REAL ESTATE DEVELOPMENT DECISIONS UNDER THE RISK AND UNCERTAINTY Raimonda Martinkutė, Aleksandras V. Rutkauskas Vilnius Gediminas technical university, Lithuania Produced real estate forms almost the half of Lithuanian national worth, and tangible asset (produced and non-produced real estate) forms almost 90% from the whole national worth. It is obvious, that management of tangible asset is the basis of country’s economy management, and economy development strategy must be related with strategies of real estate development. In addition, modern risk management theory insists that successful risk management may increase the effectiveness of country’s financial and other resources by 15-20%. So the real estate development strategy, supported by economical risk management mechanism, should become the main factor of stabile economical growth of country. In this article development possibilities of the real estate investment strategy as the main real estate management mean under risk and uncertainty will be investigated. Possibilities of risk evaluation in the processes of investment strategy preparation and realisation are revealed through analysis of concrete situations. 1. About the necessity to deal with uncertainty John Kenneth Galbraith writing in the 1970s called the latter part of the twentieth century The Age of Uncertainty (Galbraith, 1977). His theme was to “contrast the great certainties in economic thought in the last century with the great uncertainty that problems are faced in our time”. Once this uncertainty is accepted as being, as it were, in the nature of things, it is easily seen that those who enter into ventures that depend upon future outcomes have to come to terms with it. One such group of people are those who control business ventures. Consequently it is not surprising to find that it is in this area of activity where attempts have been made to develop a theory of decision-making that copes with uncertainty. Decision analysis has now become well established part of the curricula of business schools and the like, although it is fair to say that to date the basis of that theory has tended to support a mostly rational/quantitative methodology. These ideas are only beginning to gain more general acceptance, particularly with the development of techniques such as scenario analysis and more recently and still controversially, fuzzy analysis. At first sight it might be supposed that this body of theory and practice would have been taken up enthusiastically by those concerned with the development of land and buildings. After all, uncertainty lies at the root of the process of property development, which is essentially concerned, 1 with the manufacture of a product in anticipation of an unknown future demand. Indeed, if it were not for the constraint upon supply imposed by the system of town planning, it would rank as one of the most speculative of activities, involving as it does the investment of relatively large amounts of capital into a product that is fixed both in time and space. Yet the property development industry has largely ignored the methods of formal decision analysis adopted extensively in other industries (see for example Marshall and Kennedy, 1992). 2. Understanding the process of property development Even when related specifically to property, “the development process” means different things to different people. To some it is simply the construction of buildings, a physical process of production. To others it is essentially a part of a social and political process, involving the distribution and control of resources. Although there is no model of the development process that can be applied universally, for the purpose of our study, which is principally the investigation of uncertainty and risk, we can divide the process into three parts as follows: 1. Acquisition 2. Production 3. Disposal The first part of the process comprises the acquisition of the land upon which the development is to take place and the acquisition of the appropriate planning permission. The second part comprises the construction of the building or buildings and the third part comprises their disposal both for occupation and investment. As the process takes place, the developer's knowledge of the likely outcome increases but at the same time, the room for manoeuvre decreases. Thus, while at the start of the process developers have maximum uncertainty and manoeuvrability, at the end they know all but can do nothing to change their product, which has been manufactured on an essentially once and for all basis. The process is especially susceptible to risk and uncertainty because, once started, it is relatively fixed in time and place and because it aims at a very narrow consumer market. During the first part of the process, the period of acquisition, the uncertainties are of three main types: the physical characteristics of the land, the characteristics of tenure, including restrictive covenants and casements both in favour of and against the land, and the nature and extent of use that the local planning authority will permit. Most developers will attempt to identify and determine all these factors before committing themselves to the purchase of the land. Land acquisition cost is 2 often the first major commitment of capital and as it comes at the start of the development process it is then like a snowball, accumulating interest throughout the remainder of the development. During the second part of the process, the period of production, the main element of uncertainty is the cost of construction. This factor, which represents the second major capital commitment, is substantially determined at the start of the building contract with the builder. However, nearly in all cases there will be some element of fluctuation allowed in the contract and, in any event, the phasing of the construction and the length of the building period can never be fully determined at the outset. The third part of the process, the period of disposal, can be seen as comprising both the disposal of the building to one or more occupiers and its disposal as an investment. The process that transfixes all the components of real estate development is investing. More over, effective investment strategy is the main instrument for the successive real estate development. And finally, investing is the most widely exposed to risk development activity. Further we discus some possibilities how to prepare and realise investment strategy under the risk. 3. Preparation and realisation of investment strategy under the risk 3.1. The real estate investment process as jumping-of investment strategy preparation In management engineering the category of strategy is used for both to impress the importance of analysed purposes, criterions, assumptions and means and to designate the function, that depends on current information and gains the means from set of possible alternatives. So if we treat strategy as the whole of rules for achievement of system’s strategic purposes using strategic means, then the real estate investment strategy should be perceived as the achievement of strategic investment purpose. This purpose consists of the maximisation of investment value, choosing between alternative investment sources and utility maximisation. The real estate investment process (Figure 1) is the main mean for strategy preparation and should be analysed using probabilistically arranged information. 3 Figure 1. Real estate investment process Step 1: Identify investor’s objectives, goals, and constraints Risk-return preferences Wealth Step 2: Analyze investment climate and market conditions Market analysis Legal environment Sociopolitical analysis Step 3: Develop financial analysis Operating decisions Financing decisions Reversion decisions Income taxation Tax planing Wealth taxation Step 4: Apply decision – making criteria Rules of thumb techniques Discounted cash Traditional flow techniques appraisal methods Step 5: Investment decision 4 Data, characterising the realisation of chosen real estate development strategy is presented in Table 1: the price of investment, indicators of exploitation of investment, sources and conditions of finance, selling price of investment, indicators of dependence and interaction between factors and indicators of investment efficiency. In many cases the presumption is made that data, presented in Table 1, is sufficient for description of strategy. Indeed, this data is the result of consistent and comprehensive research of external and internal environments. A lot of information about various factor’s impact on investment process and about results of interaction between investment purposes, possibilities, means and consequences after choosing the strategy is accumulated in this research. Really, a traditional point of view into investment process as deterministic doesn’t create presumptions for taking into account some strategic factors of investment process such as risk and uncertainty. The purpose of this article is to take into consideration the impact of uncertainty of separate parameters, dependencies and restrictions on the results of investment strategy realisation. As we can see from Table 1, the data in Columns 2-4 is arranged differently than the data in the first column. The data, presented in Columns 2-4, which can’t be described in deterministic manner (the rent, the change of operating costs, the interest rate of first mortgage, selling price of investment, discount rate of expected incomes) is defined as probability distributions. Such way of definition of possible meanings reflects the fact, that investment process is perceived as probabilistic process. This fact must be taken into account. 3.2. The real estate investment strategy, based on probabilistic standpoint at investment process In the traditional description of strategy, where investment process is understood as deterministic process, there are three possibilities: pessimistic, realistic and optimistic. After making the assumption that each of these three possibilities may occur with determined probabilities, it is affirmed that probabilistic case of strategy realisation is chosen. But such a point of view has few contradictions. The first contradiction is the excluding of future possibilities. Of course, the perception and choosing of pessimistic or optimistic possibilities do not excite any problems or doubts. Pessimistic possibility has the lowest value from chosen level of confidence interval if the factor has possessive impact on indicators of investment’s efficiency and the biggest value – if the factor has negative impact on the investment. In the case of optimistic possibility the biggest value is chosen if the factor has positive impact on efficiency of investment. Meanwhile, the selection and understanding of realistic possibility is problematical. What is the realistic value of the factor? Practically, it is the average of pessimistic and optimistic values. But it is true only in the case of symmetrical distribution. 5 However, main reproaches may be declared against presumption that pessimistic, realistic and optimistic values of economical indicators of investment are obtained when all the factors have pessimistic, realistic and optimistic variants accordingly. Of course, the probability of such coincidence is zero. Indirect evidence of such proposition may be the circumstance, that when pessimistic and optimistic values of separate investments’ factors are comparatively compatible, optimistic and pessimistic values of received indicators of investment efficiency can’t be treated as separate possibilities of the same investment. So the natural way of determination of possibilities of investment indicators should refer to assumption about analysis of interaction between factors, taking into account their possible interdependency. Of course, such analysis of investment possibilities demands purposeful preparation of information and absolutely new technique for decision making. 3.3. The comparison of strategies and analysis of results In this section of article we’ll try to compare traditional variant of strategy (first column) with probabilistic variants (second-fourth columns). Table 1. Triplex example with a first and second mortgage Variants The table of 1 2 3 4 exogenously Under point Under statistical Under line Under log-normal forecasted data forecasting independence of dependence of distribution of factors factors factors’ possibilities 95 000 LTL 95 000 LTL 95 000 LTL 95 000 LTL Rent 350 N(350;3,5) N(350;3,5) LN(a=350; σ=3,5) Annual increase of 0,05 0,05 0,05 0,05 Number of units 3 3 3 3 Coefficient of 0,3 0,3 0,3 0,3 5% N(0,05;0,001) N(0,05;0,001) LN(a=0,05; The price of triplex dwelling rent operating expenses Increase of operating σ=0,001) expenses Holding period 27,5 years 27,5 years 27,5 years 27,5 years Depreciation 0,85 0,85 0,85 0,85 First mortgage 70000 LTL 70000 LTL 70000 LTL 70000 LTL First mortgage 12% N(0,12;0,01) N(0,12;0,01) LN(a=0,12; Financing: interest rate σ=0,01) 6 First mortgage 20 years 20 years 20 years 20 years Price of mortgage 0,04 of sum 0,04 of sum 0,04 of sum 0,04 of sum Financing expenses 2800 LTL 2800 LTL 2800 LTL 2800 LTL Refusal costs 0,06 of sum 0,06 of sum 0,06 of sum 0,06 of sum Second mortgage 10000 LTL 10000 LTL 10000 LTL 10000 LTL Payments from 1500 LTL annually 1500 LTL annually 1500 LTL annually 15 00 LTL annually after 3 years after 3 years after 3 years after 3 years Tax rate 0,28 0,28 0,28 0,28 Expected selling 140000 N(140000; 700) N(140000; 700) LN(a=140000; maturity of first mortgage second mortgage Repayment from second mortgage σ=700) price Selling expenses 0,08 0,08 0,08 0,08 Income discount rate 0,11 N(0,11;0,001) N(0,11;0,001) LN(a=0,11; σ=0,001) Matrix of factors – Cii = 1 Cii = 1 Cii = 1 Cij = 0, i≠j Cij = 0,3, i≠j Cij = 0, i≠j 5 years 5 years 5 years 5 years 2160,3 N(2109; 2189) N(2132; 2374) LN(a=2501; interdependency Property holding period The table of endogenously forecasted data Net present value σ=2480) (NPV) Internal rate of return (IRR) 13,08% N(0,1297; 0,0180) N(0,1298; 0,0174) LN(a=0,1313; σ=0,0207) Lets examine four cases, shown in Table 1: 1. Exogenously forecasted indicators of strategy are defined in deterministic manner. 2. Five exogenous variables (rents, rate of growth of operating costs, interest rate for the first mortgage, selling price of investment, discount rate of income) are defined using their possible probability distributions. In this case distributions are normal. Other indicators are the same as in the first column. 3. Five variables, mentioned in the second case are not only normally distributed, but also statistically interdependent (Cii= 1, Cij= 0,3, i≠j). 7 4. The same five variables are not statistically interdependent, but they have log-normal probabilistic distribution. Geometrical view of probability distribution for main endogenous characteristics (NPV and IRR) possible meanings is shown in Figure 2. Data, presented in the first column of Table 1, can serve only as illustration of investment process, because we can form the view of real possibilities only after making some additional presumptions. The data presented in Columns 2-4 may be used as the realisation of these possibilities. The data in the second column and Figure 2a shows the possibility distribution of NPV and IRR. From the third column of Table 1 and Figure 2b we can make the conclusion that it is very important to take into account statistical dependence of separate factors. This dependence may become the object of management because of the purpose of economical benefit. If the difference between the data, presented in Figure 2b and in Figure 2a is the function of controlled processes, so the difference between data in Figure 2c and Figure 2a is predetermined by the formation of different forms probability distributions. The comparison shows that this difference is quite a big so any search of its management possibilities is necessary. Now lets analyse investor’s utility function, which has the same shape as T. F. Knight used it in 1921. In the case of our investment its concretisation should be such as depicted in the figure 3b. Two utility functions u1(x) and u2(x) are shown in this figure and we can see that the second from them - u2(x) has higher reaction at possible decrease of cash flow under expectation or higher reaction on risk. However, speaking about possibility to define utility as point estimated curve is very conventional. Really investor can describe his utility as casual function. Stochastic image of u1(x) is presented in figure 3c. In this figure u1(x) by itself serves as the curve of average of values. Three utility values U1, U2, U3, generated from utility function u1(x), are shown in figure 2b. Their quantity depends on probability distributions p1(x), p2(x) and p3(x) of net present values of cash flows, taken from figure 2. EU value (fig. 2c) would be the same with EU1, EU2 and EU3 if the utility were evaluated according to average line. But average line is not suitable for representation of utility possibilities. So more adequate criterions are necessary. 8 Figure 2. Separate variants of real estate development strategy NP Vnorm al;distribution:Norm al IRRnorm al;distribution:Norm al M=2109;S D=2189 M=0,1297;S D=0,0180 Chi-S quare:6.130764,df=8,p=.6325884(dfadjusted) 130 130 120 120 110 110 100 100 90 90 80 80 70 70 60 60 Noofobs Noofobs Chi-S quare:13.57508,df=9,p=.1383187(dfadjusted) 50 50 40 40 30 30 20 20 10 10 0 -5000 -3000 -1000 1000 3000 5000 7000 9000 0 0,07 E xpected Category(upperlim its) 0,09 0,11 0,13 0,15 0,17 0,19 E xpected Category(upperlim its) NP Vunderdependenceoffactors;distribution:Norm al IRRunderdependenceoffactors;distribution:Norm al M=2132;S D=2374 M=0,1298;S D=0,0174 Chi-S quare:4.913710,df=6,p=.5549363(dfadjusted) Chi-S quare:3.265830,df=7,p=.8593616(dfadjusted) 160 140 130 140 120 110 120 100 90 100 80 70 Noofobs Noofobs 80 60 60 50 40 40 30 20 20 10 0 -8000 -6100 -4200 -2300 -400 1500 3400 5300 7200 9100 11000 E xpected 0 0,070,080,090,100,110,120,130,140,150,160,170,180,190,20 Category(upperlim its) E xpected Category(upperlim its) NP Vlognorm al;distribution:Norm al IRRlognorm al;distribution:Lognorm al M=2501;S D=2480 M=0,1313;S D=0,0207 Chi-S quare:18.32544,df=8,p=.0189384(dfadjusted) Chi-S quare:8.167499,df=8,p=.4173083(dfadjusted) 130 160 120 140 110 100 120 90 80 100 70 80 Noofobs Noofobs 60 50 40 60 30 40 20 10 20 0 -6000 -2400 -4200 1200 -600 4800 3000 Category(upperlim its) 8400 6600 12000 10200 E xpected 0 0,070 0,089 0,108 0,127 0,146 0,165 0,184 0,203 0,222 E xpected Category(upperlim its) 9 3a NP Vnorm al;distribution:Norm al NPVunderdependenceoffactors;distribution:Norm al M=2109;S D=2189 NP Vlognorm al;distribution:Norm al M=2132;SD=2374 Chi-S quare:13.57508,df=9,p=.1383187(dfadjusted) M=2501;S D=2480 Chi-Square:4.913710,df=6,p=.5549363(dfadjusted) 130 Chi-S quare:18.32544,df=8,p=.0189384(dfadjusted) 160 130 120 120 140 110 100 110 100 120 90 90 70 70 80 50 60 Noofobs Noofobs 60 60 40 20 10 10 -3000 -1000 1000 3000 5000 7000 E xpected 9000 0 -8000 -6100 -4200 -2300 -400 1500 3400 5300 7200 910011000 Category(upperlim its) 40 20 20 0 -5000 50 30 40 30 0 -6000 Expected -2400 -4200 Category(upperlim its) 1200 -600 4800 3000 8400 6600 12000 10200 E xpected Category(upperlim its) 3b 8000 6000 4000 2000 y(u) Noofobs 80 100 80 u2(x) 0 -4000 -2000 -2000 0 2000 4000 6000 8000 u1(x) bisector -4000 -6000 -8000 -10000 x (cfo) 3c 8000 6000 4000 2000 0 -4000 -2000 -2000 0 2000 4000 6000 8000 -4000 -6000 -8000 -10000 Figure 3. The scheme of criterion selection on the base of utility functions 10 Conclusions: 1. Including risk and uncertainty into processes of strategy preparation and realisation is the most important mean for improvement real estate investment strategy. 2. The preparation and realisation of investment strategies, where investment processes is understood like probabilistic, help to take into account possibilities of separate subject to choose an investment, to realise investment strategy and to evaluate the impact of risk on investment efficiency. 3. Adequate risk management strategy could considerable improve efficiency of resources involved e.g. financial resources. 4. Invisible factors, such as statistical interdependence between indicators of investment efficiency and distribution differences between these indicators has an appreciable impact on the whole efficiency of investment. Literature 1. Sirmans C.F., Austin J. Jaffe. The complete real estate investment handbook. A professional investment strategy. Fourth edition. USA: Prentice Hall Press. 1985. 2. Rutkauskas A.V., Rutkauskas V. Adekvataus pelno galimybių nevienareikšmiškumui investicijų portfelio sudarymas // Ekonomika, 2000, № 52, psl. 62-83. 3. Galbraith J.K. The age of uncertainty. London: BBC and Andre Deutsh Ltd. 1977. 4. Marshall P. and Kennedy C. Development valuation techniques // Journal of property valuation and investment, 1992, № 11(1), p. 57-66. 5. Gerald R. Brown, George A. Matysiak. Real estate investment. A capital market approach. England: Financial Times Prentice Hall, 2000. 6. Haugen R. A. Modern investment theory, 3rd edn. Englewwod Cliffs, NJ: Prentice Hall, 1993. 7. Liu C., Hartzell D., Grisson T., Grieg W. The composition of the market portfolio and real estate investment performance. AREUEA Journal 18, p.49-75, 1990. 8. Lusht K. M. Real estate valuation. Chicago: Irwin. 1997. 9. Spaudling D. Measuring investment performance. New York: McGraw-Hill. 1997. 10. Knight F. H. Risk, Uncertainty and Profit, Houghton Mifflin. Boston and New York, 1921. 11