Friday was the 25th anniversary of Black Monday, October 19 1987, when the Dow Jones Industrial Average fell 508 points, or 22.6%, the S&P 500 Index fell 20.5% while the Nasdaq index (which few people followed in those pre-dotcom days) fell 11.4%. It was a worse one-day fall than in 1929, and was generally expected to presage a 1930’s style depression. Needless to say, no such depression occurred; even mild recession did not arrive for another 3 years and the Dow Jones average was above its pre-crash level within two years.

This time around, we may not be so lucky.

The extraordinary thing about the 1987 crash was that without the stock market being excessively overvalued, or any significant political-economic cause, the market fell more in that one day than the 13.4% fall in the worst single day of the 1929 crash. With the market far broader and more liquid than in 1929 and with traders being universally educated in the doctrine of its “efficiency” such an anomaly was not supposed to occur. Never in the course of human experience had it done so.

That was the difference. By 1987, we were no longer in the realm of purely human traders. Whereas human experience, with the irrationality of John D. Rockefeller buying at the bottom of the market in 1929, was able to stem the worst such declines, by 1987 we had already entered the age of the machine, less philanthropic than John D. Rockefeller and totally un-shocked by a drop of any size.

The 1987 drop was caused by the failure of a primitive system of “portfolio insurance” by which automated traders used stock futures markets to limit (theoretically) the maximum possible loss of their portfolios. As the futures dropped, the portfolio insurers were forced to sell more and more futures to preserve their supposed hedge – while on the other side of the market computerized arbitrageurs took advantage of the increasing discrepancies between futures and cash markets by placing gigantic sell orders for the S&P 500 stocks themselves. The resulting decline was greater than could have been created by any human influence. Fortunately, since there was no reason for it other than machine-based insanity, the recovery was relatively quick and the economic damage not great.

Since 1987, the participation of machines in the market has exponentially increased. The development of modern machine-traded markets is colorfully described in Scott Patterson’s “Dark Pools” (Crown Business, 2012). From Patterson we discover that market ethics have deteriorated, but so has true market liquidity for non-mechanical investors. The market has splintered into “dark pools” in which algorithmically driven machines battle it out like sightless triceratops in a swamp. Also like triceratops, if disaster occurs the signal from outside minders to stop trading is far too slow to stop the stupid beast from plunging into further disaster. Thus Knight Capital lost $390 million in a few seconds when it switched on its new algorithmic system. Similarly the “flash crash” a couple of years ago caused the market to fall almost 10% in a few minutes before it recovered.

Even more alarming than the overall crashes were the trades carried out in the middle of the meltdowns for 1 cent or $99,999 per share. These orders are entered by the computers to give them priority against other orders in the order book – priority is given to orders at the highest/lowest price, rather than to those in the queue first. Consequently, if these orders are executed in a market meltdown situation, it means that there are no other players in the market at all for a particular stock – all the other computerized traders have withdrawn, sensing an unstable market.

Only a tiny percentage of the volume in U.S. markets is now carried out by human traders. Even the order flow from human retail investors or almost-human institutions is routed through computers, normally to become chum for the piranha-robots in the trading pools (the computerized algorithms are taught to seize retail orders with especial glee, since they are thought very unlikely to reflect any special knowledge that might lead the piranha-robots to unexpected losses.) Institutional orders are especially disadvantaged, since an order to buy 50,000 shares, a typical institutional size, will be front run, side run, back run and jumped all over by the algos, resulting in a sharp price move that makes the institution’s average execution price far more unfavorable than under the old human specialist system.

With no human traders and machines that withdraw from the market when turbulence arises, there is effectively no liquidity in a crisis. Hence there is nothing to stop prices falling to zero, picking up all the algo traders’ 1-cent bids on the way. The market imposes trading halts in crises, but these are only likely to shut the market altogether, rather than providing a solution to the illiquidity problem. Hence, at some point it’s likely that we will see a forced closure of the market following a decline of an arbitrary large percentage in the major indices.

Getting the market open again after that closure would be a difficult task. With almost no human traders and the machines either sidelined or programmed to panic because of the large market movement, it would probably take several days before a mechanism could be organized to reopen the market in an orderly manner. What’s more, the prices available at such reopening might well be a multiple of 1987’s 22.6% below those prevailing before the crisis.

In 1987, the market was able to stagger on, although at one point in the morning of October 20 it seemed likely that it wouldn’t. Fed chairman Alan Greenspan was able to inject liquidity into the banking system and in the event, banking system losses were minor. The main difficulty arose from margin calls. A rather smug colleague of mine, who had received a large payout a few months earlier from the London investment bank where we both worked (being persona non grata to the bank’s management, needless to say no such payout came my way) had invested it in the U.S. market, margining his position. The story of his panic when he received a call at 1.30 am London time demanding an immediate margin payment of $700,000, gave his ex-colleagues like myself a quiet moment of schadenfreude. Nevertheless the reality was cruel; a 22.6% drop exposed a lot of apparently solid fortunes to severe erosion.

Needless to say, in today’s more leveraged economy, with a drop perhaps a multiple of 22.6%, or even a market closure, the margin calls would spread like lightning across the U.S. banking system, at a time when the banks themselves would have great difficulty in responding because their own positions showed gigantic losses (with their “hedges” in the derivatives market being caught up in the general liquidity spasm.) Of course, Ben Bernanke would turn the Fed printing presses on full blast, and if as in 1987 the dollar and the U.S. government’s credit were unquestioned that would very probably work, limiting the losses to those medium-sized institutions unfortunate enough not to have good political connections.

However there is an even more alarming possibility. The crash I describe would doubtless follow a period of turbulence in financial markets generally, as did the 1987 unpleasantness (there had been a bond market crash about six months earlier.) In such a case, the banks might well be sitting on large losses on their bond portfolios, while foreign investors would be bailing out of Treasury bonds and the dollar. A Bernanke attempt to open the spigots in that event would crash the dollar, spreading the panic from the U.S. stock market to the worldwide money and bond markets, and causing the dollar to become unacceptable to international buyers. The unhappy 5-year trajectory of the 1918-23 Weimar Republic would be concentrated into a few days. Imagine the worst days of the Great Depression, when the banks closed, combined with a Weimar level of inflation, at least on imports, and that’s what you’d get.

With stocks, bonds and cash all effectively worthless, the U.S. economy would then have to rebuild itself from a state of barter, with all debts either repudiated or inflated out of existence. Probably a Gold Standard would be necessary, as the Fed would lack all credibility and there wouldn’t be much left of the banking system. This could be done, but it wouldn’t do much for middle class living standards.

There are solutions to this, but they need to be implemented. The Bernanke monetary policy needs to be reversed, with an immediate increase in the Federal Funds rate to 2%, still below the level of inflation but sufficient to bankrupt all the silly games that have depended on endless years of zero-rate funding. That would de-leverage the economy, especially the banking sector and mean that any stock market crash illiquidity could be remedied by more or less conventional means, without risking a total loss of confidence.

At the same time, a “Tobin” transactions tax should be instituted, at a very low rate of perhaps 0.01%. That would eliminate the profitability of most “algo” games, which depend on high transaction volume and low margins, and thereby rebuild the percentage of the market represented by real retail and institutional investors. With this done, a one-day crash would become limited to a merely human scale. Although a major stock market decline might still be necessary and indeed economically beneficial, it would take place over the normal timespan of 12-18 months, doubtless bringing much elimination of “malinvestment,: in the Austrian economic term, but not risking a total economic wipeout (unless the government did something REALLY stupid, which is of course always possible.)

A computer-trading-driven collapse of the U.S. stock market would provide material for economics PhD theses well past 2100. Let’s not give our descendants this intellectual treat – and spare ourselves the economic agony of living through it.

-0-

(The Bear’s Lair is a weekly column that is intended to appear each Monday, an appropriately gloomy day of the week. Its rationale is that the proportion of “sell” recommendations put out by Wall Street houses remains far below that of “buy” recommendations. Accordingly, investors have an excess of positive information and very little negative information. The column thus takes the ursine view of life and the market, in the hope that it may be usefully different from what investors see elsewhere.)