Risk management is a powerful controlling and decision making tool for corporations. Graeme Keith and I went on a quest to model our lives and prove that risk management can be a powerful tool not just for companies, but for individuals as well.
The future is uncertain and we can use risk management techniques to model that uncertainty. This risk-based forecasting will allow us to test different hypothesis, make important life choices and understand what uncertainties in life matter and what doesn’t as much as we intuitively think. The conclusions will surprise you.
First we divided our hypothetical life into important elements:
- income and unemployment
- health and living expenses
- assets and maintenance
- investments and recessions
Identifying risks for each of these elements and modelling uncertainty will allow us to test important life hypothesis. Truth be told we did something a little more complex, we started with an influence diagram, because we should always start with the objective or decision or hypothesis at hand (top down), but for the sake of this series of articles we will continue bottom up. I will talk about hypothesis building and the influence diagram we built in a later article. Just keep in mind, understanding the big picture is the first step to any risk analysis.
Part 1. Income and unemployment
All number in the table below are hypothetical and can be replaced with your relevant data. All numbers are in USD. Selecting the currency for the modelling is important as you are likely to have cash flows in different currencies, don’t forget the conversion rates. You can use forward curves for the next 5 years or so and just add inflation for the following years. We have selected a 30 year horizon for the model.
|Randomize increases and decreases in base salary.|
|Year 0 salary||100000|
|Change in salary (year on year)|
|Increase (percentage) mean||15%|
|Increase (percentage) std. dev.||10%|
|Decrease (percentage) mean||15%|
|Decrease (percentage) std. dev.||10%|
|Additional, additive, stochastic component handled by a “bonus”, which covers annual fluctuations and, fluctuating, additional sources of income|
|Bonuses / additional sources of income
(as proportion of base)
|Fired: Loss of base income (additional sources maintained)||Probability (year on year)||10%|
|Risk sampled year on year. Geometric distribution on time||Probability work in 1 year||70%|
|Loss of ability to earn, assumed covered by insurance||Probability (year on year)||1,7E-03|
|Single payout of percentage of year 0 salary||Proportion of salary||80%|
We identified a set of important risks that could affect income stability:
- losing the job for an extended period of time, both the fact the job is lost and the duration of unemployment are uncertain
- illness or loss of ability to earn
Here is an example of one unlucky scenario with not one but two lengthy unemployments (dark blue line). I used this specific example for demonstration purposes. You can also see the p10 (probable worse case scenario), p50 (median), mean and p90 (probable best scenario). The light grey lines are actual scenarios. The further into the future, the higher uncertainty.