-7

systems, in terms of number and types of incompleted transactions, and to apply the cost methods that had been formulated to estimate the economic consequences.

The first step was to establish one system as a benchmark case against which the other alternative systems could be measured. The existing system, under stress, was chosen as the benchmark case. Those operating methods and policies that were varied in the other altematives included changing delivery priorities and stock loan policies; instituting a fast transfer system; taking full advantage of the partial delivery rule; and reducing DKs, uncompares, and wrong denominations.

Since such clear-cut choices are seldom possible in real-world situations, the next step was to examine the effect of several combinations of changes. In the first phase, those changes that could be accomplished by the brokerage community, acting alone, were examined. The second phase presupposed participation by the banks. Two central depository concepts were examined: one with only broker participation; the second, with full participation of all trade elements. The final analysis compared the existing OTC clearing system with two other clearing possibilities: the first represented an expansion of the present system to include approximately 100 of the largest firms not now members of the National Over-the-Counter (NOTC) clearing facility; the second involved the adoption of continuous net settlement as the industry-wide method of clearing OTC transactions.

Section III presents a complete discussion and description of the alternative trade completion systems and the results of the analyses.

CONCLUSIONS

It should be noted that by far the greatest part of the effort was expended in building tools of analysis. The computer simulation model that was developed for this study is unique. For the first time, it is possible to make quantitative statements about the relative worth of alternative trade completion systems. Of necessity, there was much less time, available for using these methods.

-8

In general, however, the analyses that were made and that are described in Sec. III show that there are a number of modifications to the current trade completion system that could result in significant improvements in industry performance with consequent benefits to the net operating income of the industry. There is no doubt that fails, as defined in this study, cost the industry hundreds of millions of dollars in 1968.

The study also shows that steps may be taken to improve the system without altering the overall structure of the industry or requiring significant new legal action. These steps are as follows:

Reduce transfer time.

Improve communication between banks and between banks and their

customers.

Take full advantage of the existing partial-delivery rule.

O Reduce the number of uncompared trades through a system such as
the AMEX floor-derived clearance.

O Reduce delays due to wrong denomination problems through in-
ventory management.

о Institute formal systems for information on the activities in-
volved in making stock loans.

Other ideas that are either now being implemented, such as the Central Certificate Service (CCS) and nationwide clearing for OTC stocks, or are being actively studied, such as the introduction of machinereadable documents, will also help to alleviate the industry's problems in trade completion. Nonetheless, it is clear that improvements such as those listed in the preceding paragraph will provide substantial economies and efficiencies that are almost as great as those the industry is considering that require major structural changes. Probably more important, the development of a full-scale depository requires certain legal changes that may be a long time in the legislative process. However, the CCS and the planned clearing system for OTC stocks with continuous net settlement will facilitate the implementation of the improvements listed above. The introduction of machine-readable documents requires the development of new procedures and the installation

67-228 O 72 pt. 5 18

-9

of additional equipment, and these too require time. The proposed utopian solution to all of these problems, the abolition of the certificate, requires very extensive legal work and, from the examinations made to date, does not suggest any truly significant additional efficiencies or savings in the trade completion system.

-10

II. COST METHOD FOR THE TRADE COMPLETION SYSTEM

FAILS AS A MEASURE OF PERFORMANCE OF TRADE COMPLETION SYSTEMS

The measure of performance used in this study is the fail, that is, the incomplete transaction. The securities industry has long been concerned with broker-to-broker fails, but they are not adequate as an indicator of the securities industry's success in achieving the goal of completing stock transactions within the established time. The level of broker-to-broker fails, for example, can be reduced by a fails clearance system without improving the rate of delivery to customers. Here, the concept of the fail has been expanded to include not only brokerto-broker fails, but fails involving prepaying customers, COD customers, and fails by brokers to repay a stock loan on call date. Fails regarded in this broader sense are useful indicators of system performance, as an examination of the causes of fails will show.

The primary cause of fails in the real world and in the model is delay in the flow of stock certificates from seller to buyer, i.e., the certificates do not move from seller to buyer within the time allowed for completing trades. Some delays occur because of DKs and uncompares. Other delays result because certificates have been sent to transfer agents for change of title or to be exchanged for certificates of smaller denominations, because of the transmittal of certificates by mail, or because customers fail to surrender stocks within the prescribed time period. When, for any of the foregoing reasons, one broker is unable to deliver to another, the effects of the delays are spread throughout the system. For example, when a customer sells a stock and does not surrender the certificates promptly, it may be impossible for his broker to make delivery to the contra-broker. This in turn can cause the con

tra-broker to be unable to deliver the securities that are due to the purchasing customer.

A secondary cause of fails may be identified as the maldistribution of inventory. The fails due to delays in the flow of certificates can be offset by using certificates in brokers' free boxes. However,

*

*"Free box" refers to the stock certificates on hand that are possessed by a broker and are not subject to segregation requirements.

Such

-11

situations develop in which there is stock, excess to the immediate needs of some brokers, that is not made available to other brokers

who are failing to deliver. Stock loans could make these shares available to those who need them. A general stock loan practice in conjunction with a partial delivery system will permit further improvement so that the maximum number of available certificates is used to fulfill delivery requirements.

There is, as well, a maldistribution of inventory when a broker sends certificates to transfer that could have been made available by stock loan to permit another broker to make a required delivery. This maldistribution also occurs when a broker sends stock to transfer that he will need for delivery within a day or two. In other words, fails occur because stock is sent to transfer unnecessarily.

Fails are useful measures of performance per se and even more so when translated into costs. Carrying open transactions on the books is expensive. The misallocation of capital that results from uncompleted transactions causes the industry to incur imputed interest costs and benefits. Further, costs can be associated directly with actions necessary to reduce the level of fails and thereby raise the level of system performance.

COSTS

The simulation model provides information about the number and type of fails, but a cost-estimating method, to translate fails into system cost, is also required. As indicated previously, three types of costs have been identified as being relevant: the clerical and administrative costs resulting from open or incomplete transactions; the imputed interest costs and benefits from improper allocations of capital following from incomplete transactions; and the costs of improving system perfor

mance.

There are several instances when a cost incurred by one firm will result in an equal benefit to another firm. In effect, these are transfer payments. For example, a firm borrowing stock from another firm

certificates can be pledged, loaned, or delivered in settlement of trades as authorized by the owner.