates are primarily
driven by the flow of mortgage funds, rather than the state of the
economy or interest rates.
Commercial/Multifamily Mortgage Debt Outstanding
Total Outstanding Most Recent
43% (real)
Percent Change in Total Loans (1996-2001)
43% (real)
Percent Change in Total Loans 2003-2008
60% (real)
Potential & Actual Bank Loans
30
25
$ Trillions
It is important to realize that while there is plenty of debt currently
available for the best properties in gateway markets (New York, Los
Angeles, San Francisco, Boston, and Washington, D.C.), the overall
flow of mortgage funds is still at a cyclical low. For example, total
outstanding mortgage debt for commercial real estate is 5.7% below
its 2008 peak in nominal terms and nearly 15% in real terms. Further,
outstanding mortgage debt has only risen by 4% since its trough in
2012. And the Linneman Index, which measures outstanding
mortgage debt relative to GDP, is 20% below its peak in 2009, and
appears to have finally stabilized. Despite increased lending, GDP
has grown faster than commercial mortgage flows to date, resulting
in continued declines in the LREI. However, the current LREI appears
to have bottomed as has been essentially flat for the last six quarters.
In short, U.S. real estate is near the bottom of a mortgage flow of
funds cycle; but as lending ramps up, the LREI will eventually
increase. And as this cycle swings upward, the pressure will be
to notably lower cap rates, even if (as we anticipate) 10-year
Treasury rates rise.
}
20
15
10
Unused
Lending
Capacity
5
0
1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014
Potential Bank Loans
Actual Bank Loans
Actual Bank Loans As Percent of Potential Bank Loans
140
120
100
80
60
40
20
0
1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014
The last three mortgage flow of funds cycles saw real increases in
outstanding mortgages from their respective troughs of 43%
(1985-1990), 43% (1996-2001), and a staggering 60% (2003-2008).
And this cycle could dwarf previous cycles in light of the staggering
amount of unused lending capacity – nearly $20 trillion based on
$2.7 trillion currently in excess reserves – that the Fed has created at
money center banks.
Historical Bank Excess Reserves
3,000
2,500
$ Billions
Banks, the dominant real estate lender, have historically pushed real
estate lending aggressively with a lag to corporate lending. We are
already seeing extremely loose bank credit for corporate buyouts and
$3.5 trillion
Percent Change in Total Loans 1985-1990
2,000
1,500
Avg. 1959 to August 2008 = $816 million
Avg. Sept 2008 to Present = $1.4 trillion
1,000
500
0
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Volume 14 Issue 3
Money Multiplier*
14
12
10
8
6
4
2
0
1964
1959
1969
1974
1979
1984
1989
1994
1999
2004
2009
2014
*M2 / Monetary Base
Excess Reserves as a Percent of Total Reserves
120
Percent
100
80
1959 to August 2008 Average = 2%
60
May 2014 = 95%
40
20
0
1959
2,500
1964
1969
1974
1979
1984
1989
1994
1999
2004
2009
2014
Excess Reserves as a Percent of Required Reserves
2,000
Percent
corporate refinancing, as banks attempt to put nearly $20 trillion of
unused lending capacity to work. Never before has the U.S. banking
system possessed so much unused lending capacity. Current
real estate bank loans are $3.5 trillion (27% of all mortgage debt
outstanding) versus estimated unused bank lending capacity of
$20 trillion. Thus, it is hard to conceive that the forthcoming cycle in
the flow of mortgage funds will not exceed the 60% real increase
from 2003-2008. And remember that cap rates fell during that period
in spite of high (and rising) Treasury yields.
1959 to August 2008 Average = 2.1%
1,500
August 2014 = 1,900%
1,000
500
0
1959
1964
1969
1974
1979
1984
1989
1994
1999
2004
2009
2014