Continuity properties of decomposable probability measures on Euclidean spaces (Q789803)

Let \(\{X_ i\}_ 1\!^{\infty}\) be a sequence of i.i.d. random k- dimensional vectors with non-degenerate distribution and A be a linear operator on \(R^ k\) all of whose eigenvalues have negative real part. The author proves that the distribution of the random vectors \(X_ 1+e^ AX_ 2+...+e^{nA}X_{n+1}+...\) is 1. continuous, 2. either absolutely continuous or singular continuous. It is necessary to note that the statement 1. is the obvious consequence of Lévy's theorem. Let \(\{Y_ i\}_ 1\!^{\infty}\) be a sequence of independent random variables and the series \(Y=Y_ 1+Y_ 2+...+Y_ n+..\). converges. Then \(\sup_{a} P\{Y=a\} = 0\) iff \(\prod^{\infty}_{k=1}\sup_{a}P\{Y_ k=a\}=0\), the proof of which is literally transfered to the multidimensional case [\textit{P. Lévy}, Sur les séries dont les termes sont des variables éventuelles indépendantes. - Stud. Math. 3, 119-155 (1931; Zbl 0003.30301); see also \textit{W. Hengartner}, \textit{R. Theodorescu}, Concentration Functions. (1973; Zbl 0323.60015), chapter 2]. The author's proof of the statement 2. is the same as the proof of the analogous theorem for convergent series \[ \eta_1 + \xi_1 \eta_2 + \cdots + \xi_1\cdots\xi_{n-1} \eta_ n + \cdots \] where \(\{(\xi_ i, \eta_ i)\}_1^\infty\) is a sequence of i.i.d. random vectors, \(\xi_ i \neq 0\) with probability 1 [see the reviewer, Teor. Veroyatn. Primen. 19, 163-168 (1974; Zbl 0321.60053); English translation in Theory Probab. Appl. 19, 163-168 (1974), and Litov. Mat. Sb. 20, No.4, 49-53 (Russian) (1980; Zbl 0456.60031)].