Testing the random walk hypothesis: evidence for the Budapest stock exchange