Pricing Insurance Risk Book Errata

publications
pricing
insurance
risk
pir
Errata list of PIR. Report issues to mynl@me.com.
Author

Stephen J. Mildenhall

Published

2024-06-23

Modified

2025-01-30

Main PIR page

Reporting Errors

If you discover a mistake in PIR please email me at mynl@me.com.

Other Errata

  • p. 20: Table 2.4 μ for SCS should be 1.805 not 1.805, making the mean severity 1
  • p. 24-5: line -1 and 1, Cat Hu is extremely skewed… though Cat Hu is thicker
  • p. 64 line 2: “Is corresponds” s/b “It corresponds”. (JAM)
  • p. 64 Figure 4.1 is created with σ=0.5 not σ=1 as written. (SJM)
  • p. 76: Example 54 says μ=21.315 but in fact μ=19.005 in order for severity to have a mean of USD 5B.
  • p. 85: The paper by Daníelsson and Jorgensen [] does report that VaR is subadditive for log-concave distributions but this is not true, as the example of the sum of two exponential distributions at p=0.7 shows (VaR of the exponential is 1.20, and the gamma sum is 2.44>2×1.2).
  • p. 88 algorithm step 6, s/b upper case Xn, not lower. (JAM)
  • p. 93 line -3 should read: “…observe first that Pr(Xx0)=1p is larger than it should be…” rather than Pr(Xx0)=p is larger. (SJM)
  • p. 98 Exercise 72 part 2 should read CTEp<=CTEp+<=TVaRp. (AM)
  • p. 101 Display equation line 3, drop q(p); it is included in the integral. (JAM)
  • p. 123: in bottom displayed equation ρ=minx{x+V(Xs} not the max as stated.
  • p. 244, Exercise 141 solution part 2, h(0.05)=1g(10.95)=0.0203 should say h(0.05)=1g(10.05)=0.0203.
  • p. 273 para above (11.2) s/b outcome-probability not event-outcome; after equation more natural than, not then. (SJ/JM)
  • p. 277 Ex 185 item 5, s/b “What conditions on p0, p1, …”. (JAM)
  • p. 372 Penultimate displayed is an expression for dS¯i/da not dS¯/da, i.e., add subscript i.
  • p. 378 Eq (14.34) first line should read M¯=P¯iS¯i, not L¯i. (JAM)
  • p. 380 line 2 should read βi(x)/αi(x)>1>S(x)/g(S(x)), not g(S(x))/S(x) which is always 1. (AB)
  • p. 411 para 1, last sentence should read “Practical applications should…”
  • p. 413-15: there are two tables numbered 15.21. The one referenced in Exercise 273 is on p. 415, not 414.
  • p. 419: Note that the covariance premium is computed by calibrating and allocating the load using unlimited losses. The expected loss component is based on limited losses.

References

References

1.
Dan??elsson, J., Jorgensen, B.N.B., Samorodnitsky, G., Sarma, M., De Vries, C.G., Daníelsson, J., Jorgensen, B.N.B.: Fat tails, VaR and subadditivity. Journal of Financial Econometrics. 172, 1–30 (2013). https://doi.org/10.1016/j.jeconom.2012.08.011