Evaluating the Decision to Own Corporate

There is no single answer as to ONEOFTHEMOST important cap- ital decisions made by corporations is whether a company should own whether they should own or their operating real estate (offices, industrial or lease its real estate. and warehouse facilities, and retail space). This decision is generally viewed by cor- porations as a trade-off between the pres- ent value of rental payments versus that of the operating costs of owning the real estate, net of expected capital appreciation and the depreciation tax benefits from ownership. The rule of thumb is that only if the present value of future rent is less than the present value of costs of self-own- ership of the space (net of depreciation

PETERLINNEMAN benefits, and expected apprecia- FRANKPFIRSCHING tion), should the firm lease rather than

7 4 ZELL/LURIEREALESTATECENTER own. However, as this paper demonstrates, WHATDOCOMPANIESDO? this analysis is fundamentally flawed, lead- ing companies to own far more corporate Germany is the third-largest economy in real estate than is economically justified. the world but remains relatively inefficient This is true in countries such as Germany, in terms of capital allocation and the man- where corporate users own as much as 75 agement of corporate capital. As a result, percent of their real estate, as well the some 75 percent of corporate real estate is United States, where roughly 40 percent is owned, one of the highest proportions in owned by corporate users. the developed world. In this decade, some The correct model for the own-versus- tentative steps have been taken by major lease decision must compare the present German companies to reduce their owner- value of profits the corporation expects if ship of corporate real estate. However for it versus the present value of expect- the most part these efforts have followed ed profits if it decides to own its real the general pattern, which is to dispose of estate. The key insight provided by this corporate real estate only when there is no corrected approach is that the own-versus- alternative for raising capital, or to find a lease decision revolves around the com- gimmick to remove the ownership from parison of the lost profits associated with the company’s balance sheet even though moving corporate capital from core oper- ownership remains. ations to real estate, versus the profits For example, in 2000, Germany’s lead- achieved by real estate owners. That is, ing retailer, Metro, transferred all of its 357 capital freed up from real estate ownership real estate assets, which included depart- generates the company’s core rate ment stores, supermarkets, and do-it-your- of return, while rents reflect the rate of self stores, into a joint venture (JV). Metro return earned by on their real retained a 49 percent ownership, with a estate capital. Since most companies have major mortgage bank owning 49.5 percent higher expected rates of return in their and a large insurance company owning the core business than are achievable through remaining 1.5 percent. The transaction was real estate ownership, this decision model worth ⇔2.7 billion, yielding approximately indicates that the vast majority of corpo- ⇔1.3 billion in cash for Metro. This trans- rate real estate should be leased. The intu- action resulted in Metro leasing these prop- ition of this result is simply that by mov- erties on a long-term basis from the JV and ing capital from low-yielding real estate to successfully moved the assets from Metro’s high-yielding core operations, companies balance sheet. In 2002 and 2003, Metro increase profits. attempted to sell the entire JV to third-

REVIEW 7 5 party investors. But after two failed by Siemens. The open-end fund subse- attempts with buyer consortia, the attempt quently acquired additional real estate, was abandoned. Metro ultimately repur- bringing in additional shareholders beyond chased the ownership shares from the Siemens’ pension trust. Recently, Siemens insurance company and mortgage bank, sold the fund management company to a resulting in a reconsolidation of the real third party. In a similar transaction, estate onto Metro’s balance sheet, largely Dresdner Bank transferred a portfolio of purging the cash raised in the original sale. 300 bank buildings to a newly established State-owned Deutsche Bahn also spun open-end real estate fund managed by its off all of its land and development proper- wholly owned subsidiary. In this case, ties into a JV. This transaction raised shares in the open-end fund were never approximately ⇔1.1 billion for Deutsche offered to other investors. In 2005, Bahn and was structured almost identically Dresdner Bank sold this fund to to the Metro transaction. Like Metro’s Eurocastle. There have also been a number transaction, this transaction was also of smaller transactions where one or more reversed when Deutsche Bahn subsequent- were sold to third-party pur- ly repurchased the shares of the JV chasers in sale-lease back transactions. For investors. The real estate assets were subse- the most part, however, German corporate quently sold in September 2007 to a JV of owners continue to retain their corporate the German construction company real estate. Hochtief and investor Redwood Grove In an effort to identify why German International LP for ⇔1.64 billion. corporations have failed in their real estate In 1999, Siemens entered into an agree- monetization efforts and are so predisposed ment transferring eighteen properties to a to corporate real estate ownership, one of newly established open-end German real the authors (Pfirsching) has conducted estate fund. Siemens received approximate- detailed research documenting that the ly ⇔750 million for these properties from problems are both technical and strategic in the fund, and immediately contributed nature. Among key technical problems are: these proceeds to its corporate pension insufficient data quality and poor data fund to cover outstanding pension liabili- management; unprofessional management ties. This resulted in the transfer of the real of the transaction; the seller’s lack of knowl- estate as well as its pension liabilities to off- edge of the value of its assets; and demands balance-sheet status. The management of by investors for re-trades. The key strategic the open-end fund was provided for a fee problems are: lack of seller commitment to by a management company wholly owned the sale; unrealistic valuations, often based

7 6 ZELL/LURIEREALESTATECENTER on book value; seller changing of the port- the depreciation tax advantages and folio of assets available for sale during the expected appreciation on the corporate sale process; and changing space require- real estate. But the appropriate approach ments of the seller throughout the sale to evaluating the economic benefits of process. Additional studies have identified leasing rather than owning corporate real other reasons for the failure of sale and estate must compare the present value of monetization efforts of corporate real profits if they own their real estate versus estate, including: the seller’s unwillingness profits if they lease. That is, differential to take book value losses; the seller’s desire expected profits—not just costs—must for complete control over all real estate drive the analysis. assets; concern about image and reputation The present value of the after-tax prof- damage to the seller; preservation of social its associated with owning corporate real peace in the company; fear of losing key estate (πo) is equal to the present value of properties and sites to competitors; the lack after-tax profits from core operations, of off-balance sheet treatment if the seller ignoring the incremental costs of owning retains long-term control of the asset; poor corporate real estate costs (Po), minus the data quality; and higher perceived long- present value of the after-tax incremental term costs. costs associated with real estate ownership As in any business effort, it is essential (C), plus the present value of the tax sav- to have a clear strategy when it comes to ings associated with the depreciation disposing of corporate real estate. A lack of allowance provided to owners of corporate such clarity inevitably leads to an ineffi- real estate (D), plus the present value of the cient and ineffective process. But in addi- expected after-tax appreciation on the cor- tion, many corporate real estate executives porate real estate (A): appear to conduct faulty analysis of the πo = Po (1-t) - C(1-t) +D+A, benefits of leasing rather than owning their where t is the corporate income tax rate. real estate. The present value of core operations profitability (ignoring incremental real estate ownership costs) is equal to the THECORRECT rate of return (r) it achieves on capital DECISIONMODEL invested in core operations, times the capital it invests in core operations (K), The typical own vs. lease analysis evalu- times the company’s cash flow valuation ates the differential operating costs associ- multiple (M): ated with owning versus leasing, net of Po = r K M.

REVIEW 7 7 That is, core cash flow reflects a return rate on return (g) on the capital invested in of r percent, earned on each of the K dol- real estate (V). profits are equal to lars at work in the core business, and annu- the expected rate of return on real estate al cash flows are translated into value via (g), times the capital invested by the land- the firm’s cash flow multiple (M). lord in real estate (V), with this cash stream The corporate owner’s present value of converted to present value via the real incremental real estate costs, should it estate’s cash flow multiple (N): choose to own, can be expressed as the pro- πLL = g V N. portional costs ( ) relative to property The landlord achieves his profits via the value (V), times theα owner’s cash flow mul- present value of rental payments (R), plus tiple, which converts annual operating the value of the depreciation tax shield he costs to value: receives by owning the property (D), plus C = VM. the present value after-tax capital gains (A), The corporate owner’sα after-tax present minus real estate operating costs (CL): value of profits if it leases (πL) its corporate πLL = g VN = R+D+A-CL. real estate, and redeploys its real estate cap- Note that for any level of landlord prof- ital (V) in core operations, are the after-tax it, rents decline with the tax benefits of present value difference between core prof- depreciation and expected property appre- its (PL) and rental payments (R): ciation. It is also noteworthy that the πL = (1-t) (PL – R). depreciation (D) and capital gain (A) com- The present value of core profits if the ponents are the same irrespective of corporation leases (PL) are the core busi- whether the property is owned by a corpo- ness rate of return (r), times capital rate user or a third party landlord (up to employed in core operations (K+V), con- differences in effective tax rates). certed to value by the corporation’s cash The landlord may (or may not) be a flow multiple (M): more efficient provider of real estate from PL = r (K+V) M. the perspective of operating costs. For That is, core profits are higher because example, a landlord may achieve lower an additional V dollars are invested in core operating costs via scale economies, or operations rather than real estate. detailed operating expertise derived from Turning to the present value of rental greater experience or specialization. In payments, it is instructive to analyze the addition, the corporate owner may lack present value of the landlord’s profits (πLL). knowledge of real estate, and make “rook- Rents must be sufficiently high for the ie” mistakes that are avoided by a profes- landlord to achieve an expected required sional landlord. Alternatively, the opera-

7 8 ZELL/LURIEREALESTATECENTER tions of the real estate may be so unique reduce rents, resulting in a lower landlord and idiosyncratic that the corporate owner expected rate of return (g). In contrast, if has lower operating costs than a landlord. the property is such that this is little land- The presence of operating cost efficiency is lord competition, rents will be high, allow- easily captured by expressing the present ing landlords to achieve a higher rate of value of the landlord’s real estate costs (CL) return (g). This is the case for less devel- as lower (higher) than those of the corpo- oped property and geographic markets, as rate owner by e percent: well as for highly idiosyncratic properties CL = (1-e) C. for which landlords require a high rate of If e equals zero, landlords and corporate return. owners have the same operating costs. Substituting the operating costs expres- Alternatively, if e is greater than zero, land- sion (CL = [1-e] C) and rearranging the lords have lower operating costs by e landlord profit expression (πLL = g VN = percent, while if e is less than zero, the R+D+A-CL 7), yields the present value of property is so idiosyncratic (or landlords so rental payments (R) as: operationally inept) that operating costs are R = g VN + (1-e) C – D – A. lower if self-provided. For example, if e That is, rent is established in the mar- equals 0.05, the landlord is 5 percent more ketplace such that the landlord receives a g efficient, resulting in landlord operating percent rate of return on operating costs, costs that are 95 percent of the operating after realizing the return benefits of depre- costs of corporate owners. ciation (D) and property appreciation (A). The presence of landlord operating cost It may seem like a long process, but we are efficiencies (e>0) lowers rents for any given (after appropriate substitution and manip- landlord rate of return (g), meaning lower ulation) able to use these relationships to landlord operating costs translate into express the differential after-tax present lower rents. That is, to a large degree, the value profitability of leasing versus self- benefits of landlord operating cost efficien- ownership as: cies are passed on to tenants in the form of π = πL–πO = (1-t) V [(r+ e)M–gN]– t(D+A). lower rents. Δ While at first blushα this expression The nature of the relevant property looks obscure, it is actually very transpar- market will determine the rate of return (g) ent. The first term simply reflects the after- that the landlord expects to earn given tax value of deploying V dollars of capital operating costs (CL), depreciation (D), and in core operations rather than real estate appreciation (A) benefits. In a highly com- ownership. If the capital is employed in petitive market, competitive pressures core operations, they generate a rate of

REVIEW 7 9 return r+ e, which is converted to present of property costs (New York, Tokyo, and value by theα core operation valuation mul- London vanilla office space), and in low- tiple (M). This core return derives both inflation environments should rent their directly from the g rate of return earned on real estate. At the other extreme, firms core capital, plus any additional capital that using highly idiosyncratic space in non- is freed by arbitraging landlord operating competitive markets (e.g., specialized man- cost efficiencies (or inefficiencies) that exist. ufacturing facilities in developing coun- If the same V dollars of capital are tries, and corporate headquarters in small invested in corporate real estate, they markets), for whom core returns are mod- generate a rate of return of g, which is est, the firm is slow growing, while real converted to present value by the real estate appreciation is high, should own estate’s cash flow multiple. Finally, set their real estate. Of course, each firm and against this are the tax benefits of depre- property will have a unique combination of ciation and capital gains achieved by self- these factors. Further, what is optimal will ownership times the corporate tax rate. change over time as capital and real estate This later effect reflects the fact that since markets change. rents are lower due to depreciation (D) A key element of the ownership deci- and property appreciation (A) benefits, sion is the value arbitrage associated with only a fraction (t) of these benefits differ- draining capital from core operations. If entially flow to self ownership. core operations generate a relatively low This differential profit expression indi- rate of return (as is the case in many old- cates that a firm should own if π>0, that line ), while real estate returns are is, if generates greater after-taxΔ pres- high (as may be the case for expanding ent value profitability. Leasing is more prof- operations into non-competitive property itable in cases when: core returns are high; markets), ownership makes economic landlord efficiencies exist; the core cash sense. But since core returns are typically flow multiple is high; real estate returns are higher than real estate returns, renting low; real estate’s cash flow multiple is low; tends to be more profitable. the corporate tax rate is low; depreciation Another way to describe the return benefits are low; and capital gains expecta- arbitrage associated with renting is that tak- tions for the real estate are low. Hence, ing capital from assets generating 7 percent firms with high core returns (financial serv- to 10 percent returns (corporate real estate) ice and tech firms), that are growing rapid- and transferring the capital to core opera- ly, while using space in highly competitive tions that generate 10 percent to 15 per- markets, where land is a large component cent returns generates substantial value

8 0 ZELL/LURIEREALESTATECENTER gains. Hence, by converting dollars from ed, and real estate returns are reduced by EBITDA into rent, the firm can create greater competition, liquidity, and trans- value, as rent sells for a higher multiple parency, the ownership of corporate real than EBITDA for most companies. Of estate should decline. course, this entails designing a lease that allows sufficient operating control for the company to achieve the core return on its THREESIMULATIONS capital. If core operating ability is compro- mised by an imperfect lease, than the core Three simulated cases demonstrate the return would reflect this lack of control. own vs. lease decision (Table I). First, con- However, in markets with sophisticated sider the case of a “typical” firm. It has 35 legal systems this should rarely be a prob- percent corporate tax rate, a 12 percent lem, except with the most idiosyncratic core return, and a 13 cash flow multiple. operating facilities. The property return is 9 percent, and has a A key insight is that the more com- 13 cash flow multiple. Landlords have 10 petitive the real estate market is, the percent lower operating costs than corpo- greater is the incentive to rent, as compe- rate owners due to the commodity nature tition reduces rents. Thus, as more corpo- of the real estate and depth of the property rate real estate is sold to landlords, a vir- market, while self provision operating costs tuous cycle is created, as if all corpora- are 3 percent of value. The property is tions own their real estate, it is unlikely expected to appreciate 3 percent annually, that a competitive landlord market and there is a 20 percent effective capital evolves. But as real estate is sold by cor- gains tax, 2.5 percent of non-land is depre- porate real estate owners into the landlord ciable annually, and land accounts for 30 market, a deeper and more competitive percent of real estate value. For this compa- rental market evolves, reducing landlord ny/property/market combination, leasing returns, causing lower rents, encouraging generates a present value greater profit less corporate real estate ownership. equal to 24 percent of the value of the real More developed capital markets and estate. That is, every $100 million deployed more competitive property markets, such in corporate real estate destroys $24 million as those found in major U.S. markets, in corporate value. should have greater corporate leasing due The second case is a company with a to greater competition and landlord oper- mere 8 percent core return, and a seven ating cost efficiency. It also suggests that as times core cash flow multiple. In addition, the markets become more globally integrat- the company is evaluating a high idiosyn-

REVIEW 8 1 Table I: Simulation parameters Case 1 Case 2 Case 3 “Typical” Loaded towards Loaded towards situation freehold rental (e,g, (e,g, specialized assets) tech company) ≡ ΔΠ ≡ pre-taxDifferential rate Rentalof return to on Freehold core business 0.12024% -93%0.080 0.060 41% Mг ≡ after tax cash flow multiple 13.000 7.000 18.000 t ≡ effective corporate tax rate 0.350 0.350 0.350 V ≡ property value 1.000 1.000 1.000 B ≡ annual non-land depreciation allowance rate 0.025 0.025 0.025 a ≡ freehold property annual capital appreciation rate 0.030 0.050 0.030 N ≡ after tax real estate multiple 13.000 15.000 10.000 e ≡ landlord specialization efficiencies relative to freehold pre-tax costs 0.100 -0.100 0.200 ≡ freehold real estate costs as a proportion of value 0.030 0.040 0.030 nα ≡ discount rate for real estate appreciation 0.100 0.080 0.120 s ≡ capital gain tax rate 0.200 0.200 0.200 T ≡ years freehold property is held 40.000 40.000 40.000 g ≡ landlord return rate 0.090 0.120 0.050 l ≡ land (non-depreciable) share of value 0.300 0.300 0.300 d ≡ discount rate for tax shield (reflects risk of tax law change) 0.080 0.080 0.080 cratic piece of real estate for which their a more attractive case for ownership, as operating costs are 10 percent lower than a there is substantial arbitrage, lower owner landlord’s operating costs, while these costs operating costs, and substantial property are 4 percent annually of real estate value. appreciation associated with owning. The property is expected to appreciate at 5 The third case considers a firm leasing percent annually. Due to a non-competi- space in a highly competitive property tive real estate market, the landlord’s return market (g=6 percent), where landlord effi- is 12 percent. All other parameters are the ciencies are high (e=20 percent), real estate same as in the first case. In this case, the multiples are low relative to core business ownership of corporate real estate generates multiples, and core returns are high (r=14 a higher present value profit equal to 93 percent). In this instance, the arbitrage percent of the value of the real estate. That associated with shifting capital from real is, owning $100 million of real estate gen- estate to core operations, combined with erates higher present value profits of $93 reduced rents attributable to landlord effi- million. Note that it is difficult to envision ciencies, create a 41 percent present value

8 2 ZELL/LURIEREALESTATECENTER Figure 1: Simulation results

0.41

0.24 0.20 Low core return, high real estate return and idiosyncratic asset

Delta Typical High core return and competitive real estate market (0.20)

(0.60)

(1.00) (0.93) profit gain associated with renting real less competitive property markets, compa- estate. That is, $100 million of corporate nies with low rates of return in their core real estate ownership destroys $41 million business will gain by owning their real in corporate value. estate, particularly if the rental market is very inefficient. Our model can be easily applied to CONCLUSION every property to determine if the firm should own or lease the property. A critical The model demonstrates that there is no insight is that shifting dollars from EBIT- single answer as to whether a company DA to rent can enhance corporate value, as should own or lease its real estate. Instead, the capital is allocated to higher return it depends upon the nature of the firm, the core businesses, generating greater bang on nature of the real estate market, the type of the firm’s limited capital, by freeing capital the real estate, and taxes. But the model from relatively low-yielding real estate to demonstrates that high-multiple firms high-yielding corporate operations. This with high core rates of return, particularly decision also allows corporate manage- if they are looking for real estate that is ment to focus its energies on its core com- readily available in a competitive real estate petencies, which generally both lowers risk environment, should lease. The model also and adds value. suggests that for idiosyncratic properties in

REVIEW 8 3