# The fixed points of the multivariate smoothing transform

## Abstract

Let $$(\mathbf {T}_1, \mathbf {T}_2, \ldots )$$ be a sequence of random $$d\times d$$ matrices with nonnegative entries, and let Q be a random vector with nonnegative entries. Consider random vectors $$X$$ with nonnegative entries, satisfying

\begin{aligned} X\mathop {=}\limits ^{{\mathcal {L}}}\sum _{i \ge 1} \mathbf {T}_i X_i + Q, \end{aligned}
(*)

where $$\mathop {=}\limits ^{{\mathcal {L}}}$$ denotes equality of the corresponding laws, $$(X_i)_{i \ge 1}$$ are i.i.d. copies of $$X$$ and independent of $$(Q, \mathbf {T}_1, \mathbf {T}_2, \ldots )$$. For $$d=1$$, this equation, known as fixed point equation of the smoothing transform, has been intensively studied. Under assumptions similar to the one-dimensional case, we obtain a complete characterization of all solutions $$X$$ to (*) in the non-critical case, and existence results in the critical case.

This is a preview of subscription content, access via your institution.

## Abbreviations

$$\mathbf {1}, \tilde{\mathbf {1}}$$ :

$$\mathbf {1}=(1, \ldots ,1)^\top \in {{\mathbb {R}}^d_\ge }$$, $$\tilde{\mathbf {1}}=d^{-1/2} \mathbf {1}\in {\mathbb {S}}_\ge$$

$$\mathop {=}\limits ^{{\mathcal {L}}}$$ :

Same law

$$\mathrm {int}\left( A\right)$$ :

Topological interior of the set $$A$$

$$\left[ \cdot \right] _{v}$$ :

Shift operator in $$\mathbb {V}$$, see (4.1)

$$\alpha$$-Regular:

See Definition 2.1.

$${\mathfrak {B}}_n$$ :

Filtration, $${\mathfrak {B}}_n=\sigma \Bigl ((T(v))_{\left| {v} \right| <n} \Bigr )$$

$${\mathfrak {B}}_{{\mathcal {I}}_{t}^u}$$ :

$${\mathfrak {B}}_{{\mathcal {I}}_{t}^u} := \sigma \left( U(\emptyset ), \{ T(v) \ : \ v \text { has no ancestor in } {\mathcal {I}}_{t}^u \}\right)$$

$$(C)$$ :

Condition imposed on the supp of $$\mu$$, see Definition 1.1

$$D, {D_{L,s}}$$ :

$$D(x) := \frac{1-{\phi }(x)}{H^\alpha (x)}$$, $${D_{L,s}}(u,t) = \frac{1- {\phi }(e^{s+t}u)}{H^\alpha (e^tu) e^{\alpha s}L(e^s)}$$

$$\mathbb {E}^\alpha _u$$ :

Expectation symbol of $$\mathbb {P}^\alpha _u$$, see (3.4).

$$\mathcal {E}_{\alpha ,c}$$ :

Extremal points of $$H_{\alpha ,c}^K$$, see Lemma 10.5

$$H^s$$ :

$$H^s(x) = \int _{{\mathbb {S}}_\ge } \left\langle x,y \right\rangle ^s \, \nu ^{s}(dy)$$ for all $$x \in {{\mathbb {R}}^d_\ge }$$. It is a $$s$$-homogeneous function, i.e. $$H^s(x)=\left| {x} \right| ^s H^s(x/\left| {x} \right| )$$ and satisfies $$(P_*^{s}H^s)(u)=k(s) H^s(u)$$, $$u \in {\mathbb {S}}_\ge$$

$$I_\mu$$ :

$$I_\mu = \{ s \ge 0 \ : \ \mathbb {E}\left\| {\mathbf {M}} \right\| ^s < \infty \}$$

$$\iota (\mathbf{a})$$ :

$$\iota (\mathbf{a}) := \inf \{ x \in {\mathbb {S}}_\ge \ : \ \left| {\mathbf{a}x} \right| \}$$

$${\mathcal {I}}_{t}^u$$ :

Stopping line, $${\mathcal {I}}_{t}^u:=\{ v \in \mathbb {V}\ : \ S^u(v)> t \text { and } S^u(v|k) \le t \, \forall \, k<\left| {v} \right| \}$$

$${\mathcal {J}}_\alpha ^K$$ :

See Definition 9.4

$$k(s)$$ :

Dominant eigenvalue of $$P^{s}$$ and $$P_*^{s}$$ satisfies $$m(s)=\mathbb {E}N k(s)$$

$$K_C$$ :

$$K_C = \left( \min _{y \in C} \min _i y_i \right)$$ for compact $$C \subset \mathrm {int}\left( {\mathbb {S}}_\ge \right)$$

$$\mathbf{L}(v)$$ :

Recursively defined by $$\mathbf{L}(\emptyset )= \mathbf{Id}$$ and $$\mathbf{L}(vi)=\mathbf{L}(v) \mathbf {T}_i(v)$$

$$L$$-$$\alpha$$-elementary:

See Definition 2.1

$$\lambda _\mathbf{a}$$ :

Perron–Frobenius eigenvalue of $$\mathbf{a}\in \mathrm {int}\left( {\mathcal {M}}_\ge \right)$$

$$M(u)$$ :

Disintegration, see Definition 7.2

$${\mathcal {M}}_\ge$$ :

Set $$M(d \times d, {\mathbb {R}}_\ge )$$ of nonnegative $$d\times d$$ matrices

$$m(s)$$ :

$$m(s) = \mathbb {E}N \lim _{n \rightarrow \infty } \left( \mathbb {E}\left\| {\mathbf {M}_{1} \cdots \mathbf {M}_{n}} \right\| ^s \right) ^{1/n}$$, where $$(\mathbf {M}_{i})_{i \in {\mathbb {N}}}$$ are i.i.d. with law $$\mu$$

$$(\mathbf {M}_n)_{n \in {\mathbb {N}}}$$ :

Sequence of random matrices (i.i.d. with law $$\mu$$ under $$\mathbb {P}$$) c.f. (3.4)

$$\mu$$ :

Law on $${\mathcal {M}}_\ge$$, defined by $$\int \! f(\mathbf{a}) \mu (d\mathbf{a}) \!:=\! (\mathbb {E}N)^{-1} \, \mathbb {E}\left( \sum _{i=1}^N f(\mathbf {T}_i)\right) \!.$$

$${\mathbb {N}}, {\mathbb {N}}_0$$ :

$${\mathbb {N}}= {\mathbb {N}}_0\! \setminus \! \{0\}$$.

$$\nu ^{s}$$ :

Probability measure on $${\mathbb {S}}_\ge$$, satisfy $$(P^{s})' \nu ^{s}= k(s) \nu ^{s}$$

$$\mathbb {P}_u$$ :

Notation for initial states, $${\mathbb {P}}_{u} \left( {U(\emptyset )=u, U_0=u, S_0=0} \right) =1$$

$$P^{s}$$, $$P_*^{s}$$ :

Operators on $${\mathcal {C}}^{}\left( {\mathbb {S}}_\ge \right)$$ defined in (1.13) resp. (3.1)

$$\varvec{\Pi }_n$$ :

$$\varvec{\Pi }_n := \mathbf {M}_n^\top \dots \mathbf {M}_1^\top$$

$$\mathbb {P}^\alpha _u$$ :

Exponentially shifted measure, see (3.4).

$${\mathbb {R}}_\ge , {\mathbb {R}}_>$$ :

$${\mathbb {R}}_\ge =[0,\infty )$$, $${\mathbb {R}}_>=(0,\infty )$$

$$\varrho$$ :

Stationary law of $$(U(t), R(t))$$ under $$\mathbb {P}_u^\alpha$$, see Theorem 4.9

$$S^u(v), S_n, S(t)$$ :

$$S(v)= - \log \left| {\mathbf{L}(v)^\top u} \right|$$, $$S_n := - \log \left| {\varvec{\Pi }_n U_0} \right|$$, $$S(t)=S_{\tau _t}$$

$${\mathbb {S}}_\ge$$ :

$${\mathbb {S}}_\ge = {\mathbb {S}}\cap {{\mathbb {R}}^d_\ge }$$ intersection of the unit sphere and the nonnegative cone in $${\mathbb {R}}^d$$

$$\tau _t$$ :

$$\tau _t:= \inf \{n : S_n >t\}$$

$$\mathbb {V}$$ :

$$\mathbb {V}=\bigcup _{n=0}^\infty {\mathbb {N}}^n$$

$$U^u(v), U_n, U(t)$$ :

$$U(v)= \mathbf{L}(v)^\top \cdot u$$, $$U_n= \varvec{\Pi }_n \cdot U_0$$, $$U(t)=U_{\tau _t}$$

$$v_\mathbf{a}$$ :

Perron–Frobenius eigenvalue of $$\mathbf{a}\in \mathrm {int}\left( {\mathcal {M}}_\ge \right)$$

$$W_n, W$$ :

Martingale, see (1.14), $$W_n \rightarrow W$$, $$W_0(u)=H^{\alpha }(u)$$

$$W^*$$ :

Particular fixed point of $${\mathcal {S}}_Q$$, see Lemma 5.2

$$W_{{\mathcal {I}}_{t}^u}^f$$ :

$$W_{{\mathcal {I}}_{t}^u}^f = \sum _{v \in {\mathcal {I}}_{t}^u} H^\alpha (\mathbf{L}(v)^\top U(v)) \, f(U(v), S(v)-t),$$

$$Z(u)$$ :

$$Z(u):=- \log M(u)$$

## References

1. 1.

Aidekon, E., Shi, Z.: The Seneta–Heyde scaling for the branching random walk. Ann. Probab. 42(3), 959–993 (2014)

2. 2.

Alsmeyer, G., Biggins, J., Meiners, M.: The functional equation of the smoothing transform. Ann. Probab. 40(5), 2069–2105 (2012)

3. 3.

Alsmeyer, G., Damek, E., Mentemeier, S.: Precise tail index of fixed points of the two-sided smoothing transform. In: Random Matrices and Iterated Random Functions, Springer Proceedings of Mathematics and Statistics, vol. 53, pp. 229–251. Springer, Heidelberg (2013)

4. 4.

Alsmeyer, G., Meiners, M.: Fixed points of inhomogeneous smoothing transforms. J. Differ. Equ. Appl. 18(8), 1287–1304 (2012)

5. 5.

Alsmeyer, G., Meiners, M.: Fixed points of the smoothing transform: two-sided solutions. Probab. Theory Relat. Fields 155(1–2), 165–199 (2013)

6. 6.

Athreya, K.B.: Change of measures for Markov chains and the $$L \log L$$ theorem for branching processes. Bernoulli 6(2), 323–338 (2000)

7. 7.

Basrak, B., Davis, R.A., Mikosch, T.: A characterization of multivariate regular variation. Ann. Appl. Probab. 12(3), 908–920 (2002)

8. 8.

Bassetti, F., Matthes, D.: Multi-dimensional smoothing transformations: existence, regularity and stability of fixed points. Stoch. Process. Appl. 124(1), 154–198 (2014)

9. 9.

Biggins, J.: Martingale convergence in the branching random walk. J. Appl. Probab. 14(1), 25–37 (1977)

10. 10.

Biggins, J., Kyprianou, A.: Measure change in multitype branching. Adv. Appl. Probab. 36(2), 544–581 (2004)

11. 11.

Biggins, J.D., Kyprianou, A.E.: Seneta-Heyde norming in the branching random walk. Ann. Probab. 25(1), 337–360 (1997)

12. 12.

Biggins, J.D., Kyprianou, A.E.: Fixed points of the smoothing transform: the boundary case. Electron. J. Probab. 10(17), 609–631 (2005)

13. 13.

Blum, M.G.B., François, O., Janson, S.: The mean, variance and limiting distribution of two statistics sensitive to phylogenetic tree balance. Ann. Appl. Probab. 16(4), 2195–2214 (2006)

14. 14.

Boman, J., Lindskog, F.: Support theorems for the Radon transform and Cramér–Wold theorems. J. Theor. Probab. 22(3), 683–710 (2009)

15. 15.

16. 16.

Buraczewski, D.: On tails of fixed points of the smoothing transform in the boundary case. Stoch. Process. Appl. 119(11), 3955–3961 (2009)

17. 17.

Buraczewski, D., Damek, E., Guivarc’h, Y., Mentemeier, S.: On multidimensional Mandelbrot’s cascades. J. Differ. Equ. Appl. 20(11), 1523–1567 (2014)

18. 18.

Buraczewski, D., Damek, E., Mentemeier, S., Mirek, M.: Heavy tailed solutions of multivariate smoothing transforms. Stoch. Process. Appl. 123(6), 1947–1986 (2013)

19. 19.

Buraczewski, D., Damek, E., Zienkiewicz, J.: Precise tail asymptotics of fixed points of the smoothing transform with general weights. Bernoulli (2014) (to appear)

20. 20.

Buraczewski, D., Kolesko, K.: Linear stochastic equations in the critical case. J. Differ. Equ. Appl. 20(2), 188–209 (2014)

21. 21.

Caliebe, A.: Symmetric fixed points of a smoothing transformation. Adv. Appl. Probab. 35(2), 377–394 (2003)

22. 22.

Caliebe, A., Rösler, U.: Fixed points with finite variance of a smoothing transformation. Stoch. Process. Appl. 107(1), 105–129 (2003)

23. 23.

Chevillard, L., Rhodes, R., Vargas, V.: Gaussian multiplicative chaos for symmetric isotropic matrices. J. Stat. Phys. 150(4), 678–703 (2013)

24. 24.

Cohn, H., Jagers, P.: General branching processes in varying environment. Ann. Appl. Probab. 4(1), 184–193 (1994)

25. 25.

Doney, R.A.: A limit theorem for a class of supercritical branching processes. J. Appl. Probab. 9, 707–724 (1972)

26. 26.

Dunford, N., Schwartz, J.T.: Linear Operators, Part I, General Theory. Wiley, New York (1958)

27. 27.

Durrett, R., Liggett, T.M.: Fixed points of the smoothing transformation. Z. Wahrsch. Verw. Gebiete 64(3), 275–301 (1983)

28. 28.

Feller, W.: An Introduction to Probability Theory and its Applications, vol. II, 2nd edn. Wiley, New York (1971)

29. 29.

Goldie, C.M.: Implicit renewal theory and tails of solutions of random equations. Ann. Appl. Probab. 1(1), 126–166 (1991)

30. 30.

Guivarc’h, Y.: Sur une extension de la notion de loi semi-stable. Ann. Inst. H. Poincaré Probab. Statist. 26(2), 261–285 (1990)

31. 31.

Guivarc’h, Y., Le Page, É.: Simplicité de spectres de Lyapounov et propriété d’isolation spectrale pour une famille d’opérateurs de transfert sur l’espace projectif. In: Random walks and geometry, pp. 181–259. Walter de Gruyter GmbH & Co. KG, Berlin (2004)

32. 32.

Guivarc’h, Y., Le Page, É.: Spectral gap properties and asymptotics of stationary measures for affine random walks. ArXiv e-prints. arXiv:1204.6004, April (2012)

33. 33.

Iksanov, A., Meiners, M.: Fixed points of multivariate smoothing transforms with scalar weights. ArXiv e-prints. arXiv:1402.4147, February (2014)

34. 34.

Iksanov, A.M.: Elementary fixed points of the BRW smoothing transforms with infinite number of summands. Stoch. Process. Appl. 114(1), 27–50 (2004)

35. 35.

Iksanov, O., Negadaĭlov, P.: On the supremum of a martingale associated with a branching random walk. Teor. ĭmovīr. Mat. Stat. 74, 44–51 (2006)

36. 36.

Jagers, P.: Branching Processes with Biological Applications. Wiley Series in Probability and Mathematical Statistics-Applied Probability and Statistics, vol. II. Wiley-Interscience, Wiley, London, New York, Sydney (1975)

37. 37.

Jagers, P.: General branching processes as Markov fields. Stoch. Process. Appl. 32(2), 183–212 (1989)

38. 38.

Janson, S.: Functional limit theorems for multitype branching processes and generalized Pólya urns. Stoch. Process. Appl. 110(2), 177–245 (2004)

39. 39.

Jelenković, P., Olvera-Cravioto, M.: Implicit renewal theorem for trees with general weights. Stoch. Process. Appl. 122(9), 3209–3238 (2012)

40. 40.

Jelenković, P., Olvera-Cravioto, M.: Implicit renewal theory and power tails on trees. Adv. Appl. Probab. 44(2), 528–561 (2012)

41. 41.

Kahane, J.-P., Peyrière, J.: Sur certaines martingales de Benoit Mandelbrot. Adv. Math. 22(2), 131–145 (1976)

42. 42.

Kelley, J.L.: General Topology. D. Van Nostrand Company Inc., Toronto, New York, London (1955)

43. 43.

Kesten, H.: Random difference equations and renewal theory for products of random matrices. Acta Math. 131, 207–248 (1973)

44. 44.

Kesten, H.: Renewal theory for functionals of a Markov chain with general state space. Ann. Probab. 2(3), 355–386 (1974)

45. 45.

Knape, M., Neininger, R.: Polya urns via the contraction method. Comb. Probab. Comput. 23, 1148–1186 (2014)

46. 46.

Kolesko, K., Mentemeier, S.: Fixed points of the multivariate smoothing transform: the critical case. ArXiv e-prints. arXiv:1409.7220, September (2014)

47. 47.

Kyprianou, A.E.: Slow variation and uniqueness of solutions to the functional equation in the branching random walk. J. Appl. Probab. 35(4), 795–801 (1998)

48. 48.

Kyprianou, A.E., Rahimzadeh Sani, A.: Martingale convergence and the functional equation in the multi-type branching random walk. Bernoulli 7(4), 593–604 (2001)

49. 49.

Leckey, K., Neininger, R., Szpankowski, W.: Towards more realistic probabilistic models for data structures: the external path length in tries under the Markov model. In: Proceedings of the Twenty-Fourth Annual ACM-SIAM Symposium on Discrete Algorithms, pp. 877–886. SIAM, Philadelphia (2012)

50. 50.

Liu, Q.: Fixed points of a generalized smoothing transformation and applications to the branching random walk. Adv. Appl. Probab. 30(1), 85–112 (1998)

51. 51.

Liu, Q.: On generalized multiplicative cascades. Stoch. Process. Appl. 86(2), 263–286 (2000)

52. 52.

Liu, Q.: Asymptotic properties and absolute continuity of laws stable by random weighted mean. Stoch. Process. Appl. 95(1), 83–107 (2001)

53. 53.

Lyons, R., Pemantle, R.: Random walk in a random environment and first-passage percolation on trees. Ann. Probab. 20(1), 125–136 (1992)

54. 54.

Mandelbrot, B.: Multiplications aléatoires itérées et distributions invariantes par moyenne pondérée aléatoire. C. R. Acad. Sci. Paris Sér. A 278, 289–292 (1974)

55. 55.

Melfi, V.F.: Nonlinear Markov renewal theory with statistical applications. Ann. Probab. 20(2), 753–771 (1992)

56. 56.

Menshikov, M., Petritis, D.: On random walks in random environment on trees and their relationship with multiplicative chaos. In: Mathematics and Computer Science, II (Versailles, 2002), Trends in Mathematics pp. 415–422. Birkhäuser, Basel (2002)

57. 57.

Menshikov, M., Petritis, D., Popov, S.: Bindweeds or random walks in random environments on multiplexed trees and their asymptotics. In: Discrete Random Walks (Paris, 2003), Discrete Mathematics and Theoretical Computer Science Proceedings, AC, pp. 205–216 (electronic). Association of Discrete Mathematics and Theoretical Computer Science, Nancy (2003)

58. 58.

Menshikov, M., Petritis, D., Popov, S.: A note on matrix multiplicative cascades and bindweeds. Markov Process. Relat. Fields 11(1), 37–54 (2005)

59. 59.

Mentemeier, S.: On Multivariate stochastic fixed point equations: the smoothing transform and random difference equations. Ph.D. thesis, Westfälische Wilhelms-Universität Münster (2013)

60. 60.

Mentemeier, S.: The fixed points of the multivariate smoothing transform. ArXiv e-prints. arXiv:1309.0733v2, September (2013)

61. 61.

Mirek, M.: On fixed points of a generalized multidimensional affine recursion. Probab. Theory Relat. Fields 156(3–4), 665–705 (2013)

62. 62.

Neininger, R., Rüschendorf, L.: A general limit theorem for recursive algorithms and combinatorial structures. Ann. Appl. Probab. 14(1), 378–418 (2004)

63. 63.

Neininger, R., Rüschendorf, L.: A survey of multivariate aspects of the contraction method. Discrete Math. Theor. Comput. Sci. 8(1), 31–56 (2006). (electronic)

64. 64.

Nerman, O.: On the convergence of supercritical general (C-M-J) branching processes. Z. Wahrsch. Verw. Gebiete 57(3), 365–395 (1981)

65. 65.

Nolan, J.P.: Stable distributions—models for heavy tailed data. In Progress, Chapter 1 Birkhauser, Boston. http://academic2.american.edu/~jpnolan/ (2012)

66. 66.

Olofsson, P.: Size-biased branching population measures and the multi-type $$x\log x$$ condition. Bernoulli 15(4), 1287–1304 (2009)

67. 67.

Rhodes, R., Vargas, V.: Gaussian multiplicative chaos and applications: a review. Probab. Surv. 11, 315–392 (2014)

68. 68.

Rüschendorf, L.: On stochastic recursive equations of sum and max type. J. Appl. Probab. 43(3), 687–703 (2006)

69. 69.

Stadtmüller, U., Trautner, R.: Tauberian theorems for Laplace transforms in dimension $$D>1$$. J. Reine Angew. Math. 323, 127–138 (1981)

70. 70.

Zolotarev, V.: One-Dimensional Stable Distributions, Translations of Mathematical Monographs, vol. 65. American Mathematical Society, Providence (1986) (Translated from the Russian by H. H. McFaden, Translation edited by Ben Silver)

## Acknowledgments

The major part of this work was done while the author held a position at the Institute of Mathematical Statistics, University of Münster, which was supported by Deutsche Forschungsgemeinschaft via SFB 878. An earlier version of a part of these results has been published within the author’s Ph.D. thesis , written under the supervision of Gerold Alsmeyer, Münster, to whom I’d like to express my gratitude; as well as to Dariusz Buraczewski, Ewa Damek, Yves Guivarc’h, Konrad Kolesko, Daniel Matthes and Matthias Meiners for very helpful discussions on the subject. The author is grateful to the referees for a very careful reading of the paper and many helpful comments, that led to an improvement of the presentation.

## Author information

Authors

### Corresponding author

Correspondence to Sebastian Mentemeier.

Research supported by Deutsche Forschungsgemeinschaft (SFB 878).

## Appendix A: Inequalities for Laplace transforms

### Appendix A: Inequalities for Laplace transforms

If $${\phi }$$ is the LT of a r.v. $${\mathbb {R}}_\ge$$, then $$t^{-1}(1-{\phi }(t))$$ is again a LT of a measure on $${\mathbb {R}}_\ge$$ (see Feller , XIII (2.7)]). Consequently, it is decreasing and thus for all $$t \in {\mathbb {R}}_\ge$$, $$0 < a < 1$$:

\begin{aligned} \frac{1-{\phi }(at)}{at} \ge \frac{1-{\phi }(t)}{t} \quad \Rightarrow 1-{\phi }(at) \ge a (1-{\phi }(t)), \end{aligned}

as well as, for $$b\ge 1$$,

\begin{aligned} 1- {\phi }(bt) \le b (1-{\phi }(t)). \end{aligned}

This proves the first four inequalities in the subsequent lemma:

### Lemma 15.1

Let $${\phi }$$ be the Laplace transform of a distribution on $${{\mathbb {R}}^d_\ge }$$, $$u \in {\mathbb {S}}_\ge$$, $$t \in {\mathbb {R}}_\ge$$ and $$\mathbf{A} \in M(d \times d, {\mathbb {R}}_\ge )$$. Then

\begin{aligned} 1 - {\phi }(atu)&\le 1- {\phi }(tu) \quad \text { for } a < 1, \end{aligned}
(15.1)
\begin{aligned} 1-{\phi }(atu)&\ge a(1-{\phi }(tu)) \quad \text { for } a < 1, \end{aligned}
(15.2)
\begin{aligned} 1- {\phi }(btu)&\ge 1- {\phi }(tu) \quad \text { for } b >1, \end{aligned}
(15.3)
\begin{aligned} 1- {\phi }(btu)&\le b(1-{\phi }(tu)) \quad \text { for } b > 1. \end{aligned}
(15.4)
\begin{aligned} 1- {\phi }(tu)&\le 1- {\phi }(t\mathbf {1}) \end{aligned}
(15.5)
\begin{aligned} 1- {\phi }(t\mathbf{A}u)&\le 1- {\phi }(t\left| {\mathbf{A}u} \right| \mathbf {1}) \le 1- {\phi }(t\left\| {\mathbf{A}} \right\| \mathbf {1}) \end{aligned}
(15.6)
\begin{aligned} 1 - {\phi }(t\mathbf{A}u)&\le \left( \left\| {\mathbf{A}} \right\| \vee 1 \right) (1-{\phi }(t\mathbf {1})) \end{aligned}
(15.7)
\begin{aligned} 1 - {\phi }(tu)&\ge 1 - {\phi }(t(\min _i u_i) \mathbf {1}) \ge (\min _i u_i) (1-{\phi }(t\mathbf {1})) \end{aligned}
(15.8)

### Proof

Let $$Z$$ be a r.v. with LT $${\phi }$$. For all $$u \in {\mathbb {S}}_\ge$$, $$\left\langle u,Z \right\rangle \le \left\langle \mathbf {1},Z \right\rangle$$. Thus

\begin{aligned} 1 - {\phi }(tu)&= \mathbb {E}_{} \left( {1-e^{-t \left\langle u,Z \right\rangle }} \right) = \int _0^\infty t e^{-tr} {\mathbb {P}}_{} \left( {\left\langle u,Z \right\rangle > r} \right) dr \\&\le \int _0^\infty t e^{-tr} {\mathbb {P}}_{} \left( {\left\langle \mathbf {1},Z \right\rangle > r} \right) dr = 1- {\phi }(t\mathbf {1}). \end{aligned}

From (15.5) and (15.1) now (15.6) follows:

\begin{aligned} 1- {\phi }(t\mathbf{A}u)&= 1-{\phi }(t\left| {\mathbf{A}u} \right| \mathbf{A} \cdot u) \mathop {\le }\limits ^{(15.5)} 1-{\phi }(t\left| {\mathbf{A}u} \right| \mathbf {1}) \\&=1-{\phi }(t\frac{\left| {\mathbf{A}u} \right| }{\left\| {\mathbf{A}} \right\| }\left\| {\mathbf{A}} \right\| \mathbf {1}) \mathop {\le }\limits ^{(15.1)} 1- {\phi }(t\left\| {\mathbf{A}} \right\| \mathbf {1}). \end{aligned}

Then (15.7) follows by applying (15.1) resp. (15.4) in (15.6).

In order to prove (15.8), observe that

\begin{aligned} \left\langle u,Z \right\rangle = \sum _{i=1}^d u_i Z_i \ge \min _i u_i \sum _{i=1}^d Z_i = \min _i u_i \left\langle \mathbf {1},Z \right\rangle . \end{aligned}

Then the argument is the same as given for (15.5), with an additional use of (15.2).   $$\square$$

## Rights and permissions

Reprints and Permissions