“…Seen in this light, Logue's (1991) assertion that soft dollar research does not pass a cost-benefit test is suspect. Burgunder and Hartmann's (1986) assertion that in the absence of Section 28(e) protection managers would continue purchasing research "until the last hard dollar spent for the research equaled the value of that research to the clients" is simply wrong. So, too, is the conclusion that soft dollars necessarily lead managers to purchase too much research.…”