WorksMathematics
The Laplace transform of the time of ruin in the Cramér–Lundberg model with Erlang claims: spectrum, exact moments and matrix reduction
Abstract
We study the Laplace transform of the time of ruin in the classical Cramér–Lundberg model with claims having an Erlang distribution of shape parameter 2. We derive an integro-differential equation with two boundary conditions. For every positive discount rate, we show that its characteristic polynomial has one negative root and two distinct positive roots. Boundedness excludes the growing exponential term and yields a representation of the transform as a sum of two decaying exponentials. We construct a finite-dimensional matrix reduction of the convolution chain. For a model parameter set, we obtain exact expressions for the ruin probability, the mean time of ruin, and higher moments, including the third moment at zero initial surplus. We prove analyticity in the discount rate and justify the resulting moment series. We consider the analytic continuation of the representation to negative discount rates separately. For Erlang claims with an arbitrary positive integer shape parameter, we establish a closed linear system and the asymptotics of the root continuing the zero root at zero discount.
manuscript No. (will be inserted by the editor)
Mathematics Subject Classification (2020) 91G05 · 60G51 · 45J05 · 34A30 · 60G40
Introduction
Let be the first time the surplus of an insurance company enters the negative half-line. The object of study is the Laplace transform of the time of ruin
At this transform coincides with the ruin probability , and its derivatives with respect to the discount rate are related to the moments of the time of ruin. Such transforms belong to the Gerber–Shiu framework of the time value of ruin [5, 6]. Sundt and Teugels [12] studied the model in which the surplus earns interest at a constant force of interest.
For claims with a rational Laplace transform (in particular, for Erlang-distributed claims), the representation of in terms of the roots of the Lundberg equation was obtained by Gerber and Shiu [5]. The defective renewal equations arising in such problems were studied in [8], and recurrence relations for the moments of the time of ruin were obtained in [9]. Here the same case is treated by elementary means. The additions are exact rational expressions for the mean and the higher moments in the model case, analyticity in the discount rate with an explicit radius of the moment series, and a finite-dimensional matrix reduction of the convolution chain with an explicit spectrum.
For Erlang-distributed claims, the integro-differential equation (IDE) admits a reduction to an equation with constant coefficients. The reduction of an IDE to an ordinary differential equation (ODE) is also used in risk models with investment [1, 2, 3, 7]; unlike the case considered here, the equations arising there have variable coefficients. Promyslov [10, 11] also studies the construction and the asymptotics of the solutions of such IDEs. We focus on the case of claims with the Erlang distribution (shape 2, rate ), in which an equation of the third order arises. For it one must write out the characteristic polynomial and then establish which of its roots are admitted by the boundary conditions and by the boundedness of the solution.
We prove that for every the characteristic polynomial has one negative and two distinct positive roots, and as the negative root continues the zero root of the limiting problem. On this basis we obtain the representation of the transform as a sum of two exponentials. For the model parameter set we find the exact ruin probability and the moments of the time of ruin. We additionally construct the matrix reduction of the convolution chain and the analytic expansion in , and for claims with the distribution we obtain a closed linear system.
§ 2 derives the equation and the boundary conditions, § 3 studies the roots, and § 4 provides the exact solution and the matrix reduction. § 5 treats the moments, § 6 treats the numerical illustrations, and § 7 and § 8 treat the general case of and the properties of the transform in the parameter .
The equation for the Laplace transform and the boundary relations
The surplus process is , where is a compound Poisson process of intensity , the jump sizes have the density and the survival function
and the net profit condition holds. We extend to by setting it equal to 1, since ruin has already occurred there.
Proposition 1 (the equation and the boundary relations). For the function belongs to , tends to zero as , and satisfies
whence at
Proof. Conditioning on the first claim time , which is exponentially distributed with parameter , gives for a bounded measurable the integral equation
where , and the contribution of the region equals thanks to the extension by 1 for . The right-hand side is continuous in on , hence so is . The convolution with belongs to and its derivative is ; therefore the right-hand side is continuously differentiable, and differentiating in we obtain (3) on together with the inclusion . Differentiating (3) repeatedly, we obtain in turn and , the derivatives at zero being understood as one-sided. The first relation in (4) is the identity (3) at . The derivatives and vanish at since , whence the second relation. Finally, convergence to zero as follows from : almost surely , and therefore .
Remark 1. The second relation in (4) uses the condition in an essential way. For exponential claims () it takes the form , because and , and the linear system for the constants of integration changes. This case is checked in Section 6.
The characteristic equation and the signs of its roots
Lemma 1 (raising the order). The operator , , turns (3) into the linear ordinary differential equation
because and . Conversely, if satisfies (5), then the residual lies in , and the values and coincide with the left-hand sides of (4), since , and . Therefore a solution of (5) satisfies (3) if and only if both relations (4) hold.
Proof. From (2): and . For direct differentiation gives , , and the factor accounts for . The kernel of is two-dimensional, as stated.
Substituting into (5) gives the characteristic polynomial
Theorem 1 (the roots of ). Let . (i) At the polynomial has the root and two distinct positive roots . The term corresponding to the zero root is excluded by the condition ; therefore . (ii) For each the polynomial has exactly one negative root and two distinct positive roots , and
The corresponding exponential term grows, therefore its coefficient vanishes by the boundedness condition, and
Proof. (i) We have , , and as . Hence one nonzero root lies in and the other in . Together with these are the three roots of the cubic polynomial.
(ii) To prove the uniqueness of the negative root, we put and consider
Then and as . Differentiating gives
the second factor decreases linearly and has at most one zero, while . Hence changes sign at most once, and after that the sign remains negative. ends_mid=1 After its maximum decreases strictly, so the equation has exactly one positive root , and .
Next, , and as . Therefore there exist roots and . Together with these are the three roots, and all of them are simple. The growing exponential term is excluded by the boundedness of , and the two remaining ones give (8).
Formula (7) follows from the implicit function theorem:
where and at . We note that the root count in (ii) does not use the condition ; the latter is needed only for (i) and (7).
Corollary 1. The two-exponential representation of the solution (8) holds for all , and the exponents are the two positive roots of (6). The third exponential term grows for and does not enter the bounded solution. For example, for , and the roots are , , to the digits shown.
Remark 2. The growing exponential term does not obstruct the computation of via the truncated boundary value problem with the condition (the truncation error has order ), but it makes the direct computation of the profile of the solution unstable for large . Therefore the closed form (8) with the two relations (4) is preferable for computations.
The exact solution of the model case
From now on , , , . Then , , . The polynomial has the roots , , .
Theorem 2 For this parameter set
and follows from the boundary conditions and is not imposed from outside.
Proof. The constants , solve the linear system (4) at . The first relation gives , that is, . The second gives , that is, ; hence , and then , . To check the original equation (3) we use the exact convolution of an exponential term:
The coefficients of vanish after the substitution because of . The remaining terms have the form , where
For the constants found above, and . Hence the residual of the original IDE vanishes identically, including all tail terms of the truncated convolution. Finally, . ⊓⊔
We now turn to an independent matrix reduction of the convolution chain.
Lemma 2 (the convolution chain). For a locally integrable put
Then , , , , , . At the quantities vanish and .
Proof. Differentiation with respect to the upper limit gives . For it is more convenient to change the variable : , whence
(the boundary term is zero). The relations for and and the representation are immediate, and follows from . ⊓⊔
Put .
Proposition 2 (the closed linear system). The vector satisfies the system with constant coefficients
and , where . The value is determined by the boundedness of the solution and by the condition as . For the latter condition follows already from boundedness, because the third exponential term grows. In both cases one uses the condition for a left eigenvector of the matrix corresponding to the eigenvalue . In a numerical computation on it is replaced by the truncated condition , with an error of order .
Proof. Substituting and into (3) by Lemma 2 gives the first row of (11); the remaining rows are definitions. The matrix is constant, hence , and the truncated boundary conditions are linear in ; hence the system has size 5. The condition is equivalent to the vanishing of the coefficient of the exponential term . The eigenvalues of are (of multiplicity two) and , where are the roots of (6). For one of them is positive, which produces the instability discussed after Corollary 1.
The mean time of ruin and resonance
Denote
If the mean is finite, differentiating (3) with respect to at gives
and from the relations (4) it follows that
Theorem 3 (the identity for the mean time of ruin). Let and let be integrable on . Then
Proof. Integrating (3) at (the function is integrable by Theorem 1(i)) and using Fubini’s theorem together with gives , that is, . Next we integrate (12). We have and
Therefore the terms with cancel and we are left with . The conditional mean is obtained by dividing by . In the model example the integrability of follows from the exact formula of the next theorem.
Theorem 4 (the resonant structure of in the model case) For , ,
and and .
Proof The finiteness of the mean follows from the analyticity established in Proposition 3 by the argument given in the proof of Theorem 5; therefore (12) applies. We apply to it the operator . Unlike the substitution of a single exponential directly into , this step accounts for the truncated convolution completely. We obtain
Both exponents on the right are positive roots of ; therefore the particular solution carries a factor . The condition excludes the constant term, and the bounded solution has the form
The substitution into the resulting ODE gives , , after which the boundary conditions (13) give , . This yields (14). Substituting back into the original equation (12) with the exact convolution (10) and its -derivative (for the terms ) shows that both the coefficients of and the terms of the form vanish. Hence no extraneous solutions were introduced by raising the order. Finally, and
so that agrees with Theorem 3. Division by gives the conditional mean . ⊓⊔
Theorem 5 (the hierarchy of moments) For claims with the Erlang distribution of shape parameter , all the moments
are finite. Put . Then for we have
Proof. The two-exponential representation (8) extends analytically in to a neighborhood of zero; this is justified in Proposition 3, whose proof does not rely on the results of this section. The finiteness of the derivatives of the continuation at zero proves in turn the finiteness of all moments: if, the previous moments being already finite, the next moment were infinite, the corresponding difference quotient as would also diverge by monotone convergence, which contradicts the differentiability of the continuation. Therefore differentiation in under the expectation sign is allowed.
Differentiating the representation (8) also shows that is a finite sum of terms with and positive . Hence is integrable and . Further,
while the density and the tail do not depend on . Since , differentiating (3) gives the first statement. The integration over with Fubini’s theorem, as in the proof of Theorem 3, gives the second. ◻
Corollary 2 (the second moment and the variance in the model case) From (14) we have ; hence
The subsequent moments are computed in the same way: first one solves the equation (a resonant ODE with terms ), and then integrates .
Remark 3. The natural guess fails for . Denoting the right-hand side by , from we see that the identity would require the invariance of with respect to the convolution with . Table 2 shows that the relative discrepancy grows with . The integral identity of Theorem 3 is not a pointwise identity in .
| by (14) | ratio | ||
| 0.0 | 0.04166667 | 0.04166667 | 1.0000 |
| 1.0 | 0.01715030 | 0.00986602 | 0.5753 |
| 2.0 | 0.00579443 | 0.00221183 | 0.3817 |
| 5.0 | 0.00013071 | 0.00002458 | 0.1881 |
| 10.0 | 0.00000013 | 0.00000001 | 0.1019 |
Table 2. The mean time of ruin by the exact formula (14) and the naive continuation .
Numerical illustrations
The values below are computed from the characteristic equation and the two-exponential representation (8). They illustrate the proved formulas rather than replace their derivation.
Table 1 agrees with Theorem 1: the root is negative for , and its linear estimate is most accurate for small . The matrix boundary value problem from Proposition 2 reproduces the values within the printed digits. In a direct computation of the profile through , however, the growing component may accumulate an error near the right edge of the truncated domain; therefore the accuracy of the value at the initial point should not be carried over to the whole profile without a separate estimate.
| 0.00 | 0.000000 | 0.000000 | 1.500000 | 4.000000 | 0.40000000 | 0.3333333333 |
| 0.05 | −0.012461 | −0.012500 | 1.504960 | 3.999168 | 0.39819879 | 0.3312678779 |
| 0.10 | −0.024847 | −0.025000 | 1.509844 | 3.998337 | 0.39642779 | 0.3292373739 |
| 0.20 | −0.049403 | −0.050000 | 1.519388 | 3.996682 | 0.39297243 | 0.3252766458 |
| 0.50 | −0.121499 | −0.125000 | 1.546406 | 3.991759 | 0.38323597 | 0.3141241706 |
| 1.00 | −0.237275 | −0.250000 | 1.586909 | 3.983700 | 0.36876893 | 0.2975812980 |
| 2.00 | −0.456733 | −0.500000 | 1.655298 | 3.968101 | 0.34470970 | 0.2701787511 |
| 5.00 | −1.060387 | −1.250000 | 1.801974 | 3.925079 | 0.29482670 | 0.2141231862 |
Table 1. The roots of the characteristic equation (6) and the constants (8) for , , . The third column is the linear estimate (7).
For the test set , we have , the positive roots 1.078835 and 4.171165, , . Hence , , and .
For exponential claims (, , ) the characteristic equation has the roots 0 and , and the solution makes the residual of the equation vanish identically. Here and the conditional mean equals 1. The second boundary relation from Remark 1 becomes the equality , whereas the form (4) for would give here the wrong equality .
Finally, the slope of at zero equals in the main model case. This gives an independent consistency check of Table 1 against Theorem 4 without finite-difference estimates of the derivative.
The case of claims with the distribution
The constructions of Proposition 2 and § 2 extend to the claims . Put , a polynomial of degree $k + 1.ats
Theorem 6 (the finite-dimensional system and the root shift for general ). For claims with the distribution, the system
with the initial conditions , , (), is closed and has dimension ; one has and . The characteristic polynomial of its matrix is
that is, the spectrum consists of of multiplicity and the numbers , where are the roots of . At the number is a root of , and for this root moves into the negative half-line,
For a fixed value of the formula is universal, but through the shift depends on .
Proof. For differentiation with respect to the upper limit gives and for , and gives the second recurrence. The identities and follow from the definitions of and . The matrix of the system has a block structure: the - and -blocks are lower bidiagonal matrices with diagonal , the first row couples them with , and the expansion of the determinant along the first row gives (16). For (17) we use , and . The implicit function theorem gives , as claimed. □
Table 3 checks (15): for , the matrix method reproduces the classical value up to a term of order , where is the smallest positive root of , and confirms the estimate (17).
The deviation at is explained by the truncation of the computation domain: the smallest positive root of is , hence the condition is imposed with an error of order . At the right-hand side of (17) differs from the exact root by . The reason is not a slow convergence: at the root coalesces with the smallest positive root of ; therefore lies outside the disk of convergence of the series in , and (17) has here only an asymptotic meaning (at the same formula gives of Proposition 4).
| discrepancy | by (17) | ||||
| 1 | 0.166666666667 | 0.1666666667 | −0.009993 | −0.010000 | |
| 2 | 0.333333333333 | 0.3333333333 | −0.012461 | −0.012500 | |
| 3 | 0.500000000000 | 0.5000000000 | −0.016487 | −0.016667 | |
| 4 | 0.666666666665 | 0.6666666667 | −0.024051 | −0.025000 | |
| 5 | 0.833328976420 | 0.8333333333 | −0.041614 | −0.050000 |
Table 3. The values (the matrix method, ) and the root shift at for , .
Remark 4. For all the roots of are real, and the transform is a combination of exponentials without oscillations. Already at and , in the example considered, at the complex conjugate pairs and appear, respectively. Therefore the result of Theorem 1 on the two positive roots applies specifically to . For Rouché’s theorem, applied to the equation () in the half-plane , gives for exactly one root with and exactly roots with , among which complex conjugate pairs appear (damped oscillations). The matrix reduction (15) remains valid.
Properties of the Laplace transform and the expansion in powers of $\delta`
The properties of as a function of follow from the definition (1) itself and do not depend on the method of computation. Recall that a function is called completely monotone on if it is infinitely differentiable and for all and all .
Lemma 3 (completely monotone functions) The function is infinitely differentiable and completely monotone on :
In particular, decreases, is convex, and lies below its value at .
Proof By (1) we have , where is the positive measure of the distribution of the time of ruin on the set . That the Laplace transform of a positive measure is a completely monotone function is checked by differentiating under the integral sign (the converse is Bernstein’s theorem, [4 Ch. XIII, § 4]). For a fixed pick . Then , and the integral is finite. This justifies the differentiation under the integral sign. □
Corollary 3 In the model case , , for all the following estimates hold:
The second estimate follows from convexity; for it is trivial.
We turn to the analyticity at zero. Put . The substitution of the two-exponential representation (8) into the boundary system (4) gives
and hence
Proposition 3 (analyticity of the two-exponential representation) The branches , the coefficients and the function , for each fixed , are analytic in in a neighborhood of zero.
Proof The positive roots and are simple and distinct by Theorem 1; therefore the implicit function theorem in the complex domain gives analytic branches and , which remain distinct in a small neighborhood of zero. Moreover, on the corresponding root: otherwise, from it would follow that . Therefore the denominators in (18) and (19) do not vanish, which proves the statement.
For negative inside the domain of analyticity, the formulas (8)–(19) define the continuation of the branches selected at . For boundedness no longer excludes the third exponential term; therefore the identification of this continuation with the expectation (1) requires a separate justification; at it is given after Proposition 4 (at the argument is the same).
Theorem 5 and Proposition 3 give the moment series
and in particular, in the model case,
Proposition 4 In the model case the singularity of the constructed analytic continuation of the function closest to zero lies at , where
Numerically and .
Proof. A singularity of the formulas (18)–(19) can arise only from a coalescence of branches, since does not vanish at the root. Hence the equations must hold simultaneously. From and , at we obtain and . The real solution gives and . The case is impossible, since . The two remaining complex conjugate solutions are obtained by replacing with the other roots of the cubic equation; the modulus of the corresponding equals . As we move from zero to the left, at the real singularity the branch coalesces with the branch continuing the zero root, while the branch remains simple.
The point found is a genuine singularity, not a removable one. There and , hence
If is the right-hand side of (19), then direct differentiation gives at the critical point . Hence the branch point of square-root type survives.
For illustration, the values of the analytic continuation itself are , , , and to the digits shown. For this continuation coincides with : by Theorem 5 the coefficients of the moment series equal , the series converges in the disk , and it remains to apply Tonelli’s theorem. For the corresponding exponential moment is infinite, so that is the rate of exponential decay of the distribution of on .
Proposition 5. In the model case, exactly,
Proof. We expand both positive roots from (6) to the third order in and substitute the two decaying exponential terms from (8) into the boundary system (4). Solving for gives
The comparison with (20) shows that . The conditional moment is obtained by dividing by . The numerical values of Table 4 agree with this exact result.
| 0.005 | 0.0100411 |
| 0.01 | 0.0100272 |
| 0.02 | 0.0099995 |
| 0.03 | 0.0099721 |
| 0.05 | 0.0099177 |
Table 4. Extraction of the third moment: the quantities approach linearly in .
Remark 5. The approximation by the terms up to the second order in (20) has a cubic remainder. Indeed, the difference between and at equals respectively , , , to the digits shown. The division of these values by gives quantities close to .
Conclusion and open questions
For claims with the Erlang distribution of shape parameter 2, we obtain a closed description of the Laplace transform of the time of ruin: the two admissible exponential terms are determined by the characteristic equation and the boundary conditions. This makes it possible to compute both the ruin probability and the moments of the time of ruin, keeping the full truncated convolution in all the checks.
For the general distribution a finite-dimensional system is constructed; the structure of the roots, however, is more complicated: in addition to real exponential terms, complex conjugate pairs are possible, and the explicit formulas for the coefficients of the decaying terms require a separate analysis. Extensions to models with proportional investment and two-sided jumps, where the variable coefficients or additional unbounded terms require a different boundary value setting, are separate directions of further work.
Declarations
Conflict of interest. The author declares no conflict of interest.
Availability of data and materials. No datasets were generated or analyzed during the current study.
References
- [1]Belkina, T.A., Konyukhova, N.B., Kurochkin, S.V.: Singular boundary value problem for the integrodifferential equation in an insurance model with stochastic premiums: analysis and numerical solution. Comput. Math. Math. Phys. 52, 1384–1416 (2012)
- [2]Belkina, T.A., Konyukhova, N.B., Kurochkin, S.V.: Singular initial-value and boundary-value problems for integrodifferential equations in dynamical insurance models with investments. J. Math. Sci. 218, 369–394 (2016). Transl. from Sovrem. Mat. Fundam. Napravl. 53, 5–29 (2014)DOI
- [3]Belkina, T.A., Konyukhova, N.B., Kurochkin, S.V.: Dynamical insurance models with investment: constrained singular problems for integrodifferential equations. Comput. Math. Math. Phys. 56, 43–92 (2016)DOI
- [4]Feller, W.: An Introduction to Probability Theory and Its Applications. Vol. II, 2nd ed. Wiley, New York (1971)
- [5]Gerber, H.U., Shiu, E.S.W.: On the time value of ruin. N. Am. Actuar. J. 2, 48–72 (1998)DOI
- [6]Gerber, H.U., Shiu, E.S.W.: The time value of ruin in a Sparre Andersen model. N. Am. Actuar. J. 9, 49–69 (2005)DOI
- [7]Kabanov, Yu., Pukhlakov, N.: Ruin probabilities with investments: smoothness, integro-differential and ordinary differential equations, asymptotic behavior. J. Appl. Probab. 59, 556–570 (2022)DOI
- [8]Lin, X.S., Willmot, G.E.: Analysis of a defective renewal equation arising in ruin theory. Insur. Math. Econ. 25, 63–84 (1999)DOI
- [9]Lin, X.S., Willmot, G.E.: The moments of the time of ruin, the surplus before ruin, and the deficit at ruin. Insur. Math. Econ. 27, 19–44 (2000)DOI
- [10]Promyslov, P.: Existence of a classical solution to the integro-differential equation arising in the Cramér–Lundberg non-life insurance model with proportional investment. arXiv:2604.05143 (2026)arxiv.org/abs/2604.05143
- [11]Promyslov, P.: On the integro-differential equation arising in the ruin problem for annuity payment models. arXiv:2601.01447 (2026)arxiv.org/abs/2601.01447
- [12]Sundt, B., Teugels, J.L.: Ruin estimates under interest force. Insur. Math. Econ. 16, 7–22 (1995)DOI