Lapas attēli
PDF
ePub

-42

with their direct mail clearing operation. There are no draft costs for the continuous net settlement system, and the other settlement costs were assumed to be negligible.

The costs of incomplete transactions under each of the three alternative OTC clearing possibilities are shown in Table 10. Alternative 1, shown in the first column, is essentially the existing system or benchmark cost. Thirty-five percent of the trades are cleared daily net through the NOTC and the remaining 65 percent are settled directly between brokers.

[merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][subsumed][merged small][ocr errors][merged small][subsumed][subsumed][merged small][merged small][ocr errors][subsumed][ocr errors][subsumed][subsumed][ocr errors]

NOTE: MC mail clearing; DN = daily net; and CNS = continuous net settlement.

[blocks in formation]

Similar results for an expanded version of the existing system are shown in the second column. In this case, as explained earlier, the 100 largest OTC firms have been included in the daily net clearing arrangement, increasing the total number of trades so cleared to 90 percent. The remaining 10 percent continues to be settled directly between brokers.

The third and final alternative is based on complete conversion to the continuous net settlement system of the NCC.

-43

As for all prior analyses, the simulation model was used to generate the level of each kind of incomplete transaction, and these volumes in turn were used to estimate the costs shown in Table 10.

A total of $91 million is shown as the net cost of fails for the existing system. The two largest components are broker-to-broker fails and the settlement cost that consists primarily of the cost of bank drafts. Clearing-related operations department cost is also significant as is the cost of fails-to-deliver to customers.

Expansion of the daily net system, as shown in the table, results in a 29 percent reduction of the net cost of fails with the biggest savings in broker-to-broker fails, clearing-related operations department costs, and other settlement costs. With the improvement in the broker-to-broker picture, there is a related improvement in the level and cost of broker-to-public cash customer fails. Because of the expanded use of the NOTC clearing facility, clearing charges more than double but are still a relatively small part of the total cost.

With the introduction of continuous net settlement the net cost

of fails decreases to $39 million per year. This is a drop of 57 percent when compared to the existing system and of 40 percent when compared to the expanded daily net system. Costs associated with brokerto-broker fails are down to an insignificant level. Costs of brokerto-public cash customer fails are down by 42 percent when compared to the existing system. Settlement costs and clearing-related operations department costs are reduced to zero. These savings are offset in part by the $21 million per year charge for clearing that is levied by the NCC.

These results make it clear that, although substantial gains could be achieved by expanding the daily net clearing system, the continuous net settlement system is to be preferred.

-44

IV. CONCLUSIONS

The broad objective of this study has been the development of methods for analyzing the existing trade completion system and various possible alternative systems. A concomitant task was the evaluation of different combinations of changes in operating practices and pro

cedures.

To achieve this objective, a mathematical simulation model and related cost methodology were developed. Data were collected with which to describe the characteristics of the trade completion process. The model has been used to test some alternative configurations of the system at high volumes and to estimate their impact on costs.

THE SIMULATION MODEL

In spite of early misgivings as to the feasibility of developing a workable model of the trade completion system, such a model has been constructed and its use has been successfully tested in evaluating various proposals for reducing the level of fails, which was chosen as the measure of system performance.

By virtue of the development of the model and the explorations that were made with it, a deeper understanding of the delays and maldistributions that interrupt the flow of stock certificates and result in fails has been obtained. Although the model has proven to function successfully, further work should be done to refine it and to expand its usefulness beyond the alternatives examined in this report. For example, the model might be used to assist in determining whether and to what extent there is a relationship between the characteristics of different firms and the difficulties that they experience in completing transactions.

*

COST METHODOLOGY

In order to judge the relative desirability of alternative systems,

*

For other examples, see R. L. Petruschell, D. J. Dreyfuss, L. E. Knollmeyer, and J. Y. Lu, Reducing Costs of Stock Transactions: A Study

-45

cost estimates were required for use in conjunction with the results of the simulation runs. Estimates were made of (1) the clerical and administrative costs associated with fails and stock loans; (2) the imputed interest cost of fails; and (3) the cost of making alternative system improvements. Some special cost estimates were required for evaluation of alternative clearing systems for OTC securities. these cost estimates are conservative.

All of

A basic clerical and administrative cost of four cents per dollar of broker-to-broker fails to deliver was derived by regression analysis from the 1968 I&E reports for the 386 NYSE reporting members. Clerical costs for other types of fails were assumed to be the same for brokerto-institution fails and half as much for broker-to-cash customer and customer-to-broker fails. Using information obtained from a large stocklending broker, it was estimated that clerical and administrative costs for stock loans averaged $5 per loan.

For the securities industry as a whole, the imputed interest cost related to fails between brokers nets out. There remains, therefore, only the interest on broker-to-customer and customer-to-broker fails to be included in estimates of the interest cost of fails for all brokers. This imputed interest is a very significant factor in evaluating changes to reduce the level of fails between brokers and their customers.

The savings obtainable from any system change must be reduced by the cost of introducing and continuing the improvement. Conservative estimates were made of the cost of the changes considered to be required for alternatives examined. In most cases, the savings exceeded the costs of effecting the improvements by substantial amounts.

COST OF FAILS IN ALTERNATIVE TRADE COMPLETION SYSTEMS

Of the many alternative systems tested in the model, those considered the most significant or interesting were selected for inclusion in this report. All of those chosen have been described in detail in Sec. III and are summarized in Table 11.

of Alternative Trade Completion Systems, Vol. I, Summary of Results, The Rand Corporation, R-552-ST, December 1970, Sec. II.

Table 11

SUMMARY OF COST OF FAILS AND INPUT CHANGES FOR ALTERNATIVE TRADE COMPLETION SYSTEMS

[blocks in formation]

NOTE: Delivery priorities are identified by a four letter group showing order of delivery for broker-to-broker, B;
broker-to-institution, I; broker-to-cash customer, C; and broker repayment of a stock loan, S. Blank data spaces in-
dicate that the input in the given alternative is the same as for the Benchmark case.

The word "All" in the Partial Deliveries column means that partial deliveries were made to brokers and institutional
customers but not to cash customers.

This applies to both cash and institutional customers.

-46

[graphic]
« iepriekšējāTurpināt »