Demystifying Monte Carlo Analysis - What it is (And what it Isn't)

When it comes to risk analysis, Monte Carlo statistical analysis is often seen as the gold standard – yet many companies still don’t use it.

Why is that? 

I’ve heard countless reasons for avoiding Monte Carlo risk analysis, but the objections usually fall into two categories: 

1. “It’s too complex and academic.” This may have been true 20 or 30 years ago, but with today’s fast computers and widely available software, it’s now simple and cost-effective. 

   
   

2. “We don’t have good data to model it.” This is a myth. Monte Carlo analysis is specifically designed to work with uncertain data – that’s exactly what it was created for. 

   
   

In the hundreds of Monte Carlo simulations I’ve facilitated, teams are often surprised at how straightforward the process is and how much it improves their decision-making. 

Why the Misunderstanding About Quantified Risk Analysis (QRA)? 

A large part of the confusion around Monte Carlo analysis stems from the knowledge gap in many organisations. Often, well-meaning managers and engineers are given the title of “Risk Manager” without the necessary training in advanced risk management techniques. 

Our so-called “best practice” standards don’t help either. They can be contradictory across industries and rarely provide clear guidance on Quantified Risk Analysis (QRA). 

If you’re a Risk Manager, I highly recommend reading How to Measure Anything – Finding the Value of Intangibles in Business and The Failure of Risk Management by Douglas Hubbard. You’ll also find a wealth of resources on “Quantified Risk Analysis” by searching on Google or LinkedIn. 

What is Monte Carlo Analysis? 

At its core, Monte Carlo analysis is a statistical algorithm that uses random number generation to simulate thousands of potential outcomes based on a defined range of inputs. It gives you a distribution of possible futures rather than a single point estimate. 

Or, if you prefer a more imaginative analogy – think of it as a time machine that allows you to explore all the possible future timelines of an event, much like the Marvel multiverse. 

The output of a QRA is typically shown as a histogram, an S-curve, or both, which visually represents the range and likelihood of different outcomes. 

From this you can quickly identify the risk exposure your organisation or project is facing, based on your risk Apatite and or Tolerance.

Typical Quantitative Cost Risk Analysis (QCRA) Results

 This is a far better way to present risk exposure to management, that the traditional heatmap.  It may take a little explanation, but senior managers and executive boards are normally pretty astute people, so they will get it. 

Why Use Monte Carlo Analysis? 

Quantified Risk Analysis using Monte Carlo is designed to measure uncertainty. As I mentioned in my previous blog, the purpose of measurement is to improve our understanding – and any measurement is better than none if it moves us closer to clarity. 

If you’re currently using qualitative risk scoring, you can easily transition to quantitative analysis by assigning numerical ranges to your qualitative categories.

For example: 

• Risk X has a high likelihood and a medium impact.  

• High likelihood: 50%–80% 

• Medium impact: $50,000–$200,000 

  
  

What Monte Carlo Analysis Is NOT 

Let’s clear up some misconceptions: 

• It’s not difficult. 

• It doesn’t require perfect data. It works best with uncertainty. 

• It’s not just for academics or insurance companies. 

• It’s not just a project cost estimating tool. 

• It’s not a dark art. 

  
  

What Monte Carlo Analysis IS 

• An evidence-based modelling tool 

• The only method that can quantify the effectiveness of risk management 

• Used globally across industries and sectors 

• Applicable to everything from project risks to board-level strategic decisions 

  
  

That said, getting the best results does require someone with a solid understanding of risk management and uncertainty estimation.

When I first started as a junior risk manager, I received software training but no guidance on how to extract the best information from subject matter experts or how to challenge risk biases – in hindsight I can see how much this limited the quality of my analysis. 

Unfortunately, this is still what happens in some large consultancies, the junior risk analysts are normally the ones tasked to run the risk models for their clients.

A Short History of Monte Carlo Analysis 

Monte Carlo simulation was developed by Stanislaw Ulam, a scientist on the Manhattan Project (the U.S. effort to develop nuclear weapons during World War II). After the war, during his recovery from brain surgery, Ulam became fascinated with predicting outcomes in the card game Solitaire. Given the massive range of possibilities, he realised that the best way to estimate his chances of winning was to play many games and track the results. 

Ulam shared his idea with his colleague John von Neumann, who saw its potential for nuclear science. They used early computers to simulate the random behaviour of neutrons during nuclear explosions – and the Monte Carlo method was born. 

Early simulations took hours to compute, but modern computers can now process complex models in seconds. 

How to Build a Monte Carlo Model 

The key components of a Monte Carlo model are the input ranges for each potential outcome (or risk). This is usually captured as a lower and upper boundary – for example: 

• Lower value: $12,500 

• Upper value: $56,000 

  
  

Some models use a three-point estimate (lower bound, most likely value, and upper bound), but if you’re starting out, I recommend sticking to a simple lower and upper range. Aim for a 90% confidence interval – meaning you’re comfortable that 90% of the potential outcomes fall between these two values. 

Along with this you will need to decide the probability that the risk might occur.

If you want to dive deeper, Douglas Hubbard explains the benefits of this approach far better than I can. 

Some Final Thoughts for You 

Monte Carlo analysis is a practical, evidence-based tool that helps you make better decisions in the face of uncertainty. It’s not mysterious or overly complex – and it’s a skill that can dramatically improve your approach to risk management. 

If you’re ready to take your risk analysis to the next level, start small, ask questions, and embrace the uncertainty – because that’s exactly what Monte Carlo was made for. 

Fox Risk are specialists in both Quantitative Cost Risk Analysis and Quantitative Schedule Risk Analysis.   We have access to some of the best know QRA analysis software platforms and are here to help. 

Contact me for is you want to know more.