Skip to content

Talking Decision Analytics Series Part 4: Risk Metrics Continued

| February 28, 2023 | By

Management may wish to avoid the risk of an event such as an earthquake, but ultimately risk events, have to be translated into metrics that can be used directly to guide decisions.

First, recall that risk events can describe disruptions to the physical system, such as inventory stockouts, production shutdowns and work stoppages. Then there are financial risk events, such as not meeting an earnings target or a forecast for revenue growth.  These events have to be defined by management.

Risk metrics come in different flavors.  

  • Measures of the severity of an event, such as the amount of demand not met due to a stockout, or degree to which we miss a financial forecast.  
  • Measures of the frequency of an event - Risk events typically happen only occasionally, or rarely, or they may be something that has never happened, but which management judges is something that *might* happen.

Some authors define “risk” as the product of the severity times the probability of a risk event. However, a high-impact risk event that might occur with very low probability might then produce a low risk score, which might not command the resources needed to mitigate the severity of the event, which may threaten the survival of the company. In practice, management has to balance the probability of an event with its severity to make a judgment whether they want to spend the money needed to reduce the event into something manageable. 

The term “risk” can also be used to describe frequent, operational events such as missing the load for an important customer, not getting a driver home on time, or incurring stock outs which, while uncommon, are not rare. However, each of these examples have impacts that may have broader impacts, such as loss of a major account, loss of a driver, and damage to company reputation.

A green semi-truck driving down a highway.