Hj. Witte, SOME CHARACTERIZATIONS OF EXPONENTIAL OR GEOMETRIC DISTRIBUTIONS IN ANONSTATIONARY RECORD VALUE MODEL, Journal of Applied Probability, 30(2), 1993, pp. 373-381
Let (F(n)}n greater-than-or-equal-to 0 be a sequence of c.d.f. and let
{R(n)}n greater-than-or-equal-to 1 be the sequence of record values i
n a non-stationary record model where after the (n - 1)th record the p
opulation is distributed according to F(n). Then the equidistribution
of the nth population and the record increment R(n) - R(n - 1) (i.e. R
(n) - R(n - 1) approximately F(n)) characterizes F(n). to have an expo
nentially decreasing hazard function. To be more precise F(n) is the e
xponential distribution if the support of R(n - 1) generates a dense s
ubgroup in R and otherwise the entity of all possible solutions can be
obtained in the following way: let for simplicity the above additive
subgroup be Z, any c.d.f. F satisfying F(0) = 0, F(1) < 1 can be chose
n arbitrarily. Setting lambda = - log(1 - F(1)), F(n)(x) = 1 - F(x - [
x])exp(- lambda[x]) for-all x greater-than-or-equal-to 0 is an admissi
ble solution coinciding with F on the interval [0, 1] ([x] denotes the
integer part of x). Simple additional assumptions ensuring that F(n)
is either exponential or geometric are given. Similar results for expo
nential or geometric tail distributions based on the independence of R
(n - 1) and R(n) - R(n - 1) are proved.