2. Settler reputation for stock clearing performance (B4, C4) These patterns are interesting because they describe settlers' reputations for stock clearing performance which are completely unrelated to actual performance. It is a fairly common finding of multivariate analysis that ratings of performance often contain a reputational component which is not related to the performance being rated. Notice that two geographic patterns emerged from sample B and one pattern emerged from sample C. None of these patterns were related in any way to the time required for stock clearing. These patterns are interesting because broker/dealers rating transfer agent promptness in transferring title have indicated that securities with a small number of shares outstanding are most likely to get prompt title transfer. 10. (Unidentified settler's effects) (B11, C11, C13) One pattern from sample B and two patterns from sample C clearly describe settler characteristics but have no more specific identifiable meaning. They are essentially unidentified patterns describing settlers. It is important to note with care the ten influences which have no effects upon the time required for trade to be cleared in the stock clearing system as it was operating in August-September 1968. It may be reasonable to assume that recent trading volume, transfer agent promptness in transferring title, stock splits and tender offers should all have some effects on the time required for a trade to be cleared. It also may be reasonable to expect that trades being cleared over large geographic distances should take longer to be settled than trades being cleared within the same town. In fact, our data failed to support these reasonable hypotheses. The stock clearing system as it was performing in August-September 1968 was operating more slowly than the traditional industry standard for reasons other than these, and therefore the potential contribution made by influences like large geographic distances had no observable effects upon its performance. REVIEW OF EFFECTS OF STOCK CLEARING HOUSE Having examined the effects of a stock clearing house upon fails by the foregoing multivariate approach - including in that work buys from around the country and from the New York, American, and over-the-counter markets – we chose to focus upon the effects of a stock clearing house on fails in over-thecounter securities traded within New York City. We selected from our data only those trades in over-thecounter securities which were completed entirely within New York City. That is, the buying broker and the selling-settling broker were both located in New York City if direct clearing was used. Both the buying broker and the settling broker were located in New York City if a stock clearing house was used for clearing. We developed two samples for this analysis so that we could observe directly the stability of the results from two samples (see the descriptions of Samples I and J in Appendix B). The comparative performance of direct stock clearing and stock clearing through a clearing house for settlements between parties in New York City is shown in Table 9. It can be seen that use of a stock clearing house usually but not always — results in a slightly higher percentage of trades, dollars, and shares being cleared on settlement day. It can also be seen that both clearing systems perform well below the industry standard since both permit failures in settlement of from 60% to 80% of the trades-dollars-shares. Several things must be remembered in interpreting these findings. First, the methods used by the stock clearing house for clearing over-the-counter securities are those common for stock clearing in New York City in 1968; that is, a net TABLE 9 COMPARISON OF DIRECT CLEARING AND CLEARING THROUGH A (showing only over-the-counter trades) share balance for each member of the stock clearing house is drawn for each security at the close of each business day. Members' share balances are not carried forward from day to day. Also, members complete the exchange of certificates among themselves without involving the stock clearing house in any way except in the creation of the net share balance order each day. (Money is balanced through the stock clearing house.) Second, this comparison of the stock clearing performance for direct-clearing and house-clearing controls for geographic distance and general type of security, but it fails to control for characteristics of the parties to the trade which have been shown in our multivariate analysis to be related in an important way to stock clearing performance. These results show rather little advantage to house-clearing when compared with direct-clearing of over-the-counter securities within New York City. However, the findings need to be treated as provocative, rather than definitive, since they are tied to a particular house-clearing method (the daily net balance method without carrying forward prior balances), and since certain influences known to affect stock clearing performance have not been controlled by the analytical method used to develop these findings. USE OF FINDINGS ABOUT CAUSES OF FAILS These findings can be used to form guidelines for judging the likely usefulness of suggested approaches to improving the performance of the stock clearing system. The next chapter discusses the implications of these findings for future stock clearing. VI. IMPLICATIONS FOR FUTURE CLEARING This chapter reviews the need for improvement in the stock clearing system, considers the accuracy of knowledge about causes of fails, and suggests guidelines for judging proposed remedies. THE NEED FOR IMPROVEMENT IN STOCK CLEARING The stock clearing system in the United States is operating less promptly than the standard generally accepted in the industry suggests that it should. While it is the intent of the industry that all trades should be cleared on settlement day, actually about 65% of the trades on the New York Exchange, about 45% of the trades on the American Stock Exchange, and about 20% of the trades in the over-the-counter market are cleared on settlement day. The total dollar value of trades in a fail state on any day in August-September 1968 is estimated at about 4 billion dollars with fails in over-the-counter trades accounting for as much as 70% of that total dollar value. The industry, and especially the over-the-counter segment of the industry, must either bring its stock clearing performance into line with its self imposed and publicly accepted standards of performance or it must take steps to modify those standards by increasing the time between trade date and settlement day. Since an extension of the time permitted for settlement puts the stock clearing process under increased pressures from stock price movements, and since delayed settlement increases in a variety of ways the costs of operating the trading-clearing-dividend systems for the corporations, banks, stock brokers, and customers, the industry preference probably is to maintain or even shorten the industry standard of five business days between trade date and settlement day. ACCURACY OF INDUSTRY KNOWLEDGE ABOUT CAUSES OF FAILS Industry views about the causes of fails were checked by means of a survey of buys made in six geographical locations in the United States during August and September 1968. The industry opinions about the causes of fails and the survey outcomes describing causes of fails are compared in Table 10. Of the seven causes for fails suggested by knowledgeable observers in the securities industry, three emerged from the survey as having no effects and, therefore, were rejected as causes of fails in the stock clearing system as it operated in the latter part of 1968. The inventory hypothesis was only partially confirmed, with several of the specific causes in this general hypothesis being shown to have no effects on the time required for stock clearing. While the specific hypothesis that broker/dealer clerical performance as a cause of fails was not confirmed, the implied hypothesis that broker/dealer stock clearing performance affects fails was confirmed. The industry opinion that the buying broker/dealer's risky financial status would act |