IAI Actuarial Core Principles · Risk Modelling and Survival Analysis · Introduction to copulas
Which statement about tail dependence in common copulas is correct?
The Clayton copula with a positive parameter has lower tail dependence and no upper tail dependence. Gumbel is the reverse, while Gaussian, Frank and independence copulas are tail independent, which makes Clayton suitable for modelling joint extreme lows.
- AThe Gaussian copula with correlation below 1 has positive upper tail dependence
- BThe Clayton copula with positive parameter has lower tail dependence but no upper tail dependenceCorrect
- CThe Gumbel copula has lower tail dependence but no upper tail dependence
- DThe independence copula has tail dependence coefficient 0.5
- The Frank copula has positive tail dependence in both tails
Explanation
The Clayton copula with positive parameter has lower tail dependence of 2^(-1/alpha) and zero upper tail dependence. The Gaussian copula with correlation below 1 is tail independent. The Gumbel copula has upper, not lower, tail dependence. The independence and Frank copulas have no tail dependence.
Did you get it right without looking?
One question tells you little. A timed set on Introduction to copulas shows your real accuracy, how long you take and where you lose marks.
More Introduction to copulas questions
- If X has a continuous distribution function F_X, which statement is used to obtain the uniform inputs for a copula from observed losses?
- A reinsurer uses the Fréchet–Hoeffding bounds to assess aggregate risk on two lines. Which statement about the comonotonic copula M(u,v)=min…
- Two loss variables X and Y have a Gaussian copula. A modeller replaces both marginals by new continuous distributions, each via a strictly i…
- A Gumbel copula, C(u,v) = exp(-[(-ln u)^k + (-ln v)^k]^(1/k)) with k >= 1, is fitted to the joint claims of two lines of business. Which fea…
- Two risks each have a standard uniform marginal and are joined by the Clayton copula C(u,v) = (u^(-θ) + v^(-θ) − 1)^(-1/θ) with θ = 1. Calcu…
- An insurer finds the empirical probability P(U1 > 0.99 | U2 > 0.99) is about 0.60 and P(U1 > 0.999 | U2 > 0.999) is about 0.05 for pseudo-ob…