Commercial Capitalization Rates and 10-Year Treasuries

Had the opportunity to participate in a panel on industrial real estate this month at a national commercial conference. I was taken aback by a broker and fellow panel member that truly believed that rising interest rates would not result in a corresponding increase in capitalization rates. The panelist’s denial of a likely increase in capitalization rates was reminiscent of the mass of people at the peak of the housing bubble that truly believed that home values would never decline. I created an Axiom for these people:

There is No Law of Real Estate That Says Values Will Never Go Down. It’s a question of when property values will decline, not if.

Fundamental to the income approach to value in the appraisal process is the infamous IRV formula:

12-1-16 formula

As capitalization rates rise, value decline assuming that all other factors remain the same, and vice versa. For my attorney friends, fellow economists and speakers of Latin, the term is ceteris paribus. Overlooked, however, is that interest rates increase when either or both inflation rises or the demand for capital increases due to growing economic activity. Under both of these, rents typically escalate, assuming that supply is not outpacing demand. Thus, increasing capitalization rates do not necessarily equate to declining property values. Value is a function of both the capitalization rate and net operating income.

So what is the relationship of commercial capitalization rates and 10-year constant maturity Treasury Bonds? The following graph shows the quarterly commercial capitalization rates for the U.S. for all commercial property types (from Real Capital Analytics) and corresponding 10-year Treasury rates. There are two differing estimates of quarterly Treasury rates. The first uses the rate the market closed at on the last day of the month. The second does an average for the quarter of the daily closing rates. Both are included in the following graph, with little variability between the two. For analytical purposes, the quarter-end yield was selected (the black line).

12-1-16 table1


When examining the spread between commercial capitalization rates and 10-year Treasury bonds, the following metrics were calculated:

12-1-16 table2
There is a relationship between capitalization rates and 10-year Treasury bonds that is positive and statistically significant. The following table shows the correlations for both the parametric Pearson’s Product Moment Correlation Coefficient and the nonparametric Spearson’s Rank Correlation Coefficient, along with the respective P-Values.

12-1-16 table3
How much of the variability within capitalization rates is a function of interest rates? While there is no doubt a statistically significant positive relationship between cap rates and 10-year Treasury Bonds, much of variability in cap rates over time cannot be explained just by knowing interest rates. From a modeling perspective, 10-year Treasuries account for approximately 30 percent of all the variability in capitalization rates.

Since many commercial transactions take up to a year to close, perhaps more important than the 10-year Treasury rate at the time of closing is what the Treasury yield was across that time period. The graph below has the same capitalization rates but with the 10-year Treasury now shown as the trailing four quarter – trailing 12-month (TTM) – moving average.

12-1-16 table4
Following are the related metrics under the TTM (less-variable) 10-Treasury yields:

12-1-16 table5
The current spread between the capitalization rate and the TTM 10-year Treasury is 458 basis points. This implies that, compared to the mean and median spreads, current capitalization rates are from 53 to 84 basis points lower than normal. My expectation is that capitalization rates will increase. To move to what has been the norm, cap rates would increase from the current 6.19 percent to a range from 6.73 to 7.02 percent. To maintain current values under this assumption, however, requires net operating income to increase in a range from 8.5 to 13 percent. I do not anticipate that large of a rental increase to occur, however, under current market conditions. This implies that values will be under pressure.

In some commercial transactions underway today scheduled to close by year-end, buyers have renegotiated a lower price due to the increase in interest rates. Offsetting this somewhat, however, is the expectation that under a Trump presidency, the 3.8 percent Net Investment Tax put in place to help fund Obamacare (effectively an increase to the capital gains tax) will be repealed.

In the future, other factors will be examined for potential explanation of the variability in cap rates including:

  • Alternative Investment Returns (S&P500, Nasdaq, Commodities, Oil)
  • Prior and Current Real Estate Returns
  • CommerInflation Rates (though this is redundant somewhat to interest rates)

The bottom line is that there is a positive, statistically significant relationship between capitalization rates and 10-year Treasuries. Compared to rates since 2001, current capitalization rates are 53 to 84 basis points lower than normal. My expectation is for an increase in capitalization rates from current levels even if interest rates do not rise further.

If you have any other suggested factors to include in a model for capitalization rates, please email back.



  1. Mr. California

    UNBELEIVABLE! I remember having a heated debate with the broker of a residential RE office at the peak of the market years ago….he truly believed that home prices would continue to grow at 22% per year indefinitely in his market…I did some napkin math for him and he still argued that in 3 years, the buyer profile would still be able to buy these overpriced houses!!!

  2. Anna

    If the current spread between cap rate and treasuries is 458 vs 375 and 403 mean and median, wouldn’t that mean cap spreads are more likely to move down in mean reversion rather than up?

  3. Anna

    If current spread is 458 vs 375 average doesn’t it imply cap rates are too high and should move down?

Leave a Reply