Securities and Futures Markets: Assessment of SEC and CFTC Studies of October 1989 Market Volatilities

Published by the Government Accountability Office on 1990-07-17.

Below is a raw (and likely hideous) rendition of the original report. (PDF)



                   _ __.__.-_-.---.-.----
.Jlll\    I!)!)0
                    SECURITIES AND
                    FUTURES MARKETS
                    Assessment of SEC
                    and CFTC Studies of
                    October 1989 Market
_   _...   .._   .._.   I   I   _...,,   _“ll.-l”...   .   .   l-.*-.-.   ----
             United States
GAO          General Accounting Office
             Washington, D.C. 20548

             General Government Division


             July 17, 1990

             The Honorable Christopher J. Dodd
             Chairman, Subcommittee on Securities
             Committee on Banking, Housing,
               and Urban Affairs
             United States Senate

             Dear Mr. Chairman:

             On May 31,1990, your Subcommittee requested an analysis of the Secu-
             rities and Exchange Commission’s (SEC) and the Commodity Futures
             Trading Commission’s (CFX) reports on the market events of October 13
             and 16,198g.l On June 27,1990, we briefed your Subcommittee on our
             analysis of the differences in methodologies and conclusions of the
             reports. We also discussed how the current debate regarding the SECand
             CFTCroles in regulating stock index futures trading can hamper potential
             solutions to the data problems we identified in the reports. As you
             requested, this report summarizes the contents of our briefing.

Background   experienced extraordinary price volatility, dropping $190 billion in
             value, $160 billion of which was lost in the last 90 minutes of the
             trading day. At 2:40 p.m. on Friday, the New York Stock Exchange
             (NOSE) halted trading in the stock of UAL.2 At 2:55 p.m. the wire service
             reported that financing for a proposed buyout of UAL was in doubt. The
             Standard & Poor’s 500 futures contract fell to its 12-point price limit:’ at
             3:07 p.m. at the Chicago Mercantile Exchange (CME). At 3:16 p.m. the
             Chicago Board Options Exchange (CBOE) stopped stock index options
             trading for the day. At 3:30 p.m. CME lifted the limit on stock index
             futures trading, but prices quickly fell further and at 3:45 p.m. prices
             reached CME’S 30-point price limit. By the close of trading the Dow Jones
             Industrial Average declined by 191 points. The decline continued into
             the opening of the markets on Monday, October 16,1989, when the Dow
             fell an additional 63 points before rebounding and closing up 88 points.

              ’Division of Economic Analysis, Commodity Futures Trading Commission, Report on Stock Index
                                                                   , May 1990; Division of Market Regulation,
                                                                   of October 13 and 16,1989, May 1990.

             ‘UAL is the parent company of United Airlines.
             “Price limits are prearranged market values that, when reached, limit for a specified time period the
             lowest price at which trades can occur.

             Page 1                                 GAO/GGD-90-1OSBR Assessment of SEC and CFTC Studies

                   The price volatility was accompanied by hourly trading volume levels
                   that rivaled those of the 1987 market crash.

                   In response to the large price movements, both SEC and CFTCreviewed
                   the market events of October 1989. The SECand CFTCstudies offered a
                   variety of conclusions on the effects of such automated trading strate-
                   gies as program trading and index arbitrage,l prearranged actions by
                   markets to slow the pace of price declines (“circuit breaker mecha-
                   nisms” or “shock absorbers”), and other issues.

                   For example, SECreported that index arbitrage and other program
                   selling strategies significantly accelerated and exacerbated the market
                   decline. On the other hand, CFTC reported that neither program trading
                   nor futures sales explain the observed price movements on these dates.

                   SECalso reported that, while a direct causal relationship between circuit
                   breakers and a decreased rate of stock market decline is difficult to
                   establish, at the very least, no harm is attributable to the imposition of
                   circuit breaker mechanisms. CFTCreported that the shock absorbers used
                   on October 13 did not appear to have moderated intraday market vola-
                   tility and that when shock absorbers were imposed in one market-but
                   not in others-volatility   increased in the unconstrained markets.

                   SECand cm had additional conclusions on issues such as the effect of
                   trading by various futures traders and the movement of trading from
                   CBOE to CME when CBOE closed early.

                   The SEC and CFTCstudies used the same data to describe the events on
Results in Brief   October 13 and 16, 1989. Neither study purports to identify the under-
                   lying causes of market price volatility or large downward price move-
                   ments. The differences in their conclusions appear to be primarily due to
                   differing points of emphasis and differing points of view. Thus, the
                   reports are most appropriately viewed as useful sources of information
                   for considering issues such as whether and how to control market vola-
                   tility. The reports do not provide definitive answers to these questions,
                   however, because they do not identify the underlying causes of market

                   %ogram trading is the simultaneous trading of a group or basket of stocks. Index arbitrage is the
                   simultaneous, or near simultaneous, sale (or purchase) of stocks that comprise or closely track a stock
                   index and the purchase (or sale) of either futures or options on that particular index to take advan-
                   tage of price differentials between the markets.

                   Page 2                                 GAO/GGD-90-108BR       Assessment of SEC and CFTC Studies

                        Both studies rely on trading data that provide a less than complete and
                        precise reconstruction of the trading that occurred. Legislation has been
                        proposed that would enhance trading data quality, but its progress is
                        being impeded by the jurisdictional debate regarding the SEC and CFIK
                        roles in regulating these markets.

Objectives, Scope,and   gies and conclusions of the SEC and CFTC reports on the market events of
Methodology             October 1989. In arriving at our conclusions, we examined the data
                        quality, analysis, and interpretations in the two reports. We also dis-
                        cussed the methodology and results of the studies with SEC,CFTC,NYSE,
                        CBOE,and CME officials. Our work was done between June 11 and July
                        13, 1990, in accordance with generally accepted government auditing

                            and CFTC used the same trading data in both studies to describe the
Data Limitations Make   SEC
                        events of October 13 and 16. CFTCand SEC staffs worked closely together
Complete and Precise    to ensure that the quantitative data were identical. SEC and CFTCofficials
Trade Reconstruction    also interviewed securities and futures traders both jointly and individ-
                        ually. Both the SECand CFTCstudies relate trade times and trading strat-
Difficult               egies to price movements in the markets in an attempt to link particular
                        types of trading strategies to price movements. However, the informa-
                        tion they used was not a complete and precise reconstruction of who
                        traded what and exactly when. Whether the trading data they used
                        affected their analysis of events is unclear, but complete and precise
                        information could have reduced that possibility and provided a better
                        means to reconstruct trading activity.

                        Problems with the data that were used are varied and complex. As we
                        have reported previously, the times CFTCuses for futures market trades
                        are imputed on the basis of a range of data and may be subject to incor-
                        rect sequencing.” CFTCofficials told us, however, that the accuracy of
                        trade execution times at CME was better than normal because, except for
                        one short period, the continuous decline in prices made it easier to asso-
                        ciate specific trades with unique times. CFTC also has a large-trader
                        reporting system that allows it to reconstruct the bulk of all trading
                        during the day.

                        “Futures Markets: Strengthening Trade Practice Oversight (GAO/GGD-89-120, Sept. 1989).

                        Page 3                              GAO/GGD-90.108BR     Assessment   of SEC and CFTC Studies

We have not yet reviewed information systems in the securities markets.
However, the SEC report indicates several problems in reconstructing
stock trading. SECobtained futures-related equity trading data on pro-
gram trading and index arbitrage strategies from NYSE. This information
was available from firms using these strategies because of recently
imposed NISE rules. The information SEChad on times for these trades is
based upon the time the exchange received the order (order entry time)
rather than the exact time the trade was executed. SEC officials told us,
however, that the markets in October 1989 did not experience the same
order-handling backlog as in October 1987, and that the order entry
times provide good estimates of the actual execution times.

During the course of the day on October 13, program trading and index
arbitrage represented about 11 percent of the sells and 2 percent of the
buys at NBE. Even when prices declined significantly during the after-
noon, SEC'S report indicates that these trades represented at most about
60 percent of the trading volume of Standard & Poor’s 500 stocks and 36
percent of total NYSE volume. Thus, most of the trading during the day
and even the majority at crucial times was unaccounted for by the pro-
gram trading data.

SW   does not have a large-trader reporting system to help it reconstruct
nonfutures-related equity trading. NOSE officials told us that complete
and precise information is available for all stock trades, but the data are
maintained by individual stock, and reconstructing trading for all stocks
is a very time-consuming process. SEC tried to use this information to
reconstruct nonfutures-related equity trading, but reported that because
the information was designed for insider trading and manipulation
investigations in single or small groups of stocks, it was difficult to use
in reconstructing trading across the entire market. SECalso reported that
developing information on trading strategies from these data is ineffi-
cient because major institutional investors trade in so many different
ways through so many different accounts and firms that tracking the
activity is a complex process. Thus, to determine most of the strategies
for nonfutures-related equity trading, SF,Cobtained anecdotal informa-
tion from a sample of large traders.

Page 4                      GAO/GGD-90-108BR   Assessment   of SEC and CFTC Studies
                       The conclusions of both studies focus on a single market event during a
Generalizations        narrow, volatile time period. According to the research literature,
Cannot Be Made From    studies of multiple events over a wider time period would produce con-
the Conclusions of     elusions that better lend themselves to generalizations. Information from
                       such studies, as well as other data to account for the effects of all fac-
Single Event Studies   tors on price movements in markets, would be needed to provide con-
                       vincing support for conclusions about causal relationships between
                       program trading (or other activity) and price volatility.

                       Although the SEC and CFTC reports make informal comparisons to other
                       major market events and to normal trading conditions, for analytical
                       purposes, those other events and time periods are not included with
                       data from October 1989.

                       SECand CFTCemphasize different factors and take different points of
SEC and CFTC           view to reach their conclusions. This sometimes results in different con-
Reports Have           clusions, each of which is supported by basically the same trading data.
Different Points of    As we have discussed, determining the soundness of these conclusions
                       would require better data and a broader scoped study of the issues.
Emphasis and Points
of View                The best example of the different regulatory perspectives is the agen-
                       cies’differing treatment of index arbitrage and other program trading.
                       Their conclusions are consistent with the positions they have each taken
                       since the October 1987 market crash. SECstated that index arbitrage and
                       other program selling accelerated and exacerbated the market decline.
                       This conclusion was based on the association between relatively high
                       levels of index arbitrage and program selling and the times of large
                       market price declines. CFTC, on the other hand, concluded that program
                       trading does not explain the observed price movements. This conclusion
                       was based primarily on two findings: (1) index arbitrage was not
                       extraordinarily large in October, either in absolute terms or in relation
                       to total NISE volume, and (2) although NYSEprogram stock sales were
                       concentrated during periods of greatest price declines, this trading did
                       not account for the predominant share of total stock sales during those
                       intervals, because nonprogram stock sales also increased. Thus, both the
                       SECand CFW conclusions are consistent with the trading data in the
                       reports but involve subjective interpretation and judgment.

                       Page 6                      GA0/GGD90408BR   Assessment of SIX and CFI’C Studies

                     Legislation pending in Congress would significantly improve the trading
DebateOver Roles     data available to SEC and CFTC.The proposed market reform bills (H.R.
Should Not Impede    3667 and S. 648) would provide SEC with specific information on the
Efforts to Improve   trading activity of all large traders. The CFTC reauthorization bill in the
                     Senate (S. 1729) would require futures exchanges to tim e all trades inde-
Trading Data or      pendently, precisely, and completely. These bills have been delayed
ResolveIntermarket   becauseof congressional debate over which agency should have jurisdic-
Problems             tion over stock index futures. Although both securities and futures
                     exchanges are proceeding with plans to improve their data systems, pas-
                     sage of the legislation would emphasize the importance of these issues
                     and ensure continued progress.

                     The jurisdictional debate regarding the agencies’roles may also detract
                     from the good working relationships needed to develop processesto
                     resolve the shifts in selling pressure between markets that apparently
                     occurred in October 1989 becauseof the uncoordinated implementation
                     of circuit breakers between the futures and options markets. After the
                     October 1987 market crash, the President’s W o rking Group on Financial
                     Markets” proposed circuit breakers that could be imposed when large,
                     rapid market declines occur that threaten to create panic conditions. Cir-
                     cuit breakers similar to those proposed were adopted across all markets.
                     Several exchanges subsequently developed other measures to control
                     price movements that have not been coordinated or adopted across all
                     markets. W h e n markets decline by a specific amount, these market-
                     specific measures are imposed temporarily to slow the pace of decline
                     and enable traders to evaluate their positions.

                     Implementation of market-specific measures that had not been coordi-
                     nated between futures and options markets caused problems on
                     October 13, 1989, when price lim its at CME were activated. CBOE, which
                     does not have similar price lim its, closed for the day 9 m inutes after CME
                     reached its first price lim it. CBOE officials told us that they could not
                     keep their market open becauseCME'S price lim it shifted unmanageable
                     selling pressure from the futures to the options market. They added that
                     CBOE did not reopen its market when CME'S price lim it expired, because
                     (1) the futures market continued to fall toward the next CME price lim it
                     and (2) the reopening process would have required more tim e than was
                     available in the remainder of the trading day. CME officials told us that
                     one reason the futures market fell to the second lim it so quickly was

                     “In March 1988 the President created the Working Group on Financial Markets to identify issues,
                     make recommendations,and seek resolution of the complex problems raised by the market crash of
                     October 1087. The Working Grwp is ehaired by the Secretary of the Treasury, and its members are
                     the chairmen of the SEX, CFTC, and Federal ReserveSystem.

                     Page 6                               GAO/GGD-90-108BR     Assessment of SEC and CITC Studies



                           th a t selling pressure th a t w o u l d normally g o to C B O E w a s a ffecting their
                           m a r k e t. For th e rest o f th e a fte r n o o n , th e y a d d e d ,th e fu tures m a r k e t
                           w a s a n o u tle t for trading th a t w o u l d h a v e ta k e n p l a c e o n th e o p tio n s
                           e x c h a n g eif it h a d b e e n o p e n . This selling pressure h e l p e d k e e p th e
                           fu tures m a r k e t a t th e s e c o n dlim it.

                           In o u r reports th a t discuss intermarket issues,w e h a v e e m p h a s i z e dth e
                           n e e d for c o o r d i n a t e d unified r e s p o n s e sto intermarket p r o b l e m s 7 This
                           n e e d c o u l d b e m e t in several ways, including b e tter coordination
                           b e t w e e n e x c h a n g e sa n d regulators, m o r e active participation b y th e
                           P resident’sW o rking G r o u p o n Financial M a r k e ts, a n d /or consolidation
                           o f decision m a k i n g o n intermarket issues in a single b o d y . T h e p r e s e n t
                           d e b a te r e g a r d i n g th e S E C a n d C F T Croles is n o t c o n d u c i v e to th e tim e l y
                           resolution o f intermarket problems.

                           B e c a u s eo f th e short tim e frame within w h i c h y o u r e q u e s te d th e infor-
A g e n cyC o m m e n ts   m a tio n c o n ta i n e d in this report, w e d i d n o t o b ta i n o fficial a g e n c y c o m -
                           m e n ts. W e did, h o w e v e r , o b ta i n informal c o m m e n tsfrom S E CCFTC,  ,       NYSE,
                           C B O E , a n d C M E representatives. T h e y p r o v i d e d technical c o m m e n tsto
                           i m p r o v e th e report’s accuracy, a n d their views h a v e b e e n incorporated
                           w h e r e appropriate.

                           A s a r r a n g e d with th e S u b c o m m i tte e ,unless y o u publicly a n n o u n c eits
                           c o n te n ts earlier, w e p l a n n o further distribution o f this report u n til 3 0
                           d a y s from its issue d a te . A t th a t tim e , w e will p r o v i d e copies o f this
                           report to S E C , CFTC,N E E , C B O E , a n d C M E a s well a s o th e r interested

                           7 S e e for
                                   , e x a m p l e ,C l e a r a n c ea n d SettlementReform: T h e S tock, O p tions,a n d Futures Markets A r e
                           S till at Risk ( G A O W - 9 0 - 3 3 , Apr. 1 9 9 0 ) .

                           Page 7                                        G A O /G G D - 9 0 - 1 0 8 B R Assessment   of S E C a n d C F T C S tudies

The major contributors to this report are listed in appendix I. If you
have additional questions, please call me at 275-8678.

Sincerely yours,

Craig A. Simmons
Director, Financial Institutions
 and Markets Issues

 Page 8                       GAO/GGD-SO-108BR Assessment   of SEC and CFTC Studies
Page 9   GAO/GGD9@108BR   Assessment   of SEC and CFI’C Studies
Appendix I

Major Contributors to This Briefhg Report

                        Michael Burnett, Assistant Director
General Government      Cecile Trop, Assistant Director
Division, Washington,   Patrick Dynes, Senior Evaluator
l-kc!                   John Maurello, Senior Evaluator

                        Dan Johnson, Senior Evaluator
Chicago Regional

  (zaasoo)              Page 10                     GAO/GGD@O-108BR Assessment of SEC and CITC Studiew