Abstract
In the following chapter, the actual valuation of the retail banking startup in Poland is performed. It will be shown that the use of the static discounted cash flow framework alone can lead to wrong decisions and cannot explain the investment pattern observed in the Polish retail banking market. Using real option valuation, the investment decisions can be rationalized by including the value of the options embedded in the project. Still, the investment timing decision appears premature and suboptimal. Merging real option valuation and game theory, the conditions are explored under which the observed investment race can be explained.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
This information has been published primarily in broker reports and is the basis for the economic analysis.
Merrill Lynch (1998a), p. 3; Merrill Lynch (1998e), p. 7
Schroders (1998a), p. 7; Halaba (1998), p. 3
Dresdner Kleinwort Benson (1998), p. 26
For a definition of the term and its characteristics see Walter (1997), pp. 24–26.
Copeland, Koller, Murrin (1994), pp. 286–292 recommend an explicit forecasting period of around 10 years, but state that a longer rather than a shorter forecast period is preferable.
See, for instance, EIU Tradewire (1997). The longest-term explicit forecast only reaches until 2002. This figure (5% p.a.) has also been used for the subsequent periods in the model. Using international long-term inflation rates as a reference, there does not seem to be the need to adjust the 5% figure. Over the 1990 to 1997 period the average inflation rate in the EU amounted to 3.6%. Looking at a basket of poorer EU countries (Greece, Ireland, Portugal, Spain) the average increases to 6.9%. Taking newly industrialized Asian countries as additional benchmark, their average inflation rate for the same period was 5.2%. Increasing the observation period by 10 years to cover 1980 to 1997, average inflation increases to 5.5% for the EU, 10.7% for the poorer EU countries and 6.0% for the newly industrialized Asian countries (IMF (1998), pp. 156–157 ).
This is the approach followed by all three banks under consideration (Merrill Lynch (1998a), p. 3; Merrill Lynch (1998e), p. 7; Schroders (1998a), p. 7).
The target number of customers and number of ATMs per branch follow the BBG blueprint (Merrill Lynch (1998e), p. 7). For BSK, the per branch customer number is also 2,500 while it is 2,000 for BH (Dresdner Kleinwort Benson (1998), p. 30). The importance of a highly automated branch layout, including ATMs, is also strongly emphasized for BSK and BH (Merrill Lynch (1998a), p. 3; Schroders (1998a), p. 7 ).
For all three startups, staffing levels are assumed to fall in the 5–7 FTE range (Merrill Lynch (1998a), p. 3; Merrill Lynch (1998e), p. 7; Schroders (1998a), p. 7; Thomson Bankwatch (1997), p. 4.)
Significantly, the BSK branches are named STAREO (standard retail outlet) with their design being standardized to a very high degree (Banking on A5400 (1997), p. 7; Dresdner Kleinwort Benson (1998), p. 31).
With this setup operational flexibility can be maintained in the future and the economics can be kept as free as possible from influences of the real estate market.
BSK quotes an investment budget (IT and construction) of PLN 800,000–1,200,000 per branch. These costs are based on a highly standardized design and include one or more ATMs per site (Dresdner Kleinwort Benson (1998), pp. 30–31). The BBG investment target per branch is USD 260,000 or PLN 910,000. This figure, however, seems very low, particularly in comparison to BSK. Considering that the three ATMs alone would cost approximately USD 100,000, only USD 160,000 would remain for other branch hardware components, security installations and construction. In their assessment, Merrill Lynch also assume that actual IT cost will exceed the budgeted figures and consequently the branch budget will go up (Merrill Lynch (1998e), p. 15 ).
The rollout module used for the simulation follows the BBG setup. They target a network size of 200 branches by the end of 2000, with the first 40 branches to go live by September 1998 and 100, respectively 60 to follow in 1999 and 2000 (Merrill Lynch (1998e), pp. 7, 15). The target number quoted for BSK varies between a total of 220 (Halaba (1998), p. 3) and 400–500 (Merrill Lynch (1998a), p. 3; Dresdner Kleinwort Benson (1998), p. 30). BH on the other hand is targeting only 60100 branches (Halaba (1998), p. 3; Dresdner Kleinwort Benson (1998), p. 30). Their value proposition, however, seems to be based more strongly on electronic banking as discussed earlier.
Own calculations based on annual reports. The banks used as examples are: the Argentinean subsidiary of Brazil’s Banc Itail (Banc Itail) launched in 1995, Hypo Servicebank of Germany (HSB) which started in 1991, Banco Commercial Portugues of Portugal (BCP) launched in 1989 and their separate mass retail network (Nova Rede) which opened its first branches also in 1989.
In 1995, Poland had, for instance, only 14.8 land phone lines per 100 inhabitants. This figure is projected to grow to 25.6 by the year 2000 (Palmer (1996), p. 11 ).
Efficient transaction processing is one of the focal points of the banks in their new retail efforts (Merrill Lynch (1998a), p. 3; Dresdner Kleinwort Benson (1998), p. 26). Centralization of transaction processing allows for significant efficiency gains (Bream (1998), p. 158 ).
The Polish payment system KIR consists of a paper-based (SYBIR) and an electronic (ELIKSIR) part.
This figure is based on international experience. It can, however, be backed up and put into perspective by benchmarking it to comparable figures from Germany. The employment figures quoted so far for branches and central operations imply a total of 1,637 employees, once the full network size of 200 branches is reached. This figure can also be found in table 4.5. Put differently, it means that 3.27 employees are required per 1,000 customers. It is important to bear in mind that this figure reflects a modern and advanced multi-channel retail banking format with highly automated transactions, as presented in section 2.2. From Walter (1997), p. 202, who discussed the retail banking system in Germany, it can be inferred that the average German multi-channel retail bank had a ratio of 4.67 employees per 1,000 customers. The figure for the Polish model bank is thus at 70% of German levels. This gap can be explained by two facts. In contrast to the Polish bank, German retail banks in general offer a broader product line, including retail brokerage and mortgages, which are advice intensive businesses. Additionally, in a 1997 study, McKinsey estimated the improvement potential for labor productivity in German retail banking at about 30% vs. international best practice (McKinsey Global Institute (1997), p. 17). This evidence makes the projected figure for the Polish retail banking startups look plausible.
Due to the hyperinflationary past, almost a quarter of retail deposits are still held in foreign currency, with USD and DM being the most popular currencies (NBP (1997), pp. 86–87).
The need to offer attractive pricing when entering the Polish retail banking market is also stressed by Dresdner Kleinwort Benson (1998), p. 26.
NBP (1997), product leaflets
For a detailed overview of the MOR approach see Schierenbeck (1994), pp. 69–132. The approach is sometimes also called Funds Transfer Pricing (Uyemura, Kantor, Pettit (1996), p. 104 ).
Ministerstwo Finansów (1997), p. 56
The underlying formula for deposits is: NIM = MOR“(1-MR)-rate (Buschgen (1993), p. 642).
Average from January to June 1997. The product prices are calculated every month as the unweighted average of the BPH, BSK and WBK rates and matched with the respective opportunity rate in that month, considering the minimum reserve requirements in force at that time.
The position “excess marketable securities” from the balance sheet. 41° Citicorp (1998), p. 6
This is modeled as linear decrease, starting in the year 2000.
This is in line with the BBG forecasts (Merrill Lynch (1996), p. 16) and is also used as assumption by Dresdner Kleinwort Benson (1998), p. 10 in their model.
The volume split along maturities follows the market average with 3 months at 60%, 6 months at 23% and 12 months at 17% (NBP (1997), p. 15).
The split between PLN and USD savings accounts follows the market average (NBP (1997), p. 15).
The average customer is based on a 20/40/40 target customer mix.
Salomon (1997), p. 9; PlanEcon (1997), p. 123; McQuaid (1997), p. 3.
Based on Hempel, Simonson (1991), p. 40.
Buschgen (1993), pp. 635–638
Own calculations based on broker reports, Bank/Prawo i Gospodarka (1998). As part-time work is basically non-existent in Polish banking at the moment, costs per employee closely approximate costs per FTE. The higher costs of BH and BRE, and to some extent BBG, can be explained by a stronger wholesale focus of these banks.
For a similar approach using an average of 12 m` for German banks, see Walter (1997), p. 134.
At 1999 cost level. The figure consists of personnel cost (39%), occupancy cost (29%), as well as other costs such as telecommunications (32%).
Merrill Lynch (1998e), p. 15
The depreciation rate for IT equipment is five years in Poland (Price Waterhouse (1997). A linear five year depreciation is therefore assumed for all investments in the model.
The year 2 numbers in table 4.6 support a drop in acquisition costs to 60% for Bank 24, 40% for Banc Itaú and even 10% for First Direct. While the latter is probably an exception due to its tremendous success right from the start, the 40% figure is in line with Walter (1997), p. 179, who looks at the development over time of new customer acquisition costs for German direct banks.
In the US, the annual customer attrition rate amounts to 15%, which can be split into 6% voluntary and 9% involuntary (death, moving) attrition (Huber, Lane, Pofcher (1998), p. 149). Given the lower mobility and sophistication of consumers in Poland, the long-term attrition rate should probably move around 6% p.a., and represent mainly involuntary attrition.
This is in line with Walter (1997), p. 179, who quotes first year customer acquisition cost of German telephone banks at DM 1,200 (PLN 2,200).
This means that they are almost 50% lower than the costs of the project phase for a German direct bank, which have been estimated at DM 18 million or PLN 34 million (Walter (1997), p. 166). Given the fact that the majority of these costs are personnel and service related costs, it is plausible, considering the much lower salary level in Poland.
According to broker reports, the banks have used external consultants in the process of designing their retail bank startup (Merrill Lynch (1998b), p. 3; Merrill Lynch (1998e), p. 7).
Based on Hempel, Simonson (1991), pp. 34–35
As all three banking initiatives are launched by an existing parent bank, the increased equity requirements in the first two years for start-up banks do not apply.
Bank/Prawo i Gospodarka (1998)
The existence of an internal capital market for bank holdings has been corroborated by Houston, James (1998), pp. 74–77. Otherwise, loan growth would be slower, being limited by the bank’s own growth in retained earnings.
The 8% are based on the assumption that of a total 12 million households in Poland, only 51% are currently banked. Considering that by 2003, a higher share of households should have a bank relation, the actual market share at that time is likely to be lower than 8%.
For the BSK retail banking launch, branches are assumed to break even between year 3 and 4 (Merrill Lynch (1998a), p. 3); for BH, the estimated time is 18 months (Schroders (1998a), p 7). The difference could be explained by the different banking formats used as well as by different assumptions concerning buy or lease.
Own calculations based on broker reports and Bank/Prawo i Gospodarka (1998).
Depending on the angle adopted by the authors, there is an overlapping classification of capital budgeting and valuation. Copeland, Koller, Murrin (1994), pp. 59–60 classify, for example, the capital budgeting problem of selecting from among several investment alternatives as valuation, while Seitz (1990) treats the valuation of acquisition targets in a capital budgeting context.
As cash flows are projected based on optimal management from the commissioning firm’s point of view, a controlling interest in the project is implicitly assumed.
See, for instance, Mercer (1992).
Mercer (1992), pp. 195–214; Smith (1988), pp. 92–101. The capitalization of earnings approach capitalizes an estimate of accounting earnings that the business can be expected to generate from normal operations (“normalized earnings level”), using a capitalization rate based on the anticipated growth in earnings and the associated risk. This approach does not reflect actual cash flows and can only he applied meaningfully to a project with an operating history.
Miller (1995), pp. 185–189, Mercer (1992), pp. 252–258. There are two variations of the market approach. It is possible to value the venture based on prices paid in recent acquisitions of similar banks, or using publicly traded companies as comparables. In both cases adjustments are necessary to account for differences between banks.
Kaplan, Ruback (1995), pp. 1076–1081. The general consensus in the valuation literature also seems to favor the DCF approach (Kasper (1997), Miller (1995), Damodaran (1994) and Copeland, Koller, Murrin (1994)). Only Mercer (1992) strongly advocates the market approach and states that the DCF method, though theoretically sound, should be used carefully and only as a supplement but not as a substitute.
Lessard (1996); Godfrey, Espinosa (1996), pp. 80–81. It has been recommended specifically in an Eastern European context by Birch (1993), p. 18; Laurent (1993), p. 96; Ferris, Joshi, Makhija (1995), pp. 54–57.
Decree of the Minister of Privatization from February 20, 1990, as quoted in Jermakowicz, Jermakowicz (1993), p. 42.
Copeland, Koller, Murrin (1994), pp. 478–479
Copeland, Koller, Murrin (1994), p. 479. Though this claim is not always supported (see, for example, Kirsch, Krause (1996)), diverging results can be attributed to an inconsistent application of the methodology (Kaden, Wagner, Weber, Wenzel (1997); Volkart (1997), pp. 108–113). The mere existence of such a discussion, however, shows that correct application is anything but trivial.
Copeland, Koller, Murrin (1994), pp. 131–134; Damodaran (1994), p. 149. This is particularly true of Anglo-Saxon countries, while the equity method is also widely used in Germany for industrial companies. ( Volkart (1997), pp. 106–107 ).
See, for instance, Froot, Stein (1998).
Based on Parrino (1994); Copeland, Koller, Murrin (1994); Damodaran (1994).
Based on Parrino (1994); Copeland, Koller, Murrin (1994); Damodaran (1994). 476 Miller (1995), p. 30
Based on Copeland, Koller, Murrin (1994).
Potential dividends is the total possible payout to equity holders after adjustment for required increases in equity.
Martin, Cox, MacMinn (1988), p. 304; Kasper (1997), pp. 145–146
Lessard (1996), p. 61; Godfrey, Espinosa (1996), p. 84
Rynek Gieldowy (1997), p. 29
This can be derived from the fact that the standard CAPM is a single-period model that derives the required return assuming that the risk-free rate and the market premium, defined over the same period, are constant. The actual duration of that period is unspecified by the model but needs to cover the time span of the analysis.
The transformation into a linear relation follows Bamberg, Baur (1993), pp. 46–48.
R’ equals 86% for (a) and 72% for (b). F-test as well as t-tests are significant at the 1% level for both types.
Martin, Cox, MacMinn (1988), p. 305
The arithmetic mean is typically used in the academic literature (see, for instance, Stutz (1995); Patterson (1995); Berger, Ofek, Swary (1996); Gilson, Hotchkiss, Ruback (1998)). This is in contrast to both Copeland, Koller, Murrin (1994), pp. 260–264 and Damodaran (1994), p. 22 who recommend the geometric average. Indro, Lee (1997), p. 89 recommend a weighted average of the two. With Kaplan, Ruback (1995), p. 1083, there is some empirical evidence in support of the long-term arithmetic mean.
Ibbotson Associates (1996)
There appears to be little dissent on the use of the long-term government bond rate as benchmark and of the longest time series available (Patterson (1995), pp. 108–110; Copeland, Koller, Murrin (1994), p. 260 ). Using the long-term rate allows for consistency with the discount rate used within the CAPM framework. Using the longest time series available improves the quality of the estimate since it helps to eliminate short-term anomalies.
Ibbotson Associates (1996), p. 117 state that market risk premia seem randomly distributed as their annual serial correlation is equal to -0.01, which is not significant. The importance of this statement is clarified, when looking at the typical argument set forth by the proponents of the geometric average. The example follows Copeland, Koller, Murrin (1994), p. 261. They consider the purchase of a non-dividend paying stock at 50. After year one, the stock is at 100, after year two, back to 50 again. Based on this, they conclude that the geometric average of the returns, which is 0%, is closer to reality than the arithmetic average, which is 25%. Within the boundaries of their example, this conclusion is correct, as stock returns and thus serially correlated values are analyzed. The true mean of a distribution of serially correlated values is the geometric average. However, as the finding of Ibbotson Associates shows, market risk premia, which are the differences between the return on stocks and long-term bonds are not serially correlated. The true mean of a distribution of independent values is the arithmetic average.
Chan, Karolyi, Stulz (1992), p. 139 state that “there is now substantial evidence that stock markets are reasonable well integrated.” For an overview of the factors driving the trend towards further integration of capital markets see, for example, Bryan, Farell (1996). For an opposing view see Erb, Harvey, Viskanta (1995), p. 80.
In 1995, the CEE countries did approximately two-thirds of their foreign trade with developed market economies. Looking specifically at Poland, Germany was by far the largest trade partner with a share of around 30% (United Nations (1997), pp. 128–139 ).
Ali, Mirza (1996), p. 44; Vinton (1997), pp. 1–2
Patterson (1995), pp. 108–110; Copeland, Koller, Murrin (1994), p. 260
The US market risk premium (arithmetic average) for the 1970–1995 period would, for instance, amount to only 2.8% in contrast to the 7.0% for the 1926–1995 period (Ibbotson Associates (1996)).
Harvey (1991), p. 139. Sinquefield (1996) finds no statistically significant difference between estimates of the market risk premium for the US and a world portfolio for the 1975–1994 period. The result supports the claim that the US market risk premium can act as a proxy for the world’s.
The integration of capital markets has two opposing effects on the cost of capital. Risk premia should decrease as investment barriers are removed and the risks can be spread over more investors. As a secondary effect, a higher degree of integration of the national economies will lower the benefits from global diversification, as the correlation between national markets increases. Setting both effects off against each other approximately results in a 10% discount from the US figure (Stulz (1995), pp. 37–38 ).
Comparing this market risk premium with other recommendations in the literature, it appears to be rather on the low side. Damodaran (1994), p. 23 recommends, for instance, an ad-hoc premium of 8.5% for CEE countries vs. 5.5% for the US. Godfrey, Espinosa (1996), p. 89 quote a 27.3% USD-discount rate for an offshore project in Poland, compared to 11.5% for a US project and Ferris, Joshi, Makhija (1995), p. 57 recommend grossing up the discount rate for a comparable Western company by 10 to 15 percentage points. The figure is, however, in line with a 4–4.5% risk premium covering the term ß’MRP used by Merrill Lynch (1998e), p. 18, while Dresdner Kleinwort Benson (1998), p. 9 use 7% for Poland.
Fuller, Kerr (1981), p. 1002 provide empirical support for the validity of the “pure play” approach.
Copeland, Koller, Murrin (1994), p. 264; Klemkosky, Martin (1975), p. 1128
Consistency requires that both ß and the market risk premium are measured against the same index.
BARRA predicted-betas measured against an international index are used (BARRA Beta sourcebook (1997))
BARRA (1997); own calculations
Moreover, Fuller, Kerr (1981), p. 1004 show that “pure play” ß-factors, which are not adjusted for leverage, give more accurate results than their re-levered equivalents.
The formula is adapted from Copeland, Koller, Murrin (1994), p. 277. The weighted average cost of capital has been replaced by the cost of equity (r,) to account for the fact that equity-based free cash flows are used in the analysis.
As discussed in section 4.2.2, the long-term real growth rate for retail deposits and loans is estimated at 3.5%. The figure is used here as proxy to describe real cash flow growth. It is close to Dresdner Kleinwort Benson (1998), p. 10, who assume a 4% long-term real earnings growth in their valuation model.
The 5% p.a. figure equals the longest-term forecast available (for the year 2002 ). Comparing it to international long-term inflation rates, there does not seem to be the need to further adjust the figure (see section 4.2.2 for more detail).
The target number of customers will not be reached before mid-2003. Until then, there will be minor investments in the headquarters to cope with the increasing number of customers. However, by the end of 2000, all the core infrastructure will be in place and more than 95% of the total investment budget will have been spent.
By applying macroeconomic data to a microeconomic problem, i.e. the value of a certain bank, the main implicit assumption is that changes in the macro variable, i.e. the aggregate volume of deposits and loans, are exactly mirrored in the changes of the micro variable (i.e. retail deposits and loans at a certain bank). This is true under the condition that the changes in aggregate volume are equally distributed among customers and non-customers of the bank. Another condition is that market shares are constant over time unless explicitly modeled. Both conditions appear realistic. The Gini coefficients of the Polish income distribution are stable over time (Gorecki (1994), p. 36) and retail customers are loyal, leaving market shares virtually constant over time (McKinsey Global Institute (1997), exhibit 14 for evidence on the German retail banking market).
See particularly Davis (1998).
Except for specialized consumer finance companies or mortgage banks, retail banks will typically have an excess of deposits. Based on international banking statistics from OECD (1997), the 65% can be put in an international perspective, looking at the figures from savings and co-operative banks in Germany, Spain and the US, as for them, the retail business is typically more important than for commercial banks. The average 1995 ratio of non-bank loans to non-bank deposits ranges from 61% for Spanish co-operatives to almost 89% for German savings banks. The numbers are likely to be inflated, as these banks still do a sizable amount of corporate business, particularly for small and medium sized companies. Taking Polish banks as a reference, the numbers are lower, as the ratio for the nine NBP spin-offs amounts to 59.8% and to 45.3% for the co-operative banks.
As a measure of correlation, the Bravais-Pearson correlation coefficient has been used (Schulze (1990), pp. 128–130). It amounts to -0.0696.
Rynek Gieldowy III and íV/1997; own calculations
Citicorp (1998), p. 40. In the legal area changes typically take time and are also hard to predict.
The contribution to gross income in Western markets is proxied by using “Net fees and commissions” for savings and co-operative banks, as percent of “Gross income”, as stated in OECD (1997). Savings and co-operative banks are used, as for them, the retail business is typically more important than for commercial banks. Still the results are likely to be inflated, as “Net fees and commissions” also includes current account maintenance fees as well as some commissions from the wholesale business. The share of gross income ranges from 8.9% for Spanish co-operative banks in 1994 to 15.3% for US savings banks in 1995.
See Trigeorgis (1993b) for a discussion of real options and their interaction with financial flexibility.
Dresdner Kleinwort Benson (1998), p. 31
There are 42 cities in Poland with more than 100,000 inhabitants, covering a total of 11.6 million people (Dresdner Kleinwort Benson (1998), p. 24).
The term compound option can consequently be defined as an option that yields both an asset and another option, when exercised ( Benavides, Johnson (1998), p. 3 ).
Kulatilaka, Trigeorgis (1994), p. 794
See, for instance, Trigeorgis (1996), pp. 166–167.
Berger, Ofek, Swary (1996), p. 268 find an average value for the option to abandon of 12% for US industrial companies. The figure is, however, very volatile, depending on firm-specific factors.
See Thaler (1993); DeBondt, Thaler (1995) for surveys of this field.
DeBondt, Thaler (1985) and (1987); Jegadeesh, Titman (1993)
Roll (1986); Morck, Shleifer, Vishny (1990)
DeBondt, Thaler (1995), p. 391
DeBondt, Thaler (1995), p. 391
DeBondt, Thaler (1995), p. 391
Darity (1986), pp. 214–215
Lissaker (1991), p. 87
Lissaker (1991), pp. 96, 192
Sea As Spremann (1987), p. 9 notes, a formal contract is not a prerequisite for the existence of an agency relationship.
Spremann (1987), p. 9; Shavell (1979), p. 55
The literature typically focuses on mergers and acquisitions. Given the common growth theme, its results can also be applied to investments in new areas of business.
Business Eastern Europe (16.03.1998), p. 9
PER and P/BV ratios from Merrill Lynch, Dresdner Kleinwort Benson and Schroders for eight Polish banks over the 1996–1998 period are combined to derive average market valuation ratios. The assumption is that the specific retail project is not reflected in the average market ratios. If it were, the error would go in the “right” direction, as the value of BIG Bank SA would simply be overstated. Their application to the BIG Bank SA data yields a stable value estimate at PLN 140–150 million, with the extremes being PLN 80 and 200 million. For detailed calculations see Appendix F.
Berk (1997), pp. 17–18. The paper provides a decision rule applicable to companies with options traded on their stock. Dixit, Pyndick (1994), p. 136 state that for reasonable parameter values, investment should occur only if the gross present value is two to three times the investment costs.
Szymanski, Troy, Bharadwaj (1995), p. 18
See, for instance, Huff, Robinson (1994), pp. 1376–1377; Golder, Tellis (1993), pp. 166–169 or the survey articles of Kerin, Varadarajan, Peterson (1992); Szymanski, Troy, Bharadwaj (1995) or Clement, Litfin, Vanini (1998). In their survey of empirical work to date, Clement, Litfin, Vanini (1998) find that 15 out of 18 studies support the existence of a pioneering advantage. Murthi, Srinivasam, Kalyanaram (1996), p. 335 confirm the existence of a pioneering advantage even after controlling for influences other than order of entry. However, another important finding is also that pioneering by itself does not guarantee success in a market (Clement, Litfin, Vanini (1998), p. 222 ).
Kerin, Varadarajan, Peterson (1992), p. 46; Szymanski, Troy, Bharadwaj (1995), pp. 29–30
Arnold, Quelch (1998), pp. 10–11 identify five types of first mover advantages which are particularly important in an emerging markets context: favorable government relations, pent-up demand, marketing productivity, access to scarce marketing ressources and consequent learning (i.e. transferability of learnings to other emerging markets). Of these, both pent-up demand and higher productivity of marketing expenditures appear most applicable to the situation under consideration.
Kerin, Varadarajan, Peterson (1992), p. 35
Clement, Litfin, Vanini (1998), p. 208; Farnell, Robinson, Wernerfelt (1985), p. 1098
Pentor (1997). The exact distribution cannot be further specified.
It is correct to use the probabilities implied by the original underlying process, although the ENPV and not the gross project value is being discounted back. It becomes clearer by framing the problem in a single-period model. In such a case, deferring the investment would have been valued using the joint value of the options to defer, expand and shut for the given period (see, for example, Smit, Trigeorgis (1998), p. 10). The value of the option to defer is derived using the probabilities from the original underlying process.
See, for instance, Fudenberg, Tirole (1991), pp. 9–10.
This term has been used by Smit, Trigeorgis (1998), p. 11 to describe a similar outcome in a one-period game.
BSK, for instance, has probably superior experience in IT implementation as they are the first bank in Poland to have implemented a modern IT banking system for their existing operations (Thomson Bankwatch (1997), p. 5).
Rights and permissions
Copyright information
© 2000 Springer Fachmedien Wiesbaden
About this chapter
Cite this chapter
Müller, J. (2000). Applying real option valuation to a retail banking startup in Poland. In: Real Option Valuation in Service Industries. Gabler Edition Wissenschaft. Deutscher Universitätsverlag, Wiesbaden. https://doi.org/10.1007/978-3-322-99299-4_4
Download citation
DOI: https://doi.org/10.1007/978-3-322-99299-4_4
Publisher Name: Deutscher Universitätsverlag, Wiesbaden
Print ISBN: 978-3-8244-7138-6
Online ISBN: 978-3-322-99299-4
eBook Packages: Springer Book Archive