Our approach to risk quantification
Stochastic risk models
Business processes are generally volatile and therefore difficult to forecast for planning and strategic decisions.
Many different scenarios might occur in the future.
"Risk" is the inherent uncertainty in how the future will be and how business objectives will be met.
The only way to deal with the uncertainty is to include it actively in the planning or decision making process, to find the right balance between daring and prudence, risk and reward.
This can be done with the help of stochastic risk models. They represent an improvement on single point or 3 point (i.e. worst, best and most likely case) estimates.
Going beyond qualitative risk profiling
Typically, risk identification, assessment and prioritisation processes describe the potential impact of a risk with one or two parameters (i.e. ‘risk magnitude’ is a function of financial impact and likelihood of occurrence).
However, a risk scenario will normally produce a broad range of outcomes.
Take a classical hazard risk - fire. Outcomes range from a hole in the carpet because of a dropped cigarette to a 100 million loss for a burned down factory.
Additionally, multiple events of one risk scenario can add up to become even more costly than the original feared ‘big event’ (example: mid-size grocery shop being afraid of one major supermarket opening, but finally sees its turnover eroded by 10 decentralised petrol stations selling groceries as well).
Where historical information or hard data exist one can use standard statistical or actuarial methods to estimate complete risk statistics.
Often, however, statistics are not available or not applicable because of completely new problems (e.g. internet)
Then we have to rely on expert opinions.
Our Quant1 process quantifies the down-side potential of risks by
- capturing experts’ opinions of loss severties and loss frequencies
- calculating statistics of individual loss scenarios
- calculating the statistics of aggregated total losses on an organisation
- Quant1 is supported by our software tools and facilitation skills
Modelling the outcomes of optional strategies
We offer a range of financial analysis and stochastic modelling services.
We have the capability to build individually tailored stochastic business models to evaluate specific questions/problems to support decision making.
For example, consider the stochastic model of a company’s next years profit, taking various risks/uncertainty factors into account and evaluating different strategies
Each strategy introduces risk to the profit distributions. Which one to choose?
Normally it’s not easy to choose a single strategy, because there is often no clear winner.
We can rule out Strategy 1 (it looks as volatile as Strategy 2, but at lower mean).
Is Strategy 2 or Strategy 3 better?
Strategy 3 has a higher mean, whereas Strategy 2 has lower volatility (= risk).
Which one is better depends on risk appetite. There are ways to evaluate further.
Using risk appetite to optimise strategy
The risk/return diagram uses Standard Deviation as a proxy for ‘Risk’ on the x axis and mean expected profit as ‘Return’ on the y axis.
Each risk distribution curve is plotted as a single point on this diagram.
Strategy 3 is "sub-optimal" because a higher return can be achieved at the same risk or the same return can be achieved at much lower risk.
Which is the optimal choice? Risk appetite and return expectations can guide the decision.
Imagine that management promised a profit of at least 175 million, then all strategiesthat have expected profit below 175 million should be excluded.
Imagine further that management wants to have a 5% chance (risk) at worst of having a negative profit (= loss) next year, then all strategies above a risk level of 106 should be excluded.
Strategy 2 therefore offers the best combination of risk and return.