Technology Investment Decision-Making under Uncertainty in Mobile - PowerPoint PPT Presentation

Technology Investment Decision-Making under Uncertainty in Mobile Payment Systems Robert J. Kauffman, Jun Liu and Dan Ma School of Information Systems Singapore Management University

Introduction • 2012, the year in payments • Mobile payments: • Near field communication (NFC)-enabled • Cloud-based Apple Passbook • Third party app

Motivation • Uncertainties • Consumer demand • Multi-sided business platform • Revenue model and collaboration • NFC smartphones and merchant terminals • Technology standard and regulation • Banks are key stakeholders in payments. • Decision making under uncertainty • Research Questions: • How can a bank maximize the business value of m- payments technology adoption under uncertainty? • How long can a bank postpone its investment and commitment to a specific technological solution?

Cost, Benefits and Deferral Geometric Brownian Motion • Investment time Value 12 Investment Cost Benefit Flows horizon [0, T ] 10 8 • Investment cost I 6 4 follows geometric 2 Brownian motion: Time 0 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 • Benefit flows B from time t to T follow geometric Brownian motion:

Model Preliminary Invest I Receive Benefits Flows Wait 0 t T • No competitor, no correlation • The value equals to integration over the interval ( t , T ) is: • Expected investment cost I is:

Real Option Value • Deferral Option: 𝑂𝑄𝑊 𝐵 = 𝑊 − 𝐽 + 𝑆𝑃𝑊 = max⁡ (𝑊 − 𝐽, 0) 𝑆𝑃𝑊 = 𝑛𝑏𝑦 0, 𝐽 − 𝑊 • By Bellman optimality equation , 𝑠 𝑔 𝑆𝑃𝑊𝑒𝑢 = 𝐹(𝑒𝑆𝑃𝑊) , we obtain: 1 𝐶𝐶 + 1 2 𝜏 𝐶2 𝐶 2 𝑆𝑃𝑊 2 𝜏 𝐽2 𝐽 2 𝑆𝑃𝑊 𝐽𝐽 + 𝛽 𝐶 − 𝜃 𝐶 𝐶𝑆𝑃𝑊 𝐶 + 𝛽 𝐽 − 𝜃 𝐽 𝐽𝑆𝑃𝑊 𝐽 + 𝑆𝑃𝑊 𝑢 − 𝑠 𝑔 𝑆𝑃𝑊 = 0 Two boundary conditions: 𝑆𝑃𝑊 𝐶, 𝐽, 𝑈 = 0 , 𝑆𝑃𝑊 𝐶, 𝐽, 𝑢 ≥ 0⁡⁡⁡⁡∀⁡0⁡ ≤ ⁡𝑢⁡ < ⁡𝑈

Numerical Analysis Description Value Description Value 𝐽 0 𝐶 0 ( t ) Initial investment $10 million Initial benefit flow $0.1-1.0 million 𝛽 𝐽 𝛽 𝐶 Rate of cost change -0.1 Rate of benefit change 0.7 𝜏 𝐽 𝜏 𝐶 Cost uncertainty 0.2 Benefit uncertainty 1.0-0.1 𝑠 T Maximal deferral time 5 years 𝑔 Risk-free discount rate 6% Δt N No. of simulated paths 100,000 Duration of time step 1 month Investment timing benchmark simulation Optimal timing t = 14 month; Maximal payoff is $4.10 million

Sensitivity Analysis Benchmarking, t = 14 When T = 6 years, t = 13 When T = 4 years, t = 15 When α B = 0.8, t = 11 When r f = 0.5, t = 13 When α B = 0.6, t = 18 When r f = 0.7, t = 16 Sensitivity analysis of benchmark simulation with respect to T , r f and α B

Jump Diffusion Process • Benefit flows when jump events happen: Benefit flow = Continuous Benefit Flows + Jump Value 𝑒𝐶 = 𝛽 𝐶 + 𝜇𝑙 𝐶𝑒𝑢 + 𝜏 𝐶 𝐶𝑒𝑨 + (𝑍 − 1)𝐶𝑒𝑟 Description Value Description Value λ Mean jumps number 0.05 % change of benefits 0.5 K Upward Jump at t = 20, t = 12 Benchmarking, t = 14 Catastrophic Jump at t = 10, t = 14

Jump Diffusion Simulation Catastrophic Jump at t = 40, t = 15 Upward Jump at t = 10, t = 9 Upward Jump at t = 4, t = 14 Catastrophic Jump at t = 20, t = 20 • Least-Squares Monte Carlo Method (Longstaff and Schwartz) With Jump

Conclusion • Managerial implication • Investment in Multi-sided business platform • First mover advantage vs. second mover advantage • Rational expectations of senior management • Contribution • A new modeling perspective on how financial economics theory support m-payments decision-making • Help senior manager estimate optimal timing and payoffs from m-payments • Use of jump diffusion process to model the dynamically changing of value of IT investment