Free SOA Exam ASTAM (Advanced Short-Term Actuarial Mathematics) Aggregate Models Practice Questions

Aggregate loss models on SOA Exam ASTAM test compound frequency-severity models, the Panjer recursive method, fast Fourier transform (FFT) approaches, and stop-loss reinsurance pricing.

155 questions 62 easy 56 medium 37 hard 2026 syllabus

Sample Questions

Question 1 Easy
Which of the following distributions belongs to the (a,b,0)(a, b, 0) class with a>0a > 0 and b≥0b \geq 0?
Solution
D is correct.

The (a,b,0)(a,b,0) class consists of counting distributions whose probabilities satisfy
pkpk−1=a+bk,k=1,2,3,…\frac{p_k}{p_{k-1}} = a + \frac{b}{k}, \qquad k = 1, 2, 3, \ldots
with p0>0p_0 > 0.

For the negative binomial with parameters r>0r > 0 and β>0\beta > 0,
pk=(r+k−1k)(β1+β)k(11+β)r,p_k = \binom{r + k - 1}{k} \left(\frac{\beta}{1+\beta}\right)^{k} \left(\frac{1}{1+\beta}\right)^{r},
so
pkpk−1=r+k−1k⋅β1+β=β1+β+(r−1)β/(1+β)k.\frac{p_k}{p_{k-1}} = \frac{r + k - 1}{k}\cdot\frac{\beta}{1+\beta} = \frac{\beta}{1+\beta} + \frac{(r-1)\beta/(1+\beta)}{k}.
Hence
a=β1+β>0,b=(r−1)β1+β.a = \frac{\beta}{1+\beta} > 0, \qquad b = \frac{(r-1)\beta}{1+\beta}.
With r≥1r \geq 1 we get b≥0b \geq 0, and b=0b = 0 exactly in the geometric case r=1r = 1. So the negative binomial is the (a,b,0)(a,b,0) member with a>0a > 0 and b≥0b \geq 0.

For comparison, the Poisson gives a=0a = 0, b=λb = \lambda, and the binomial gives a=−q/(1−q)<0a = -q/(1-q) < 0, b=(m+1)q/(1−q)b = (m+1)q/(1-q); both are in the class but neither has a>0a > 0.
Question 2 Medium
An exponential severity with mean θ=500\theta = 500 is discretized using the method of rounding with span h=200h = 200. Compute the masses at grid points 0, 200, and 400.
Solution
B is correct.

Under the method of rounding with span h=200h=200, grid point k⋅hk \cdot h collects all probability mass within [(k−0.5)h,(k+0.5)h)[(k-0.5)h, (k+0.5)h). For k=0k=0: the interval is (−100,100)(-100, 100), but since support is [0,∞)[0,\infty), this is [0,100)[0, 100): p0=F(100)−F(0)=(1−e−100/500)−0=1−e−0.2≈0.1813p_0 = F(100) - F(0) = (1-e^{-100/500}) - 0 = 1 - e^{-0.2} \approx 0.1813. For k=1k=1 (grid point 200): interval [100,300)[100, 300): p200=F(300)−F(100)=e−0.2−e−0.6≈0.8187−0.5488=0.2700p_{200} = F(300) - F(100) = e^{-0.2} - e^{-0.6} \approx 0.8187 - 0.5488 = 0.2700. For k=2k=2 (grid point 400): interval [300,500)[300, 500): p400=F(500)−F(300)=e−0.6−e−1.0≈0.5488−0.3679=0.1809p_{400} = F(500) - F(300) = e^{-0.6} - e^{-1.0} \approx 0.5488 - 0.3679 = 0.1809.
Question 3 Hard
For a compound Poisson SS with λ=2\lambda = 2 and severity P(X=1)=P(X=2)=0.5P(X=1) = P(X=2) = 0.5, use the Panjer recursion to compute g0g_0, g1g_1, and g2g_2, then find P(S≤2)P(S \leq 2).
Solution
E is correct.

For compound Poisson with a=0a=0, b=λ=2b=\lambda=2, and f1=f2=0.5f_1 = f_2 = 0.5: g0=e−λ=e−2g_0 = e^{-\lambda} = e^{-2}. For s=1s=1: g1=b1f1g0=21(0.5)(e−2)=e−2.g_1 = \frac{b}{1}f_1 g_0 = \frac{2}{1}(0.5)(e^{-2}) = e^{-2}. For s=2s=2: g2=b2f1g1+2b2f2g0=22(0.5)(e−2)+2⋅22(0.5)(e−2)=0.5e−2+e−2=1.5e−2.g_2 = \frac{b}{2}f_1 g_1 + \frac{2b}{2}f_2 g_0 = \frac{2}{2}(0.5)(e^{-2}) + \frac{2 \cdot 2}{2}(0.5)(e^{-2}) = 0.5e^{-2} + e^{-2} = 1.5e^{-2}. Therefore P(S≤2)=e−2(1+1+1.5)=3.5e−2≈0.4726P(S \leq 2) = e^{-2}(1 + 1 + 1.5) = 3.5e^{-2} \approx 0.4726.

Guides & Articles

About FreeFellow

Jeffrey Ting, founder of FreeFellow
Jeffrey Ting
FSA, CFA · Founder

FreeFellow was built by Jeffrey Ting, a credentialed actuary and CFA charterholder who has passed thirteen of the hardest exams in finance, all on his first attempt. He paid four-figure prep fees along the way.

So he started writing his own questions, then lessons, then mock exams, until it grew into a full prep platform covering 40 exams with more than 45,000 original practice questions.

01
Cost shouldn't decide who gets in.

A CFA charter, a CPA license, an actuarial credential. Each opens a real career, but prep fees add to the cost of getting there. FreeFellow keeps its original question bank and written lessons free so you can study even if a paid course is out of reach.

02
Free should mean free.

The original question bank, the worked solutions, the topic lessons, mixed practice, and your readiness score stay free, forever. Full access needs a free account, and practice usage limits apply. Fellow adds AI grading on supported exams, spaced-repetition flashcards, full-length mock exams, analytics, and a study plan that adapts to your progress.