Skip to content
All library documents

Estimating Measured Return in Category 3 PRIIP MRM

Article Quant Q&A · Author: DrDoolittle
Read the original

Summary

This exchange addresses how to calculate the measured-return input in an example for the market risk measure (MRM) of a Category 3 PRIIP. The questioner compares two interpretations: averaging a set of assigned daily log returns and multiplying by the number of observations, or averaging returns across the full observed history and multiplying by that number. The first approach does not reproduce the example’s reported figure, while the second comes close.

The answer supports the second interpretation, referring to the guidance as requiring the mean return over the observed period multiplied by N. It does not provide a worked calculation or independently validate the example, so the conclusion is a reading of the cited regulatory instructions rather than a demonstrated derivation. The exchange also leaves an associated question about calculating the risk-free return over the recommended holding period unanswered.

Key ideas

  • The question concerns the measured-return input used in a Category 3 PRIIP MRM example.
  • The answer interprets the procedure as the mean return across the observed period multiplied by N.
  • The exchange does not show a worked calculation to verify the interpretation.
  • The related question about the risk-free return over the holding period remains unanswered.

Tags

Risk managementStatistics
Full text
# Category 3 PRIIP MRM calculation


# Category 3 PRIIP MRM calculation












I know that there is already a very similar post about this. Unfortunately it does not correspond exactly to my question, so I will try it again myself. I am trying to calculate the MRM of a Category 3 PRIIP. After I was unsure about my result I wanted to recalculate the example shown in the flow chart.

EU Flow chart, Example on page 12

I get all results except E[RETURN MEASURED]. I thought it was required to calculate the mean value of the N assigned daily logreturns times N. But this does not even come close to the 0.004067173 from the example. But I get very close to the result if I calculate the mean of all available historic returns and multiply it by N. Can anyone tell me if the second approach is right?

The formulations are unfortunately very incomprehensible and in contrast to E[Return_risk-neutral], E[RETURN MEASURED] is not explained.

Tank you!

## Answer by user50535 (score 1)

https://quant.stackexchange.com/a/58842

the second approach is right from my reading. Page 10 step 4 third bullet says you need to calculate the mean return over the observed period and multiply by N. This is the second approach you describe above.

Query from my side - i can get all the numbers in the example except the risk free return over the RHP. how did you calculate this?

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.