U.S.Securities and Exchange CommissionWashington. D.C. 20549 (202) 272.2650

[N]@~~~@~@@)~@

AN AMERICAN PERSPECTIVE ON THE 1987 STOCK MARKET CRASH

Remarks to

10TH ASIAN SECURITIES ANALYSTSCOUNCIL CONFERENCE

sponsored byResearch Institute of Investment Analysts Malaysia

The Kuala Lumpur Stock Exchange

July 19, 1988Shangri-La Hotel

Kuala Lumpur, Malaysia

Charles C. CoxCommissioner

Securities and Exchange Commissionwashington, D.C. 20549

The views expressed herein are those of Commissioner Coxand do not necessarily represent those of the Commission,other Commissioners or the staff.

Good Afternoon,As we are all aware, stock prices throughout the world

fell precipitously last October, in a wave of sellingunparalleled in the experience of most of us here today.Prices have recovered since then and are high by anystandards other than those set in 1987. In some places,they are high by even those standards. In the next halfhour or so, I will offer some remarks on what I believe weshould make of the price plunge, how its causes have beenanalyzed in the united states, and what steps have beentaken to cope with such situations in the future.

I.

First we should be clear about our objectives. Toomuch of the discussion about the stock market, it seems tome, mistakes signs for substance. It is not enough, forinstance, simply to assume that high stock prices are good.They are signs, perhaps, that investors expect good thingsfor the economy, but whether they are intrinsically gooddepends upon whether they lead to an efficient allocationof economic resources. If, for example, they lead toinvestment in dUbious, undeserving or unproductiveenterprises, they may be detrimental from the viewpoint ofsociety as a whole. From the narrower perspective of thesecurities salesman or securities holder, higher usuallymeans better. But even these people may fear the ultimatebursting of a "speCUlative bubble," and so have an interestin seeing that prices bear some relationship to economicfundamentals.

On that basis, I have no reason to conclude that stockprices last August were any "better" than stock pricestoday, and I have little interest in market reforms thatwould inhibit price adjustments, provided that the survivalof the market system itself is not jeopardized by panic andinsolvency feeding further panic and insolvency. Lowprices, per se, are not the enemy. The object of policyshould be an efficient, durable market system and theavoidance of financial panic, not high prices.

II.

What set up the market crash? The long-term magnitUdeof last autumn's decline has been reasonably ascribed tofundamental economic factors. Interest rates, for example,historically have shown a strong inverse relationship tostock prices. However, by October 1987, the ratio of stockdividends to stock prices in the united states, reachedrecord low levels relative to the interest available on

2

u.s. government securities. 1/ But why did the marketcrash, and then rebound somewhat, rather than merely slideto a new equilibrium?

To begin with, the belief that the market wasovervalued was widespread. One survey found that shortlybefore the crash even most buyers believed the market wasovervalued. Presumably they thought they could get outbefore a correction. 21 The post-crash survey may reflectthe respondents' hindsight, but it is nonethelessremarkable.

Some simply blame greed for any overvaluation. Whygreed or even speculation should result in more longpositions -- hoping to ride the trend -- than shortpositions -- anticipating its collapse -- I don't know.What regulators can do to eliminate greed, I also don'tknow.

Others argue that professional money-managers, whoplay an ever-growing role in the market, are particularlysusceptible to a "herd" instinct. The idea is that theyfeel safer making the same mistake everyone else makes,rather than one that highlights their unique incompetence.What we can do about that, once again I don't know.Outlawing pension plan investments or mutual funds seemslike a bad idea.

A third theory rests on the so-called "illusion ofliquidity," a foolish confidence in one's ability to buy orsell quickly at a price very much like the price existingwhen the transaction is ordered. One version of theillusion theory is that many institutional investorsthought that sophisticated trading strategies, usuallyemploying stock index futures, would get them out of stock

1/ See The October 1987 Market Break: A Report by theDivision of Market RegUlation, u.S. Securities andExchange Commission, 3-9 - 3-10 (1988) ("SEC StaffStudy"). See also Report of the Presidential TaskForce on Market Mechanisms, p. 10, Fig. 6: bondyields vs. S&P 500 yield 1947-87 (1988) ("BradyReport") .

21 The survey by Professor Robert Shiller of Yale isdiscussed in Robert E. Norton, "The Battle Over MarketReform," Fortune Magazine (Feb. 1, 1988) pp. 18-19,("Norton Article") and in the National Bureau ofEconomic Research Digest, Jan./Feb. 1988, pp. 1-2("NBER Digest").

3

positions faster than the market fell, and it didn't workwhen too many tried it at once.

In the united states there has been strong interest inthe role played in the crash by derivative products,especially cash-settled futures contracts on stock indices.Essentially, stock index futures are a means oftransferring the potential loss or gain associated withholding the stocks that compose the index, without havingto buy or sell the stocks themselves. Money managers havefound them to be a useful and inexpensive tool for managingrisk in investment portfolios.

The American preoccupation with the role of indexfutures in the crash may seem unusual to many of you whoexperienced stock market crashes in the absence of largefutures markets. It seems a bit curious to me too. But itis argued that the world market crash was triggered or ledby the American crash, which, in turn may have beenaffected by futures products. Futures also have a certainscape-goat appeal, in that they are new and not readilyunderstood by laymen; they are associated with much-maligned "program trading;" and some may feel that iffutures are restricted, the small investor can be persuadedthat things have been "fixed" and lured back into themarket. Futures markets also have caught our attentionbecause they represent the most controversial point in thevarious recommendations for improving the markets.Everyone seems to agree that the market's physical capacityto process transactions should be improved, that clearingand settlement procedures should be better coordinatedamong markets, and that coordination and cooperation amongregulators is important. But when questions are broachedsuch as who should regulate futures or what rules shouldgovern their trading, the battle lines are drawn. So, if Iappear to pay undue attention to futures markets in myremarks today, I hope you will understand.

Now, to return to the illusion of liquidity. Thefutures-related strategy most dependent on liquidity is the"portfolio insurance," or "dynamic hedging" that tries tobeat a market as it falls, selling a pre-set amount foreach incremental drop in the market. The sale ordinarilyis accomplished through a stock index future since this isusually cheaper and quicker than direct sale of a largeamount of stock.

The Brady Commission, appointed by President Reagansoon after the craSh, found that, as of September 30, 1987,pension funds using portfolio insurance were invested 56%in equities, as opposed to 46% for all pension funds. Thatis some evidence of futures-induced over-confidence.

4

still, only eleven of the Brady survey's 80 respondentssaid that they used portfolio insurance. 1/ others report5.5%, although the Brady figures probably provide a betternotion of the percentage of assets covered. i/ Obviously,buyers thought they'd be able to leave the market before itfell -- buyers always think that. What the futures marketsadded to their bravado is another question. As we'venoted, markets outside the United states also plunged inOctober and futures-related strategies are not yet widelyused outside the United states. One might still argue thatfutures helped trigger the great sell-off, but not thatthey were responsible for markets outside the U.s. becomingover-valued to begin with.

In the U.S., the crash itself may have shattered theillusion of liquidity as I've described it. Portfolioinsurers are reported to have been losing customers sinceOctober. a; But I doubt that, beyond a few oversold onportfolio insurance, many really expected the markets to beliquid enough to handle a sudden rush. More likely theyexpected there wouldn't be one, at least not while theywere still in equities. Probably 1987 isn't the first timethat happened, and futures aren't essential to thescenario.

III.Now let me turn from the question of overvaluation to

the dynamics of the price plunge itself. The investorsurvey I mentioned earlier £I suggests quite a few wereready to become sellers at the slightest nudge. Variousevents in mid-october, such as a tax on takeovers proposedin the United States, have been blamed for the nudge. Butprobably the main nudge was the sight of so many other

Brady Report, p. V-IS.NBER Digest, pp. 1-2. The SEC staff estimated that atleast $55 billion worth of stock, mostly pension fundassets, was covered by portfolio insurance on October19. SEC Staff Study at p. 3-4. Prior to the Octoberdecline (including several days before October 19)pension funds held about $740 billion in commonstocks, and all pUblicly owned common stock werevalued at several trillion dollars. Brady Report, p.1.Norton Article, p. 21.See note 1, supra.

5

traders selling. Already in the week before October 19 themarket had fallen dramatically.

Need it have gone so far as it did on the 19th? Thestaff of the SEC -- the United States Securities andExchange Commission -- noted the leading role ofinstitutional selling in driving prices down. 1/ In onesense, heavy institutional trading on October 19 was nosurprise, since in recent years there has been heavyinstitutional trading every day.

However, two particular types of institutional tradingwere most closely identified with sales during the marketplunge: portfolio insurance, which I just discussed inconnection with over-valuation, and index arbitrage,another futures-related strategy.

Portfolio insurance, which calls for selling into afalling market, has obvious potential for exacerbatingprice swings. The second strategy, index arbitrage, doesnot create selling pressure, it simply transmits it fromone market to another. Thus, if sales in the futuresmarket lower the price of an index future relative to theprices of the index's constituent stocks, the arbitragerwill buy the relatively cheap futures and sell therelatively expensive stocks, profiting from the pricediscrepancy, and reducing it. It is argued, however, thatfutures markets are more volatile than stock markets. Thisappears to be true and may be attributable to the moreliberal trading rules in the futures markets; rules whichmay also make those markets so cheap and efficient thatthey have become the preferred markets for manyinstitutional investors. In any case, it is argued thatarbitrage transmits volatility from futures to stocks,creating vibrations that can touch off an avalanche. Atleast I think this is the theory. It has never beencompletely clear to me how much the critics of the futuresmarkets believe that the crash was an extreme instance offutures-induced volatility, and how much it was just anextreme instance of the old-fashioned stock market crash.The increase in everyday volatility that many ascribe toindex arbitrage may be an entirely separate phenomenon.

At any rate, futures trading did generate significantselling during the crash. A shortage of buyers in thefutures market meant that many portfolio insurance saleswere made by actual stock, not futures, sales, accountingfor about 6.4% of the October 19 volume in New York StockExchange stocks. Adding stock sales from index arbitrage,

1/ SEC Staff Study at xiii.

6

the figure comes to 14.7% of total New York stock Exchangevolume. ~ In certain intervals these two types ofprograms accounted for far more than this, 2/ but the greatmajority of sales that day had no connection to futuresstrategies. However, the more fundamental point is thathad there been no futures market, and had all selling gonedirectly into the stock market, that market might have beenunder far more stress than it was. A study for the ChicagoMercantile Exchange claimed that, even allowing forarbitrage-induced sales, selling on the New York stockExchange would have increased by a net of 85 million shares

14% of volume -- but for the existence of futures. ,101

One may argue that October's sellers would have soldmonths earlier or more gradually but for some misplacedfaith in futures, but this is difficult to prove. One mayalso argue that without futures markets institutions wouldnot employ trading strategies that call for rapidreallocations between equity and other types of assets, andwhich result in the quick disposal of large, diversifiedequity positions. But these strategies are central tomodern portfolio theory, and the stock market wouldaccommodate them if it could. It is not clear that theencouragement that the futures market gives to thesestrategies, with their heavy liquidity demands, offsets theadditional liquidity the futures market provides. Nor isit clear that these strategies are bad in themselves.

Last but certainly not least is the question of marketcapacity. One of the clear facts of October was thatmarket systems capable of the orderly execution andreporting of trading volumes larger than any previouslyexperienced were inadequate to the unprecedented pressuresof a 600 million share day. Reliable information aboutreal buying and selling interest was abnormally difficultto come by. There was in fact a very genuine fear that theentire system would shut down. The uncertainty that thissystemic overload introduced could only have made a badsituation worse. This is not an issue that has engenderedso much debate in the united states as the futuresquestion, but it is one of the first order of importance,

SEC Staff Study at App. D-50.

Preliminary Report of the Committee of InquiryAppointed by the Chicago Mercantile Exchange toExamine Events surrounding October 19, 1987, Tables 1and 2 (1987) ("Mere Preliminary Report"), p. 30.

7

and, fortunately, is one about which something is beingdone.

IV.

My Commission's initial recommendations for dealingwith possible market crises in the future fell generallyinto three categories: enhancing market capacity; reducingliquidity demands on the markets; and improving regulatorycoordination.

The capacity-expanding recommendations included thefurther improvement of stock exchange automatic orderrouting systems. The New York stock Exchange has completeda number of improvements and plans to be able to handle abillion share day in an orderly manner by late 1989. Thiswon't necessarily bring more buyers into a plummetingmarket, but it may reduce backlog-induced anxiety,misinformation, and panic selling. It will also permitmore effective index price arbitrage, to the extent the NewYork stock Exchange will tolerate index price arbitrage.

The SEC also suggested increasing specialist capitaland improving specialist discipline. Exchanges are takingor have taken both steps.

The specialist's job is to bridge temporarydisparities in supply and demand by standing readycontinuously to buy and sell for his own account. This isthe price he pays for the profitable privilege of being thecentral market in the stock in which he specializes.However, his job is not to peg prices or to throwaway hismoney. Greater specialist capital would not have preventeda huge price drop in October.

still, the prices in certain intervals suggest thatsome supply and demand disparities were temporary and mighthave been ameliorated if specialists had not been so closeto the limits of their buying power or, in some cases, hadhad a more active sense of their obligations asspecialists. Moreover, capital increases should reassureinvestors and those who lend to specialists that thetrading system is sound and able to function under extremeconditions. However, whether the specialist system itselfrequires reassessment is not a question that has beenseriously addressed.

Our Commission also suggested exploring other ideas toenhance market capacity, such as introducing certain"block-trading" techniques in the futures markets; tradingbaskets of stock on the NYSE itself in order to reducepressure on specialists in individual stocks; a

8

reassessment of the rule against short sales in a decliningsecurities market; and public dissemination of programtrading information to reduce uncertainty and fear aboutmarket conditions. The NYSE recently took a step similarto this, announcing it would publish periodic analyses ofprogram trading data.

To address sudden, extreme demands on marketliquidity, the Commission looked at the relatively largefutures positions that can be established with relativelysmall investments. Futures margins at a level morecomparable to stock market margins might reduce the size ofthe trades undertaken and to that extent dampen volatility.Therefore, at least until trading capacity can be enhanced,the SEC supported an experimental increase in futuresmargins. Margins on futures have been increased, althoughnot the level required of similar stock positions. WeshOUld, however, have no illusions that higher margins area panacea. We have no proof, for example, that leveragedfutures traders exacerbate rather than modify price swings.In fact, futures speculators appear to have been net buyersduring the October crash. ll/

Finally, the SEC called for improved market andregUlatory coordination. Most press attention focused onour suggestion to extend SEC authority to stock indexfutures, currently the responsibility of the CommodityFutures Trading Commission. 12/ Every other country I knowof has this kind of a unified responsibility, and there isnothing terribly radical about it, in and of itself. Theimportant question is how the SEC would use such newauthority, which at this point, I must acknowledge, it doesnot seem very likely to receive.

One of the most important steps taken so far inregulatory coordination has been the participation of SECChairman David Ruder in an ad hoc "Working Group onFinancial Markets," also including CFTC Chairman WendyGramm, Federal Reserve Chairman Alan Greenspan and TreasuryUndersecretary George Gould. It has reiterated proposalsby the SEC and others for a coordinated credit, clearingand settlement system across all markets to providereliable information and consistent rules on the financial

11/

12/

Mere Preliminary Report, p. 44.See SEC Recommendations Regarding the October 1987Market Break (1988) pp. 30-31 (contained in testimonyof David Ruder, Chairman, SEC, before the SenateCommittee on Banking, Housing and Urban Affairs, Feb.3,1988).

9

positions of market participants. The SEC recentlysubmitted legislation to Congress to help accomplish this.

The Working Group also recommended pre-set,coordinated, temporary trading halts in the u.S. stockmarket and in related option and futures markets in case ofextreme price declines. This type of "circuit-breaker" isintended to arrest developing panic psychology; to permitcreditors time to reassure themselves of the solvency ofthe market participants obligated to them; and to allow aninterval to disseminate radically new information aboutsupply and demand, thereby attracting bargain-huntinginvestors before prices have changed so much so fast thatpanic appears to have taken hold.

This is not certain to work, of course. The hugeprice drop of October 16, 1987, was followed by a wholeweekend when the markets were closed, but the market stillcrashed on Monday the 19th. Furthermore, it is conceivablethat the approach of the pre-established threshold for atrading halt would induce a further rush to sell or snuffout a rally that might otherwise have occurred. It is, forexample, not unusual to see a price trend accelerate as theregular end of a trading day approaches. Nonetheless, thecircuit-breaker idea has been accepted by regulators,securities exchanges and futures markets. The New Yorkstock Exchange and Chicago Mercantile Exchange, the leadingindex futures market, have announced a plan for coordinatedtrading halts and reopenings when the Dow Jones IndustrialAverage drops by 250 or more points.

Regulatory coordination on the international level hasbeen proceeding before and since the crash, although someof the most notable progress has been on law enforcementmatters rather than measures relevant to containing atrading crisis. Trading links and related clearing andsettlement links continue to be pursued, although most ofthe trading we would designate as international occursoutside these links; furthermore, clearance and settlementlinks can be no better than the systems linked, and many ofthese could stand improvement.

Perhaps more important to avoiding a crisis is thatlines of communication are being kept open, and openedfurther, among international regulators. I admit that thissometimes seems to amount to no more than interminablerounds of talk. But in the event of a crisis, there may bemuch to be said for knowing who to contact and whatreasonably can be expected of them.

since its original recommendations to expand capacity,reduce liquidity demands and improve regUlatory

10coordination, the SEC also has made several otherlegislative recommendations. These would provide it withemergency rulemaking powers during a market crisis andenable it, on a routine basis, to collect more informationon transactions by large traders and on the financial riskto which broker-dealers may be exposed through theiraffilia~es. We also proposed revising margin regUlation,to prov1de a larger role for the government on the futuresside and a larger role for the markets on the securitiesside.

Exchanges and futures markets have undertaken othermeasures themselves. As I mentioned earlier, the ChicagoMercantile Exchange and the New York Stock Exchangerecently announced plans for coordinated trading halts.The Chicago Mercantile Exchange also has an independentlimit on how far the price of the Standard & Poors 500index future can move in a day. The limit varies from 15to 25 points -- the equivalent of about 120 to 200 pointson the Dow Jones Industrial Average. Trades farther outthan that must wait until the next day. In effect, this isjust another type of temporary trading halt and is sUbjectto many of the same objections.

As for the New York Stock Exchange, after attempting,with mixed reSUlts, to unlink itself from the futuresmarket in times of extreme price movement, it has arrangedto match program trades against each other and halt tradingin individual stocks for large imbalances, while theimbalance information is disseminated. During volatileperiods, orders from individual investors will be givenpriority access to the Exchange's automated order deliverysystem.

We seem still today to be in a phase ofexperimentation. Ideally, this will provide us with newinformation Which, together with a continued sifting of theevidence from October, may clarify the proper course forthe long term. until that course is clearer, theregulatory response to the crash should remain flexible.

No one can guarantee that there will be no more suddenprice drops, although I'd be surprised to see another likeOctober 19. I would point out, however, that underconditions more extreme than any in generations, our systemheld together fairly well. Few American securities firmsfailed. These included only one that carried the accountsof public investors, ld/ and they will be compensated byour Securities Investor Protection Corporation. None of

ld/ See SEC Staff Study, p. xviii.

11

the major firms even came close to insolvency, and noclearinghouse on either the securities or futures sidefailed.

We can try to make doubly sure none of that happens inthe future, and we can continue to search for ways toreconcile portfolio-sized trades to our trading system.But we must be careful of taking permanent measures justfor the sake of taking measures. Thanks to the crash,there is a great deal of interest in analyzing the stockmarket and a political opportunity for constructive change.We should utilize these to come up with good answers tohard questions, not feel pressured by them to just dosomething -- anything -- that we may not be able to undofor a long while.