Skip to content
All library documents

Risk Numbers, VaR and Expected Shortfall in Position Risk Modeling

Article Quant Q&A · Author: Math

Summary

The note clarifies two uses of risk terminology in finance. “Risk numbers” commonly refers to quantitative measures of the risk in a position or trading desk, such as value at risk (VaR) or expected shortfall (ES). These measures summarize potential losses according to a chosen methodology and are distinct from the risk-neutral probability measure used in pricing derivatives.

“Modeling risk” means identifying the relevant sources of exposure, such as interest-rate or volatility risk, and selecting a model to quantify them. The example given is a delta-normal VaR approach. The response is deliberately general because the original question lacks context; it does not define VaR or ES mathematically, compare their strengths, or specify a confidence level, horizon, or model assumptions. The particular risk measure and model therefore depend on the position and the risk being assessed.

Key ideas

  • Risk numbers commonly quantify the risk of a position or trading desk.
  • Value at risk and expected shortfall are examples of such measures.
  • Risk modeling starts by identifying relevant exposures, such as interest-rate or volatility risk.
  • A chosen model, such as delta-normal VaR, is then used to calculate risk measures.
  • The appropriate measure and model depend on the context and are not specified in detail here.

Tags

Full text
# What are "risk" or "risk numbers?"


# What are "risk" or "risk numbers?"












I work in finance but do not have any formal education in the subject (I do have a PhD, but not in finance). I've picked up a lot of the jargon but there's one thing that I haven't figured out, and it seems too basic to ask. Hence I ask here.

What do people mean when they say they are "modeling risk" or "risk numbers?" The term "risk" here seems to have a very specific meaning, and it doesn't seem to coincide with the way the word is used in everyday speech. I'm not sure if it has anything to do with the risk-neutral measure, which is something I do understand.

## Answer by Dr_Be (score 2, accepted)

https://quant.stackexchange.com/a/22726

Very hard to tell from the given context but I would assume that "risk numbers" refers to the risk of a position (e.g. trading desk) which is typically measured in terms of a VaR (value-at-risk) or a similar measure (ES - expected shortfall).

"Modeling risk" is the task of identifying your sources of risk (risk types like e.g. interest rate risk, volatility risk) and then choosing an adequate model (e.g. VaR using delta-normal approach) in order to calculate the aforementioned risk numbers.

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.