In this article, Samuel BRAL (ESSEC Business School, Global BBA – Exchange at NUS, 2025) explains how he applied Monte Carlo simulations to support Emirates Airlines in evaluating the profitability of launching a new long-haul route under conditions of uncertainty.
Context of the project
This project was part of the course “Decision Analytics using Spreadsheets” at the National University of Singapore (NUS). I was asked to provide a quantitative recommendation to Emirates Airlines on selecting a new international route from Dubai. The available destination options included Buenos Aires, Tokyo, Cape Town, and Cairo.
Due to the complexity of airline operations and the uncertainty surrounding factors such as demand, ticket prices, no-show rates, and operating costs, a traditional static financial model would not be sufficient. Instead, I built a Monte Carlo simulation model to capture the dynamic range of possible outcomes and assess the risk-return profile of each destination.
What is a Monte Carlo simulation?
A Monte Carlo simulation is a mathematical technique used to estimate the probability distribution of outcomes when there is uncertainty in the input variables. By running thousands of simulations using random values generated from defined probability distributions, the method provides insights into the range, likelihood, and volatility of potential results.
This approach is commonly used in financial modeling, risk analysis, and engineering. For example, investment banks use Monte Carlo models to simulate portfolio returns and Value at Risk (VaR), while oil and gas companies apply them to forecast drilling success and production volumes.
Simulation approach and methodology
I built a simulation model in Excel that executed 2,000 trials per route. Each trial simulated a potential outcome based on randomly generated values for key variables. The profit was calculated using the following formula:
Profit = (Tickets Sold × Ticket Price) – Operating Costs – Compensation Costs
Here is how each component was modeled:
- Passenger demand: Modeled as a normal distribution using historical demand averages and standard deviations for each route. For example, Tokyo exhibited more stable demand, while Buenos Aires showed higher variance due to geopolitical and economic volatility in Argentina.
- Ticket price: Ticket prices were generated using
NORM.INV(RAND(), mean, stdev)to account for fluctuations caused by competitive pricing, seasonal variation, and macroeconomic factors like fuel costs and currency movements. - No-show rate: Modeled with a uniform distribution between 5% and 10%, based on IATA statistics and academic studies on airline overbooking behavior (source: IATA Global Passenger Survey, 2023).
- Aircraft assignment: Simulated using a discrete probability distribution based on the actual Emirates fleet composition (e.g., A380, Boeing 777). Larger aircraft allowed more passengers but incurred higher operating costs.
- Compensation cost: Incurred when demand exceeded seat capacity, reflecting the cost of rebooking, refunds, and customer service. These costs were calibrated using Emirates’ historical compensation data for overbooking cases (source: Emirates Annual Report 2023).
To execute the simulations, I used Excel’s Data Table function to loop through trials and capture the output profit distribution for each destination. From this distribution, I calculated:
- Expected profit (mean)
- Standard deviation of profit (volatility)
- Probability of a loss (profit < 0)
- Probability of a significant loss (loss > SGD 100,000)
Key results and insights
The simulation identified Buenos Aires as the most profitable option with an expected profit of SGD 292,247 and a 99.65% chance of profitability. However, the route also exhibited a small 0.1% risk of incurring losses above SGD 100,000 due to volatile demand and long travel distance.
Cape Town, while less profitable, offered near-zero downside risk. Tokyo had moderate returns and relatively low variance. This reflects a classic risk-return tradeoff that airlines often face: should the company pursue high-reward but volatile destinations, or opt for stable but lower-margin routes?
Additionally, I tested various overbooking strategies. An overbooking rate of 9.3% was found to optimize expected profits while keeping the cost of passenger compensation within an acceptable range. This mirrors real-world practices, where carriers like Delta and Lufthansa use algorithmic overbooking based on historical no-show patterns to maximize seat utilization (source: MIT Airline Data Project). If you want to have access to the work, here is the Excel file on the overview of all routes as well as the work for Buenos Aires.
Why should I be interested in this post?
This project demonstrates how Monte Carlo simulations transform business decision-making under uncertainty. Instead of relying on single-point forecasts, the model enabled me to quantify risk, test strategic decisions (like overbooking), and provide data-driven recommendations.
For students and professionals in finance, consulting, or operations, Monte Carlo simulation is a core technique for scenario planning and risk assessment. It enhances decision quality in fields as diverse as project finance, asset management, supply chain optimization, and policy modeling.
Related posts on the SimTrade blog
▶
▶
▶
Useful resources
IATA Global Passenger Survey 2023
Emirates Annual Report and Press Releases
About the author
This post was written in September 2025 by Samuel BRAL (ESSEC Business School, Global Bachelor in Business Administration – Exchange at NUS).