|
| I. Introduction |
|
The following simulations indicate that pre-development land, packaged as a hybrid debt security consisting of a zero coupon bond plus a 30% equity kicker, offers an effective hedge against inflation. To be more specific, vacant land, unlike bond, stocks, commercial or residential real estate, appears to have a large and significant positive regression coefficient on unanticipated inflation, i.e., changes in the inflation rate.
Therefore, land is unique among asset classes in that it may be financed with a zero coupon debt instrument, which has a large and significant negative coefficient on unanticipated inflation, plus an equity kicker which has the opposite effect. The result is that, taken together, the zero coupon bond and the kicker have a value that is remarkably invariant to changes in the inflation rate and is a powerful hedge against unexpected inflation or disinflation.
|
|
| II. Description of the Simulation Model |
|
To ascertain the inflation-related behavior of real estate, the study regressed to real estate returns (R) on both inflation rates (l) in a given year and changes in inflation rates (I) from one year to the next. This is expressed algebraically as:
R = A + B (I) + gamma (delta I) + E
Alpha corresponds to the average (over time) real return on real estate. Beta times corresponds to the average inflation rate. The sum of these two components is the expected return on real estate, assuming no change in the inflation rate. Next, gamma times the change in (delta-l) represents the reaction of real estate to inflation rate changes. Epsilon is the error term from the regression. The reader familiar with the bond pricing theory will note that this model is analogous to the stochastic process for bond returns, from which we took this formulation.
The following table shows results from the above regression equation.
|
Table
1
Regression of Real Estate Returns on Inflation Rates and Inflation Rate Changes*
|
| Asset
Class |
Period
Studied |
Alpha
(/ Year in %)
|
Beta
(Sensitivity to Inflation Level)
|
Gamma
(Sensitivity to
Inflation Rate Chg)
|
Farmland |
1960-84 |
7.0 |
.75 |
1.35** |
Residences |
1960-84 |
4.5 |
.85 |
-.15 |
Business
RE*** |
1960-84 |
4.7 |
.74** |
-.40
|
RE
Composite*** |
1960-84 |
5.3 |
.78** |
0.11 |
* The source of the real
estate return series is Roger G. Ibbotson
and Laurence B. Siegel, "Real Estate Returns: A Comparison
with Other Investments", American Real Estate
and Urban Economics Association (AREUEA) Journal,
Fall, 1984, updated by Ibbotson Associates
for this study. The source of the inflation
series is "Stocks, Bonds, Bills, and Inflation:
1986 Yearbook", Chicago: Ibbotson Associates,
1986.
** Statistically
distinguishable from zero at the 95% confidence
level.
*** Business
real estate includes office buildings, shopping
centers, rentable industrial properties, and
certain large apartment buildings. Real estate
composite is a market-value-weighted average
of farmland, residence, and business real estate
indices. All returns include net operating
income (i.e., they are total returns).
These regression results indicate that land
(for which farmland is used as a proxy) is
unique among the categories of real estate
in that it has a significantly positive gamma.
That is, it is a hedge against unanticipated
inflation. Assets with a positive gamma will
perform better when inflation is higher than
expected; those with a negative gamma, such
as bonds, will do the opposite.
Note that the regression results indicate that land has a return of 7.0% in the absence of inflation and changes in inflation; that is, it has a real return of 7%. This return is significantly smaller than the expected return for pre-development land, as indicated in Harry Roberts' report. The expected return appears to be well over 50%. This real rate of return, along with the beta and gamma of land are used to formulate a forecast model for land returns (R) as follows:
R = 7 +
XS + beta (I) + gamma (delta I)
In words, Return
on land = real return + effect of expected
inflation + effect of unexpected inflation.
Beta and gamma
define the magnitude of the effect inflation
and changes in inflation, respectively, have
on land returns. The new variable XS refers
to the "excess" return over the farmland benchmark
(based on Myers' differing use of land for
other than farm uses and their superior track
record) which ascribes to various scenarios
of management success.
We regard that
it is worth examining three management performance
scenarios: aggressive, intermediate, and unfavorable.
The unfavorable scenario is that the purchased
land rises only at the inflation rate, say
5%, which is less then the expected bond yield
of 11%. In this scenario, the bonds partially
default and the equity kicker is worthless.
The other two scenarios are studied in much
greater detail by performing simulations of
the value of the bond, equity kicker and the
total (bond plus equity kicker). Each scenario
is studied twice, once for unexpected decreases
in inflation, and once for unexpected increases.
Thus, we have:
|
Table
2
|
| Performance
Assumption |
Economy
|
Bond
of Yield at Issuance |
Land
Portfolio
Excess Return Over Benchmark |
Expected
Inflation |
Expected
Inflation Change |
| Aggressive |
Inflationary |
11%
|
25%
|
5%
|
Rises to 8%
|
| Aggressive |
Disinflationary |
11%
|
25%
|
5%
|
Falls to 2%
|
| Intermediate |
Inflationary |
11%
|
5%
|
5%
|
Rises to 8%
|
| Intermediate |
Disinflationary |
11%
|
5%
|
5%
|
Falls to 2%
|
All
yields, excess returns, and inflation rates
in the above table are in percent per year. A
25% per year excess return is remarkable compared
to other investments. Given the report of
Professor Harry V. Roberts on the performance
of the Myers' portfolio, it does not seem
unreasonable to use 25% as an aggressive
assumption; since the dollar-weighted return
on the Myers' portfolio was in the range
of four times that. The intermediate scenario
assumes that, in real estate markets are
being relatively inefficient, a skilled but
not extraordinary manager could out perform
the benchmark by 5% per year.
We assume that
bond yields track inflation exactly when inflation
rates change; that is, we assume parallel shifts
in the yield curve. Bond yields are assumed
to exceed the inflation rate by a constant
6%.
We further assume
that the 3-point shifts in the inflation rate
occur over a 6-year period, with inflation
rising or falling by 0.5 percent per year,
then stabilizing. The bond is assumed to have
a maturity of 10 years. Likewise, the simulations
cover a 10-year period.
The bond value
is determined by discounting the face (ending)
value at the current bond yield. This valuation
of the bond is referred to as the DCF (discounted
cash flow) value of the bond. The equity is
calculated as the land value minus the bond
value. If the equity value were to fall below
the DCF bond value, we would consider that
the value of the bond is limited by, and hence
equal to, the value of its collateral (the
land). There are no such occurrences in these
simulations.
| III.
Conclusion |
|
Real estate is traditionally perceived as a hedge against expected inflation and is effective as such. The hybrid structure outlined here is unique in that it is also a hedge against unexpected changes in inflation.
This results from combining a bond, which is a disinflation hedge, and land-based equity, which moves opposite the bond in response to changing inflation.
Thus, the hybrid has a near zero inflation-change gamma over most of the life of the investment. Over the full 10-year span, the gamma will be positive because that of the underlying asset, land, is positive.
The period of time over which the proposed structure has a near zero inflation change, gamma is considerable, 6 to 7 of the 10 years of the investment's life. After that, the gamma moves sharply toward that of the land itself. If one wanted to operate the investment program so that it would have a zero gamma, or as close to that as possible, one would sell the land after 6 or 7 years. This is not necessarily optimal if investment performance is good, however, because an investor who has been earning extraordinary returns for years is likely to want to continue to do so, and might by then be indifferent to the rising inflation gamma of his portfolio.
The Miller-Modigliani (M and M) separation principle suggests that the assets should be managed independently of the financing, a principle that should be kept in mind in the decision making process for the Myers portfolio. In the M and M framework, the land should be sold when the manager believes its highest value (after accounting for the opportunity cost of tying up money) has been reached.
As mentioned previously, land appears to be unique among assets (including other forms of real estate) in its ability to act as an effective hedge against unanticipated inflation.
It should be emphasized that the disinflation hedge effect of the bond applies only to interim transactions during the course of the investment program. For an investor holding the hybrid over its whole life, the bond provides a senior claim but not a disinflation hedge.
It may be possible to structure various "branches", or sets of cashflows, which have negative, zero, and low and high positive inflation betas. These attributes may prove appealing to widely differing investor clienteles and help maximize the quantity of capital that can be raised for The Myers Group Investment-Grade Pre-development Land Investment Program.
|
|
Roger
G. Ibbotson
Roger G. Ibbotson is Professor in Practice of Finance, Yale School of Management; and President of Ibbotson Associates, Inc. Professor Ibbotson received his Ph.D. in finance and economics from the University of Chicago in 1973. He coauthored the standard reference work, Stocks, Bonds, Bills, and Inflation, with Rex A. Sinquefield originally in 1977 and the book has been updated in numerous editions. A book with Gary P. Brinson, Investment Markets, was published by McGraw-Hill in 1987. Professor Ibbotson has written numerous articles including three Graham and Dodd Award winning articles; many working papers and unpublished works; has spoken before a wide variety of scholarly and professional audiences; and has appeared frequently in the broadcast media.
Ibbotson
Associates, which was founded by Professor
Ibbotson in 1977, is a consulting firm
specializing in the application of financial
theory to practical problems in asset
management, legal economics, and corporate
finance. It has offices in Chicago and
New Haven.
Laurence B. Siegel
Laurence B. Siegel is Managing Partner of Ibbotson Associates, Inc., and Instructor in Economics at the Kellogg Graduate School of Management. Mr. Siegel received his MBA degree in finance and economics from the University of Chicago in 1977. He joined Ibbotson Associates in 1979. Mr. Siegel is the author of numerous articles on investment returns, asset management, and related topics, including a Graham and Dodd Award winning article coauthored with Roger G. Ibbotson and Sinquefield's Stocks, Bonds, Bills, and Inflation, and has spoken before professional and academic groups.
Mr. Siegel
manages the Chicago office of Ibbotson
Associates, where he is the lead consultant
on numerous client projects and manages
a research staff. He also assists Roger
Ibbotson and other partners and associates
in their work. |
|
