A financial bubble is generally defined as trading in an asset
market at a price or price range that significantly deviates
from the corresponding asset market’s intrinsic value.
It could also be described as a situation in which asset
prices appear to be based on improbable, aspirational, or
inconsistent views about the future. Bubbles are insidious things.
They are incredibly difficult to spot in advance because it is often
difficult to determine intrinsic value on a real-time basis. Therefore,
it is almost always only after the fact of a bubble bursting that it
becomes apparent that one even existed. Even as the commercial
real estate (CRE) market stared into the face of one in 2008, there
were many who ardently denied the bubble’s existence.

The only good thing about bubbles (if there is any such good thing)
is that they provide useful guides to determine if a market has
entered another bubble. Therefore, by using the most recent CRE
bubble as a benchmark, the current market environment can be
analyzed to determine if another bubble has formed, or is potentially
forming. When this analysis is performed, it is apparent that the
current situation is not representative of a bubble.

What Qualifies as a Bubble?

In CRE what qualifies as a bubble? What should be used to determine
if intrinsic value is misaligned with current values? Comparing a
pure valuation metric such as price per square foot or price per
unit can be misleading since values tend to rise over time, if for no
other reason than inflation. Therefore, ceteris paribus, one should
expect real estate values to trend upward over time. Empirically,
that is generally observed. However, we can compare cap rates,
which are defined as the ratio of net operating income (NOI) to
value, because both NOI and values tend to rise over time due
to the systematic impact of inflation on both, providing a better
assessment of value. Moreover, cap rates are frequently the key
pricing metric that investors use to determine the intrinsic value
of a property. As a general rule, the lower the cap rate, the greater
the perception that market values have risen above intrinsic values
because of the inverse relationship between value and yield.
Therefore, this examination will compare the current cap rate
environment to that of the bubble years from before the recession to
determine if the market is currently experiencing another valuation
bubble. Specifically, it will look at the constituent components
of cap rates to help determine if intrinsic values are aligned with
current valuations.

Current Cap Rates Versus Bubble Cap Rates

The first step in this process is to compare the current cap rate
environment with the previous bubble’s cap rates across property
types. For this purpose, we employ cap rates from the three major
property types — apartment, office, and retail. We will compare cap
rates from 2006, 2007, and 2008 (when the CRE bubble finally
burst after the implosion of Lehman Brothers) with cap rates from
the last few years — 2013, 2014, and 2015. Therefore 2006 will be
compared to 2013, 2007 to 2014, and 2008 to 2015. These years
were chosen not only because we want to compare the current
environment to that during which the bubble burst, but also because
the recovery/expansion periods in the economy are of similar length
and make for an apt comparison. After the dot.com recession of
2001 ended, there were 28 quarters until the CRE bubble burst.
In contrast, there are 26 quarters between the end of the Great
Recession and the current quarter, the fourth quarter of 2015. This
comparison is shown in Exhibit 1. Because the fourth quarter has
not yet ended, there is no growth rate for real GDP yet.

Bubble Ex. 1

The aim is not to imply that a bubble is imminently about to burst,
but to contrast the environment leading up to the bursting of the
bubble from the last decade with the environment leading up to
today. We utilize 12-month rolling cap rates because 2015 has not
yet ended so the full calendar-year cap rate is not available. For
the other time periods the 12-month rolling cap rate is the average
cap rate for each calendar year. The simple differences between
calendar year cap rates show how remarkably similar today’s cap
rates are to cap rates from before the Great Recession. These are
shown in Exhibit 2.

Cap Rate Differences

US Office US Apartment US Retail
2006 Minus 2013 0.0% -0.1% 0.6%
2007 Minus 2014 -0.3% 0.0% 0.2%
2008 Minus 2015 0.0% 0.4% 0.7%

These strikingly similar cap rates are why many in the market today
believe that another CRE bubble has formed. Only by analyzing the
forces behind these cap rates can we make the determination that
we are not in a bubble. In order to do this it is important to analyze
the differences between the constituent components of a cap rate,
the risk-free rate of return and the risk premium, from before the
recession versus the last few years.

The Risk-Free Rate of Return

As a general rule, the higher the risk-free rate of return, the higher
the cap rate should be, ceteris paribus. This is because any
investment needs to compensate the investor for the risk they are
taking (the risk premium) plus the risk-free rate of return that they
would otherwise receive from investing in a riskless asset. Due
to the relatively long-term holding periods for CRE, the ten-year
Treasury rate is often used as the risk-free rate of return. Once again,
it is instructive to look at differences, in this case the ten-year
Treasury rate from the years before the Great Recession versus
the last three calendar years. As you can see in Exhibit 3, there are
some significant differences between rates during these periods.
These equate to about 2% for each of the first two calendar-year
comparisons and not quite 1% for the last comparison. However,
this last difference is a bit misleading. The ten-year Treasury rate
fell dramatically during the flight to quality after the implosion of
Lehman Brothers. Prior to that, the difference was closer to the
2% that we see for the other years.

Ten-Year Treasury Differences

2006 Minus 2013 1.9%
2007 Minus 2014 2.0%
2008 Minus 2015 0.9%

Ceteris paribus, the data demonstrates that based on the interest
rate difference, cap rates over the last few years should be lower
than cap rates from the years leading up the bubble bursting by
roughly 200 basis points. This therefore makes the environment
from before the Great Recession look like a bubble — cap rates were
too low relative to interest rates. However, in today’s environment,
cap rates are as low as they were before the recession, but that
seems appropriate in the context of ten-year Treasury rates being
roughly 200 basis points lower today than they were during the
bubble period. The instinct here is to conclude that a bubble also
exists today because the Fed is keeping nominal interest rates
artificially low, depressing cap rates, increasing market values
relative to intrinsic values, and creating a bubble.

However, the Fed can only affect interest rates in a limited way and
on a short-term basis. Of course the Fed sets the target Fed Funds
rate, which can be viewed as the benchmark for all other interest
rates. And the Fed’s policies influence (though do not explicitly
control) the rate of inflation, which partially determines nominal
interest rates. But the more critical component is real interest
rates, and the ability of the Fed to influence real interest rates,
especially over the long term, is limited because real interest rates
are primarily determined by the real growth rate of the economy.
Taken together, these two factors perfectly explain the low interest
rate environment of today. Inflation has been declining for a number
of decades as part of a longer-term structural change. However,
since the end of the recession it has been struggling to even reach
the Fed’s 2% target rate. Meanwhile, the real annual growth of the
economy since it began to recover in mid-2009 has been around
2%, below the economy’s long-run average and below the growth
rates achieved last decade before the bubble burst. Therefore, low
interest rates are not due to aggressive monetary policy, but are a
product of today’s slow-growth, low-inflation environment.

The Risk Premium

The risk premium is the other determinant of cap rates. The risk
premium generally derives from two sources — the fundamental
space market and the capital market. The key factor for space
market fundamentals is growth expectations. They are an important
determinant of cap rates because, ceteris paribus, investors generally
pay more for a property with rising cash flows than for one with a
flat or declining cash flows. Therefore, the greater the expectation
of cash flow growth, the greater the downward pressure on cap
rates. By looking at the trends in cash flow growth over time
between the two comparison periods (2006-2009 vs. 2013-2015), it is obvious that the current environment differs greatly from the
pre-recession environment for each of the three major property
types. The annual cash flow growth rates by property type for the
aforementioned periods are shown in Exhibit 4. Note for 2015,
because the year has not ended, the current forecasted cash flow
growth rates are utilized. However, with three quarters of the year
already completed, the actual growth rates for 2015 are unlikely to
deviate significantly from the forecasted figures.

For the period 2006-2008, growth rates had been accelerating
over time and reached robust levels by 2006 and 2007. The one
possible exception is with retail where massive overbuilding limited
cash flow growth. Problems with valuations arose in early 2008
when the cash flow growth rates began to slow, even before the
CRE bubble burst. Nonetheless, investors continued using highly
improbable forecasted growth rates (anchoring their expectations
to recent positive experience in 2006 and 2007) that were not
going to be ultimately realized. By the end of 2008, the actual growth
rates for office and retail were negative while that of apartment
was slightly positive. All of these growth rates came in well below
expected growth rates that were being used in valuations at the
time. These overly optimistic assumptions about future cash flow
growth caused a compression in the risk premium as investors
believed the strength in the market would continue, just as the
market was about to implode. Consequently, this helped push cap
rates to levels that were too low — market values spiked and deviated
significantly from intrinsic values.

Exhibit 4
Cash Flow Growth Rates by Period

Bubble 4

However, for the 2013-2015 period, cash flow growth continues to
accelerate, has not yet peaked, and remains far below the growth
rates from the 2006-2007 period. Moreover, even if the forecasted
cash flow growth rates for 2016 and beyond increase, it is highly unlikely that they will reach the growth rates from the 2006-2007
period. Although the current positive trends in cash flow growth
and current cash flow forecasts are putting upward pressure on
valuations and downward pressure on cap rates, they are not doing
so to the same extent as before the recession. Essentially, this
is the inverse of what was observed with interest rates. During
2006-2008, interest rates were generally high which put upward
pressure on cap rates. However, the overly optimistic assumptions
about cash flow growth during that period lowered the risk premium
and put downward pressure on cap rates. During the current
period, relatively low interest rates are putting downward pressure
on cap rates. However, because cash flow growth is not nearly
as strong now as it was before the recession, investors are using
more modest (or at least less aggressive) assumptions about
forecasted cash flow growth and consequently the risk premium.
Therefore, downward pressure on cap rates from projected cash
flows exists, but not nearly as much as it did before the recession.
This is causing market values today to be far closer to intrinsic
values than they were around the time the bubble burst.

Capital markets also have a profound impact on the risk premium,
but directly measuring this impact can be a challenge. For example,
the interest rate on commercial mortgage debt can be a poor
indicator because of the risk-free rate component — commercial
mortgage rates could be low simply because Treasury rates are
low, not because the risk premium is low. However, these impacts
can be measured indirectly by the level of CRE debt outstanding —
the more investors are willing to use debt, it intimates that they
prefer taking on risk and/or they perceive risk as being low. This
process results in a compression of the risk premium and consequently
cap rates. Because debt is stated in nominal terms, the absolute
level of outstanding debt is misleading because it will rise over time
as values rise. Moreover, using loan-to-value ratios can also be
misleading because inflated property values can mask excessive
risk taking because the denominator of the loan-to-value ratio
rises along with the numerator and can artificially underestimate
the level of debt and risk.

Therefore, we need to measure the level of debt relative to an
independent factor, like gross domestic product (GDP). Although
property values tend to be positively correlated with the overall economy, gross domestic product is a flow metric, not a valuation
metric, and because it encompasses the entire economy it is much
harder to manipulate or artificially inflate. Because debt outstanding
is stated in nominal terms, the nominal GDP must be utilized. In
Exhibit 5, the ratio of total CRE debt to nominal GDP is calculated
over time.

Exhibit 5

Bubble 5

At the height of the bubble in 2008, when the perception of risk
and the risk premium were excessively low, the ratio of total CRE
debt outstanding to nominal GDP was roughly 24%. This was the
historical peak of this ratio and was reflective of an environment
where the incorrect perceptions of risk pushed valuations up and
the risk premium down. However, as of the most recent quarter of
available data, this ratio has not even surpassed 20%. Therefore,
risk perceptions and the risk premium remain at relatively safe
levels. The 4% or so difference between the current and peak
ratios translates into roughly $720 billion of net new loan issuance,
based on the current nominal GDP. At a rate of $120 billion of net
new CRE debt per year, it would take six years to reach that 24%
level again, assuming no nominal GDP growth. If the economy
continues to grow as most expect, the ratio might only drift marginally
higher, much as it has over the last few years. Therefore, the current
ratio, which is roughly equivalent to the level from 2003 before
the bubble truly began to inflate, does not indicate excessive risk
taking, a mistakenly low perception of risk, or too low of a risk
premium. If anything, it reinforces the notion that the risk premium
prior to the downturn was far too low. Using the current market as a guide, the risk premium from before the recession was probably
roughly 200 basis points below where it should have been. In this
context, today’s cap rates seem more or less appropriate while
those from before the downturn seem far too low, even though cap
rates from those two periods are incredibly similar.


There is ample evidence to indicate that the current market is not
a bubble. The ten-year Treasury rate, though low, is appropriate
as a function of the current economic environment, not monetary
policy. Cash flow growth, though positive, is low relative to the
period leading up to the recession and is not resulting in underwriting
that is incongruent with probable cash flow growth rates. Lastly,
debt is not being over-utilized as was the case during the prerecession
period. Investors are generally far more realistic and
restrained during this phase of the cycle using a more appropriate
risk premium.

However, this does not eliminate the possibility that another bubble
could form in the future. As interest rates rise with the continuing
recovery in the economy, the risk premium should decline, keeping
cap rates in a relatively narrow range. However, if discipline begins
to erode and the risk premium compresses too much, it is not
too difficult to envision another environment where cap rates fall
further than they should and market values exceed intrinsic values.
We are not ruling out the distinct possibility of a severe shock
prompting an economic downturn, which might then lead to CRE
fundamentals deteriorating. However, even if such a scenario takes
place, careful analysis will need to be performed before ex post
blanket statements like “Oh, there must have been a CRE bubble”
are issued. Accepting the “bubble” explanation after a downturn
occurs without nuanced thought is sloppy logic.