Dimitris Katsianis. University of Athens Dept of Informatics & Telecommunications dkats@di.uoa.gr

Transcription

1 Techno Economic Methodology for the Evaluation of Telecommunication Investment Projects. Sensitivity and Risk Analysis Incorporation Dimitris Katsianis University of Athens Dept of Informatics & Telecommunications International Telecommunication Union- Telecommunication Development Bureau Market, Economics & Finance Unit Expert Dialogues: October 2004 Geneva, Switzerland

2 The challenge Market Demand Willingness to pay User behaviour Technology Technology variety Open provisioning Service integration Strategy Where? When? How? s2

3 Consolidation of Results and Guidelines for deployment scenarios Guidelines Projects and field trials Common conclusions Other Sources Network Studies Common framework Information gathering / exchange s3

4 Operators Suppliers Standardization body Other User inputs Components Database Cost Evolution OA&M Class Volume Class Services Architectures Radio Model Revenues... Revenues Year 2 Investments Revenues Year 1 Investments Cash Flows Revenues Investments Cash Profits Flows Investments Cash Profits Flows Cash Profits Flows Profits Year n Financial Model Decision Index Index calculation (NPV, (NPV, IRR, IRR, Payback Payback period) period) Operators Surveys Market Size Policy Tariffs Game Theory Real Options s4

5 Steps in Network Evaluation Definition of service basket Network scenarios First Simulations Main Financial results Sensitivity and Risk Analysis Evaluation Recommendation and Guidelines s5

6 The TONIC Tool Based on Office 2000 platform Excel & Access Automatic sensitivity analysis Compatibility with Risk Analysis Tool(s) s6

7 The TONIC tool & its database s7

8 s8 University of Athens Dept of Informatics & Telecommunications Cost model P(0), the price in the reference year 0 n r (0), the relative accumulated volume in year 0, T, the time for the accumulated volume to grow from 10 % to 90 %, K, the learning curve coefficient. ( ) ( ) ( ) K t T r n e r n P t P + = 2 log 1 ln ln

9 Relative cost evolution as a function of T with n r (0)= ,00 0,80 0,60 0,40 0, year DT s9

10 Ecosys project University of Athens Dept of Informatics & Telecommunications WP0 - Project Management and Coordination WP1 Market dynamics WP2 Techno-Economic methodology development WP3 Tool development for T-E modelling WP4 Broadband for all - Economics of new networks and services WP6 Convergence WP5 Mobile and wireless network economics beyond 3G WP7 Dissemination and Exploitation s10

11 The new Tool Ecosys Based on Office 2002 platform Multiplayer environments Real Options implementatio New demand models.. s11

12 Main Financial Results Net Present Value, NPV Internal Rate of Return, IRR Payback Period Financial indicators Investments Running Costs Revenues Cash Flows Depreciation Profits Taxes Retained Cash Flows Cash Balance Rest Value s12

13 Scalability of the tool University of Athens Dept of Informatics & Telecommunications Sensitivity Analysis Risk Analysis s13

14 Sensitivity Analysis University of Athens Dept of Informatics & Telecommunications What if? Approach select the most critical input parameters establish boundaries for their variation with a «95% confidence interval» Results impact on NPV at boundary input parameter values: new NPV sensitivity factor: how NPV varies (slope at base value) impact on IRR at boundary input parameter values: new IRR sensitivity factor: slope at base value, although variation usually non linear s14

15 Input: Risk Analysis Uncertainty in market parameters Market size Market share Broadband services characteristics Uncertainty in Cost parameters Cost units Cost evolution Area characteristics Outputs: University of Athens Dept of Informatics & Telecommunications Probability measures for a reduced set of parameters s15

16 Method University of Athens Dept of Informatics & Telecommunications Project setting perform simulation for every decision vector risk profile store result compare results* make decision s16

17 Risk Analysis Component Price Trials NPV Frequency Chart 52 Outliers, ,68 0,74 0,80 0,86 0,92 Service Penetration Probability,020,013, ,7 134,5 67,25 Frequency 1,09 1,54 2,00 2,46 2,91 Revenue per customer 0,55 0,78 1,00 1,23 1,45, ke Statistical Variation of the input parameters Using Monte Carlo Simulation Results: probability distribution, risk profile of the business case Extended basis for investment decisions 0 s17

18 Risk Analysis - NPV Forecast: NPV Trials Frequency Chart 988 Displayed,038 38,029 28,5,019 19,010 9,5, Certainty is 82,10% from 0 to +Infinity s18

19 Requirements for a T-E study Services Scenarios Dimensioning Commercial Network Architectures. For these services Database Serving areas T-E Model Constructions Study period (years?) Potential market Market Shares (e.g operator) Pricing Runs- Results Sensitivity and Risk Analysis Evaluation of the results Recommendation and Guidelines - Commercial viability s19

20 Example case Location base Service LBS

21 Blend of cases Positioning Tariff Model Business Profile Terminals Countries s21

22 Country Types: Country Type Large Small Description Area size 370, ,000 Size of surface area of the country (km 2 ) Area dense Size of dense urban area (km 2 ). Area urban 2,960 4,000 Size of urban area (km 2 ) Area suburban 37,000 10,956 Size of suburban area (km 2. Area rural 303, ,956 Size of rural area ( km 2 ). Population dense 50,000 10,000 Number of inhabitants in dense urban area per km 2 Population urban 4,000 1,216 Number of inhabitants in urban area per km 2 Population suburban 1, Number of inhabitants in suburban area per km 2 Population rural Total Population 65,000,000 11,000,000 Total population Number of inhabitants in rural area per square km (during busy hour) s22

23 Tariff and revenue forecasts Services a) LBS services b) M-Guide Service Study Period: 7 years Parameters Nr of Queries per day (2004) Start Price per Query ( )(2004) End Price per Query ( ) (2009) Nr of main Services Value s23

24 Demand models 12,000,000 10,000,000 8,000,000 6,000,000 4,000,000 Mobile_Users Local_users_LBS LBS_MGUIDE_users Local_users_GPRS 2,000, s24

25 Main Financial Indexes 300 NPV - IRR - Payback Period 45 M Euros NPV 33.8 IRR PayBackPeriod IRR (%) Pay-back period (years) 0 0 Large Small s25

26 Net present value Internal rate of return Payback period University of Athens Dept of Informatics & Telecommunications Main results Small 35.7 M 33.8% 5.4 years Revenues SC Investments SC Running Costs SC Cash Balance SC Maximum Finance need 36 M M Euros s26

27 Cash Balances Large and Small Country Cash Balance SC Cash Flows SC Cash Balance LC Cash Flows LC M Euros s27

28 Operational Expenditures (OPEX) SC 60 Other Cost (Content) Employees 50 Marketing Subsidization GPS Terminals 12% 43% % 8% s28

29 Revenues LC Advertising LBS Roamers LBS Local MGuide Roamers 14% 11% 3% 6% 600 MGuide Local % s29

30 Sensitivity Analysis NPV sensitivity ranges (single parameter change) University of Athens Dept of Informatics & Telecommunications Tariff Market Share Local users percentage other LBS LBS percentage Mobile Penetration Low value High value Marketing Content Profit Advertising Revenues LBS Roamers percentage GPS Terminal Subs Change in NPV in MEURO (compared to base case : 35.7 MEURO) s30

31 Risk Analysis Marketingmultiplier Trials NPV Frequency Chart 52 Outliers, Probability,020,013, ,7 134,5 67,25 Frequency MarketShareMult , ke Statistical Variation of the input parameters Using Monte Carlo Simulation Results: probability distribution, risk profile of the business case Extended basis for investment decisions 0 s31

32 Risk Analysis - NPV Forecast: NPV 10,000 Trials Frequency Chart 9,845 Displayed ,463,974-30,140,537 23,182,901 76,506, ,829,777 Certainty is 69.22% from 0 to +Infinity s32

33 Risk Analysis NPV (2) Forecast: NPV 10,000 Trials Frequency Chart 9,845 Displayed ,463,974-30,140,537 23,182,901 76,506, ,829,777 Certainty is 33.97% from 35,698,326 to +Infinity s33

34 Conclusions LBS Case Acceptable business opportunities LBS can still be an attractive opportunity for companies with or without telecom background Payback period of 5 to 6 years, with a yearly ARPU of over than 27 for more enthusiastic testbeds Worst-case scenario Risk analysis evaluation shows that almost 30% of the project cases could have significant profits and 70% of them remain positive The uncertainty level is high mainly relating to the LBS penetration and market share One-year delay of this project could be reasonable in order to answer some critical questions s34

35 Time for Questions & Answers