December
13, 2018
Robert
F. Telewicz, Jr.
Branch
Chief
Office
of Real Estate and Commodities
Securities
and Exchange Commission
Division
of Corporation Finance
100 F
Street, NE
Washington,
D.C. 20549
Re:
Cresud Sociedad Anonima Comercial Inmobiliaria Financiera y
Agropecuaria
Form 20-F for the fiscal year ended June 30, 2017
Filed October 31, 2017
File No. 001-29190
Dear
Mr. Telewicz:
On
behalf of our client Cresud Sociedad Anonima Comercial Inmobiliaria
Financiera y Agropecuaria (the “Company”), we are
writing to respond to comments raised in the Staff’s comment
letter dated November 20, 2018 (the “Comment Letter”)
relating to the above-referenced annual report (the “Annual
Report”) of the Company filed on October 31, 2017, pursuant to the Securities Exchange Act
of 1934, as amended.
We are
providing the following responses to the comments communicated by
the Staff in the Comment Letter. For convenient reference, we have
reproduced below in bold the text of the comments from the Staff.
The responses and information described below are based upon
information provided to us by the Company.
Form 20-F for the fiscal year ended June 30, 2017
Item 5. Operating and Financial Review and Prospects
A. Operating Results
Results of operations for the fiscal years ended June 30, 2017 and
2016
Changes in fair value of investment property, page 122
1.
We
understand that rents from your shopping mall leases are indexed to
tenant revenues, that you expect tenant revenue growth to be
correlated to Argentine inflation, and that you attribute the
appreciation in shopping mall fair values relative to office fair
values to those expectations. Please quantify the effect of indexed
rents on your shopping mall fair value measurements. Please explain
how the increase in nominal cash flows due to indexed rents is, or
is not, offset by the use of a nominal discount rate.
The
Company acknowledges the Staff’s comment and respectfully
advises the Staff that in the medium- and long-term the
appreciation in the fair value of shopping malls is not caused by
inflation expectations. This is so because increases in nominal
cash flows that result from increases in indexed rents are offset
by using a nominal exchange rate that incorporates a rate of
depreciation which is equal to the rate of inflation used for the
Company’s projections. Therefore, and considering the offset
effect, the fair value of shopping malls fluctuates primarily as a
result of the increase in rents in real terms and the perceived
investment risk of shopping mall properties as reflected in the
discount rate.
Notwithstanding,
the Company advises the Staff that in the short-term it is possible
that the impact of the implicit inflation rate on the nominal
exchange rate may be inconsistent with the market’s inflation
expectations, which in turn will delay the forecasted correlation
between Argentine currency depreciation and the rate of inflation
included in the model. This scenario is more likely to occur when
there is friction in the market caused by, among other factors,
foreign exchange controls, import restrictions, large influxes of
foreign direct investment, direct investment, speculative capital
inflows in the financial market and commodity price increases that
result in significant inflows of foreign currency. All of these
factors were present in the Argentine economy during the years
being analyzed. In such scenario, the rate of inflation used in
cash flow projections may be higher than the rate of depreciation
of the Argentine peso, which would lead to an increase in shopping
mall fair value. If, on the other hand, the rate of inflation is
projected at a lower rate than the depreciation of the Argentine
peso, the fair value of the Company’s assets will
decrease.
In
summary, the Company advises the Staff that the fact that rents are
indexed by inflation is not a value driver itself. However, the
significant fluctuations in the Argentine economy during the
periods under analysis which generated the macroeconomic imbalances
and distortions referred to above, have resulted in changes to the
fair value of shopping malls in the short-term.
The
Company acknowledges the Staff’s comment regarding the
quantification of the effect of indexed rents on the
Company’s shopping mall fair value measurement, and advises
the Staff that the total changes in fair value of such properties
in fiscal years 2017 and 2016 derived from changes in projected
cash flows amount to AR$1,794 million and AR$5,826 million,
respectively. Of these amounts, for fiscal year 2017, AR$834
million result from the changes in indexed rents (derived from the
net effect of changes in expected inflation and expected
depreciation of the local currency, as compared to the previous
measurement date). The corresponding amount for fiscal year 2016
amounted to AR$(169) million.
For a
detailed description on the impact of the exchange rate and
inflation on our shopping malls valuation model, as well as
assumptions used by the Company to offset short-term imbalances in
the medium- to long-term, please refer to Exhibit “A”
hereto.
2.
We
understand that you do not determine the discount rate implied by
the fair value of your office properties for which fair value is
measured using comparable transactions. IFRS 13 requires
maximization of relevant observable inputs; please explain why the
discount rate for office properties is not relevant for measuring
the fair value of other commercial investment properties, such as
shopping malls.
The
Company acknowledges the Staff’s comment and respectfully
advises the Staff that, as the question duly notes, the fair value
of office properties is calculated based on the value of comparable
market transactions (price/square meters). Therefore, given that
the valuation of office properties is not based on a discounted
cash flow methodology, the valuation of office properties does not
have an explicit WACC.
Notwithstanding the
above, and for clarification purposes, the Company has provided the
hypothetical valuation exercise set forth below. The purpose of the
hypothetical exercise is two-fold: first, to determine the discount
rate implicit in the value of office properties, and second, to
illustrate why the Company considers that the office property
discount rate as calculated, is not relevant for value measurements
of shopping malls.
The
theoretical, implicit discount rate for the office property segment
arrived at below was calculated using the inputs from market
participants, including: rent per square meter, adjustments to
EBITDA to arrive at free cash flow (i.e. Capex, income taxes) and value per
square meter of Class A/CABA office properties derived from market
sales.
As
depicted in the hypothetical valuation exercise below, if the
discount rate for office properties is applied to value the
Company’s shopping mall properties, the fair value of
shopping malls would be significantly higher.
Chart 1: Hypothetical Valuation Exercise
Additionally, for
purposes of showing on an empirical basis why the discount rate for
offices is not an adequate measure to value shopping malls, and the
differences between the two segments, the Company has included the
following charts below which show the historical trend of the
office property implied discount rate and cap rate as obtained from
market participants (where cap rate is the main valuation metric
recognized for this segment) and the historical trend of the
shopping mall discount rate. As depicted below, the cap rate for
office properties has fluctuated in a range of 7-8% due to the
segment’s prevailing characteristics and the
properties’ stabilization stage (i.e. low vacancy rates, stable rents in
dollar terms per square meter). The office property market is
dominated by a small group of investors and owners which seek
medium- to long-term leases, who perceive real estate as a safe
dollar-denominated investment option. In contrast, the shopping
mall segment is a relatively new industry in Argentina where the
first shopping mall opened in 1990, compared to markets such as the
United States and Brazil where the industry began in the
1950’s and 1960’s, respectively. Additionally, unlike
the office properties segment, the financial performance of
shopping mall properties is highly correlated with the volatile
economic activity in Argentina since the cash flow generated by
shopping malls is closely related to the purchasing power of
customers.
Chart 2: Implied WACC of Market Participant Office Properties vs
Company’s Shopping Mall Segment WACC
Source:
J.L. Ramos as of
October 2018 (Calendar year)
Company
information (Fiscal year)
Chart 3: Cap Rate & Prices for Office Properties (Market
Participants)
Source:
J.L. Ramos (October 2018)
Calendar Year
2013/14/15: Rent is calculated according to the Contado con Liquidación implicit
exchange rate derived from the price in Pesos of shares listed on
the Buenos Aires stock exchange and ADRs listed for trading on an
exchange in New York, due to the exchange rate controls imposed by
the Argentine government that generated distortions in the official
exchange rate.
Chart 4: Cap Rate & Rents - Offices Properties (Market
Participant)
Source:
J.L. Ramos (October 2018)
Calendar Year
2013/14/15: Monthly rent is calculated based on the Contado con Liquidación implicit
exchange rate derived from the price in Pesos of shares listed on
the Buenos Aires stock exchange and ADRs listed for trading on an
exchange in New York, due to the exchange rate controls imposed by
the Argentine government that generated distortions in the official
exchange rate.
Chart 5: Tenant sales per square meter & WACC
(Malls)
Source:
Company’s information.
The
Company also advises the Staff that historically there has been a
spread between the discount rate used for shopping malls and the
implied discount rate applied for offices properties. This spread
exists because investors in shopping malls have historically
expected a higher return than those investing in office properties,
primarily due to the increased volatility inherent in shopping
malls given their close correlation with the economic cycle
prevailing in Argentina.
As
described in our prior responses to the Staff’s comment
letters, there is no comparable cap rate available in Argentina for
shopping malls that the Company can use as a reference for
valuation. For illustration purposes, the Company has simulated the
valuation of shopping malls using the cap rate of market
participants for offices in Chart 6 below. As shown below, using a
relative valuation methodology (NOI/Office Properties’ cap
rate) leads to a significantly higher valuation than that given by
the discounted cash flow analysis the Company uses. For example,
for the fiscal year ended June 30, 2017, the fair value of shopping
malls obtained by using a discounted cash flow analysis was
AR$28,561 million, compared to a valuation of AR$37,541 million
when using the office property cap rate of 7.2%, a difference of
AR$8,980 million.
Chart 6: Valuation simulation of
shopping malls using office property cap rate (7.2%)* for the
fiscal year ended June 30, 2017 (amounts in
millions)
|
|
(1) Includes Patio Olmos and other shopping mall
annexes.
Note:
Exchange rate average of AR$15.54 per US$1.00. Exchange rate end of
period:
AR$16.53
per US$1.00.
*The
cap rate of market participants as provided by J.L. Ramos, for A+
properties in general.
3.
We
understand that you consider your weighted-average cost of capital
(WACC) when discounting future cash flows to measure the fair value
of your shopping malls. Please explain:
●
Why
your WACC is consistent with market participant
assumptions
The
Company acknowledges the Staff’s comment and respectfully
advises the Staff that, based on publicly available information
(primarily analyst coverage reports), there is empirical evidence
that the discount rate used in the valuation models for its
shopping malls falls within the range estimated by market
participants. Specifically, for the fiscal year ended June 30,
2016, the Company’s WACC for the shopping malls segment was
9.51% compared to a WACC of 9.54% WACC observed in the market. For
the fiscal year ended June 30, 2017, the Company’s WACC for
the shopping malls segment was 9.35% compared to an average WACC of
9.37% estimated by market participants.
Chart 7: Shopping Mall Segment WACC Analyst Coverage
●
How
your WACC is adjusted for different risks associated with each
investment property.
The
Company acknowledges the Staff’s comment and respectfully
advises the Staff that the beta coefficient is used to incorporate
shopping mall segment risk to the discount rate. In general terms,
beta is calculated by
averaging the beta of
comparable companies, including those which core business are
shopping malls.
Chart 8: Comparable Companies used to determine beta.
Source:
Bloomberg
Notwithstanding the
use of beta from comparable
companies, the Company advises the Staff that the individual risk
profile of each shopping mall (driven by, among others, geographic
location, target client segment and sales per square meter) is not
fully captured by the WACC. Instead, a shopping mall’s risks
are captured by the projection of cash flows. The Company uses its
robust set of historical EBITDA margin data as the basis to project
future growth of that metric for each shopping mall property and
incorporates into the model the individual performance of each
shopping mall.
4.
If
the discount rate for office properties is not relevant to the fair
value measurement of shopping mall properties, please explain why
it is appropriate to use your enterprise WACC to discount the
future cash flows of your shopping mall properties. That is, since
you hold both shopping mall properties and office properties, if
you believe that shopping mall discount rates are not relevant to
shopping malls, explain why your WACC would be appropriate for
shopping malls.
The
Company acknowledges the Staff’s comment and respectfully
would like to clarify to the Staff that references to the
“Company WACC” in previous discussions and responses,
referred to the Company’s shopping mall segment WACC, and not
the enterprise WACC of the Company as a whole. The Company advises
the Staff that it considers that shopping mall WACC is the adequate
discount rate because it reflects the return expected by investors
in the sector, which is higher than the expected return for office
properties.
Additionally, the
value reflected by the Company’s discount rate is consistent
with market valuation methodologies. To confirm the latter, the
Company has analyzed the WACC for each period using parameters from
other market participants (capital structure, cost of debt,
etc.).
Chart 9: WACC Malls
5.
We
understand that you use forward USD/ARS forward rates implied by
the onshore futures market for one year to measure the fair value
of your shopping malls, and then revert to long-term relative
inflation rates in the two currencies. Please explain how you
consider other observable sources for forward foreign exchange
rates. For example, we observe that certain financial data
aggregators publish estimated non-deliverable forward foreign
exchange rates for periods up to five years at certain valuation
dates.
The
Company acknowledges the Staff’s comment and respectfully
advises the Staff that, from its perspective, it considers that
Argentine market information sources such as macroeconomic
consultants, REM and/or ROFEX have better capabilities to predict
future exchange rates than the NDF off-shore curve. It is the
Company’s understanding that local information sources
incorporate market expectations more effectively in their
projections due to their familiarity with the particularities and
volatility of the Argentine market. Additionally, forward offshore
contracts have certain traits which cause their price to be less
transparent and an unreliable predictor of future exchange rates.
These traits include: over-the-counter trading (unregulated and
shallow market), significant bid-ask spread in the medium-term, and
the shallow trading of contracts with terms greater than 6-month
which are not representative and are operated by very few
counterparties (6 compared to 192 counterparties that operated in
the ROFEX).
Additionally, as
shown in the Bloomberg image reproduced below, there are no values
for terms that exceed three years. When the currency suffered a
100% depreciation this year and a further 100% depreciation for the
following year, it seems that the foreign exchange information
provided by non-deliverable forward foreign exchange rates is not
representative of the current currency situation, making it more
difficult for a market participant to validate the
information.
Chart 10: Evidence of long-term liquidity
Chart 11: Evidence of ‘Real time Bloomberg
contributors
Source:
Bloomberg.
Robert F. Telewicz, Jr.
Branch Chief
Office of Real Estate and Commodities
Securities and Exchange Commission
December
13, 2018
Chart 12: Evidence of Bid-Ask spread in medium term
Source:
Bloomberg.
Chart 13: Rofex Agents
Note: The
Central Bank of Argentina operates through a Special
Agreement.
Source:
ROFEX.
* *
*
Robert F. Telewicz, Jr.
Branch Chief
Office of Real Estate and Commodities
Securities and Exchange Commission
December
13, 2018
Please
do not hesitate to contact me (212 455-3066) with any questions you
may have regarding the above responses.
Very
truly yours,
By: /s/ Jaime
Mercado
Jaime
Mercado
cc:
Matias Gaivironsky
Leonardo
Magliocco
Eduardo
Loiacono
Mariano
Tomatis
David
L. Williams
Eric
McPhee
Robert F. Telewicz, Jr.
Branch Chief
Office of Real Estate and Commodities
Securities and Exchange Commission
December
13, 2018
Exhibit A
In its
financial modelling the Company offsets the increase in medium- and
long-term cash flows in Argentine pesos as a result of inflation is
offset by using a nominal exchange rate for cash flow projections
made in U.S. dollars. In other words, free cash flow in Argentine
pesos is estimated in nominal terms based on Argentine inflation
and expected real GDP growth. As a result, cash flows in Argentine
pesos are converted to U.S. dollars at the projected nominal
exchange rate which incorporates the same inflation rate used for
projections in the medium- and long-term. Thereafter, the free cash
flows converted to dollars are discounted at a nominal discount
rate based on the Company’s weighted-average cost of capital
for the shopping malls segment in U.S. dollars. For illustrative
purposes only, we have included the following numerical example
that illustrates the effect of inflation on the valuation model and
how such effect is offset by using the nominal projected exchange
rate:
Chart A-1: Hypothetical example of the impact of inflation on
FCFF
(Assuming GDPreal
growth = 0%)
As
depicted in Chart A-1 above, if the depreciation of the Argentine
peso is projected at the same rate as the rate of inflation and if
the discount rate remains constant, there is no change to the net
present value of the Company’s assets.
Chart A-2: Hypothetical example of the impact of inflation on
FCFF
(Assuming GDPreal
growth =
5%)
Additionally, as depicted in Chart A-2 above, if
the depreciation of the Argentine peso is projected at the same
rate as the rate of inflation, and if the discount rate remains
constant but
there is assumed GDP growth, the net
present value of the Company’s assets will increase as a
consequence of GDP growth in real terms.
The
charts below show an actual valuation exercise for one of the
Company’s shopping malls. As illustrated, the valuation
supports the same conclusions reached from the theoretical examples
in Charts A-1 and A-3 above. As shown below, a change in medium- to
long-term inflation of 8% to 5% and its resulting effect on the
exchange rate (calculated using the following equation {(1+
Consumer Price Index Argentina) / (1 + U.S. Consumer Price
Index)-1}), will not lead to a change in Alto Palermo shopping
mall’s net present value.
Chart A-3: Alto Palermo Discounted Cash Flow Valuation
(amounts in millions, assuming Argentine inflation constant at
8%)
Chart A-4: Alto Palermo Discounted Cash Flow Valuation
(amounts in millions, assuming Argentine inflation constant at
5%)
The
following Charts A-5 and A-6, show how the exchange rate in
Argentina appreciated both in periods with increasing commodity
prices and in periods with increased foreign direct investment and
speculative capital inflows in the Argentine financial markets.
Consequently, by assuming that the nominal exchange rate will track
inflation in the medium to long-term the model corrects the
unbalances caused by macroeconomic friction but only to the extent
that the situations causing such friction are temporary and
unsustainable for prolonged periods of time. The Company’s
free cash flow and the terminal value are both discounted under the
latter assumption.
Chart A-5: Evolution of real exchange rate (USD/AR) and commodity
prices
Source:
Central Bank of Argentina (Index base 2015=100) / World Bank
(Commodity Index based on nominal US dollars,
2010=100)
Chart A-6: Evolution of foreign investment in
Argentina
Source:
Central Bank of Argentina