Sunday, 2 January 2011

High Frequency Trading and Flash Crash – 4: How the Pieces Fell Into Place

After Black Monday on October 87, there were loud and bitter complaints about the chaos which had prevailed throughout the day at the exchanges. The problems were especially glaring in Nasdaq, which mostly catered to retail investors. Many sell orders there were never acknowledged much less executed.

The chorus of drawn out criticism forced the SEC to act. Finding the path of the least resistance which also answered the greatest number of complaints, it compelled the Nasdaq market markers to create the Simple Order Execution System (SOES). The SOES was a computerized stock trading platform. Market makers had to display in it their best bid/asked prices for all the stocks in which they made market and honor those prices for at least 1000 shares.

As an example, Goldman Sachs displayed its prices for Microsoft as follows: MSFT 1000 24 1/4 x 24 5/8. This meant that Goldman was ready – and obligated – to buy 1000 shares of MSFT for $24 1/4 and sell the same number of shares at $24 5/8.

Since all market makers followed the same format – the main ones then, in addition to Goldman, were Lehman (LEH), Merrill Lynch (MER), Morgan Stanley (MSC) and Bear Stearns (BSC) – if one typed the stock symbol MSFT into the SOES, instead of a single price which was the stock’s last traded price (and the only price the investors were hitherto allowed to see), one saw the following:
  • MSFT 24 3/8 x 24 ¼
GSC 1000 24 1/4 x 24 5/2
LEH 1000 24 3/8 x 24 5/8
MER 2000 23 7/8 x 24 ½
MSC 1500 24 x 24 3/8
BSC 2000 23 7/8 x 24 1/4

Initially, market makers welcomed the system, reasoning that the “computerized” trading would boost the trading volume and thus, increase their profits. They were right about the volume. But the “thus” part, as in “thus increase profits”, did not follow in the expected manner.

Look at the price tableau above. As a market maker, BSC is obligated to sell 1000 shares of MSFT at $24 ¼. LEH is obligated to buy 1000 shares of the same stock for $24.375. One could buy $1000 shares of MSFT from BSC for $24 ¼ and sell it immediately to LEH at $24.375 for a riskless profit of $125.

Why would – how could – this relation exist? The answer is that, prior to the SOES, the market makers used the phone to take the pulse of the market and adjust their prices accordingly. If the prices changed rapidly, or if a market maker was distracted, he could fall behind in updating the prices. In the old clubby world of the market makers, where no outsiders were allowed, such complacency was harmless. LEH could still buy the stock for $24 3/8 because it sold it at $24 5/8.

Now, the SOES traders would bombard BSC and LEH with buy and sell orders – buying from the stock from one for $24 1/4 and selling it to the other at $24.375 – until either LEH lowered the bid or BSC raised the asked price.

That’s how day trading began

Economics and finance professors, too, welcomed the system. The SOES was the realization of their theories about the superiority of the “market based” solutions to all the problems. The SOES, as everyone could see, was bringing transparency and equilibrium to the markets. With a computer at each home and a day trader in each family, it was only a matter of time before stock prices could be said to be trading continuously; there would always be someone somewhere trading.

Such predictions and deductions seemed logical, in the way that a Norman Rockwell painting appears logical. In his paintings, there is never a violation of the principles of perspective and the subjects – humans, animals, things, nature –are depicted in a believable state.

The shortcoming of Rockwell paintings – and the standard theories of economics and finance – is their child-like, gullible simplicity that comes from the absence of experience. But adults “have been around”. Their “absence of experience” could only be due to an inability to comprehend the environment – the inability to see the surrounding social realities that are certain to get in the way of the happy picture.

Seth and Simon were two upper class friends. They finished boarding school and went to Harvard. After graduation, they decided to make a name for themselves by becoming independently rich. They would show their dads and moms their entrepreneurial skills.

One way of getting rich was scalping Indians. Indian scalps fetched $50 each. They decide to go West and scalp Indians.

They bought the most fashionable hunting clothes, the sharpest knives, the best ropes, the latest rifles and the fastest horses and headed West. There, they rode to the Indian Territory in search of Indians.

They looked high and they looked low but couldn’t find any Indians. After a while, they got tired and fell asleep.

Seth was the first to wake up. He looked and saw that they were surrounded by 50 Indians wielding axes. Behind them, in a bigger circle, there were 100 Indians with bows and arrows. Behind them, in a still larger circle, were 200 Indians with rifles. Seth was beyond himself. “Simon, Simon,” he shouted. “Wake up. We are rich.”

I have pointed out on many occasions that the subject of finance is not people. It is finance capital in circulation. Being a thing, finance capital cannot place trades, exploit opportunities or arbitrage “inefficiencies”. It does all that through humans who do its bidding. Likewise, in saying that speculative capital – the latest form of finance capital – is self destructive, I do not mean the speculative capital, as an abstract concept, somehow manages to commit suicide. The destruction, rather, is brought about by the actions of individuals who rationally strive to maximize their profit. That process – the individual profit maximization – is what destroys the conditions, the infrastructure and finally the very system which gave rise to speculative capital. From Vol. 1:
The manager of speculative capital must employ it in activities that are consistent with [its] attributes. True, the manager can decide the specific occasions of the capital’s employment, but that selection must be made from a menu of choices predefined by the attributes of speculative capital. He cannot commit the speculative capital to long-term mortgage lending. Thus, in the absence of any real option, the manager of speculative capital turns into its agent, someone who nominally “runs” the speculative capital but must in fact follow its “agenda.” Speculative capital becomes the grammatical subject of the sentence as if it were alive: speculative capital seeks arbitrage opportunities. Of course, it does so through its agent, the fund manager, but it is the speculative capital which determines the nature of its own employment and calls the strategic shots.
Implicit in the statement that speculative capital is self-destructive is the use of force, further connoting coercion and compulsion.

Look at the price tableau for MSFT. On top, the public quote for the stock is 24 3/8 x 24 ¼. These are the best bid and offered prices. As a customer, you could not sell MSFT higher than $24 3/8 or buy it lower than $24 1/4.

But these bid/asked prices do not come from the same market maker. Only LEH is bidding $24 3/8 for the stock. And only BSC is offering it for $24 1/4. Individually, the market makers have a much wider spread. If your broker had an “order flow” agreement with Merrill Lynch, for example – and everyone had an order flow agreement with a market maker – it would cost you $24.5 to buy one share of MSFT and you could only get $23.875 if you sold it. So MER made $375 in each 1000 shares of stock that it bought and sold.

That was true for all the market makers. Would they then allow a system which brought them tens of millions of dollars in profits each year be disturbed to their disadvantage?

The answer is that they would not – voluntarily. But then, on the other side was speculative capital.

In Vol. 1, I chronicled the open fight that broke out between the market makers and retail traders after the introduction of the SOES. I urge you to read that section as a case study of how events in real life take shape and unfold. Norman Rockwell sceneries they ain’t.

For example, whenever the market makers were caught as the subject of arbitrage, because they had failed to update the prices, they refused to honor the trade. This “backing away” from the trades went on for years without either the SEC or NASD taking actions to stop the illegal practice. A Wall Street Journal article from 1995 gives a glimpse of what was taking place. Note the paper’s framing of the story by use of terms such as “unsympathetic victims” and “SOES bandits”:
Major backing-away sanctions have been rare partly because so many of the backing-away complaints have come from … ’SOES bandits,’ whom dealers accuse of bombarding the market...with orders at dealers’ expense. The 12 backing-away incidents in the Morgan Stanley case all involved orders from SOES bandits, and the Lehman matter is understood to derive from bandits’ complaints, as well … yet despite unsympathetic victims, backing away from quotation is forbidden in the markets.
Put yourself now, if you will, in the position of speculative capital. Through the SOES, you finally have a venue to the world of market makers. You see that Lehman, for example, quotes MSFT at 24 3/8 x 24 5/8. That is a large spread. Why is there no mid price, say, $24.50?

The answer is that Lehman would not allow it. Market makers were happy with the way things were and would not allow a change. So their resistance had to be overcome – crushed, if need be. As always, the agents for the deed would be humans. From Vol. 1:
For years, the complaints of investors about the artificially high bid/asked spreads in Nasdaq stocks went unheeded. In 1994, the publication of a paper by two professors who claimed “collusion” among market makers finally drew the attention of the Justice Department which began an investigation. That forced the Securities and Exchange Commission (SEC) to act. In 1996, after two years of investigation, the Commission issued a scathing report about the conduct of Nasdaq market makers and demanded a series of changes to the system. One of the more significant of these changes was requiring market makers to post outside prices that improves the current best bid and asked.
Market makers fought back nail and tooth, through lobbyists, campaign contributions, placing stories in the press and even through legislation:
Three of the biggest Nasdaq Stock Market Dealers are mounting a behind-the-scenes campaign on Capitol Hill to block the Securities and Exchange Commission from imposing new rules on their market. [A lobbyist hired by some Nasdaq firms] has drafted a proposed amendment to a securities bill that … would obstruct the SEC’s plan to enact rules requiring more open display of prices on the Nasdaq over-the-counter market. The amendment would require the SEC to study, at great length, the rules’ effect on “the competitiveness and liquidity” of the market.
But the game was lost. Wall Street Journal, December 30, 1997:
A federal judge … granted preliminary approval to 30 securities firms’ $910 million settlement of a class-action suit alleging that the firms fixed prices in the past on Nasdaq Stock Market trades … The investors’ lawsuit alleged that more than three dozen Nasdaq dealers conspired to widen spreads on trades involving 1,659 stocks.
Speculative capital won the battle of the spreads. The bid/asked spreads which were kept artificially high at about .375 of a point, fell to 1/8 of a point or $.25. But who said that spreads must move in the increment of 1/8? Why not in mere pennies? That, too, came to pass. The New York Times, 1998:
After more than centuries of using a system descended from Spanish pieces of eight, American stock markets are now appear to be moving toward having stocks priced … in dollars and cents … If Wall Street does move, it is widely expected that it would lead to better … prices for investors. That gain would come at the expense of brokers, who have resisted the move in the past … A change in pricing could shrink their profit margins.
Against the will and interest of market makers, bid/asked spreads thus collapsed to what they currently are: a fraction of a penny for HFTers.

Speculative capital had one last vulnerability which its foes used to prevent it from dominating the daily stock trade. From Vol. 1:
To compensate for the shrinkage of profit margins, the size of the capital must constantly increase.
So if the size of speculative capital decreases -- or is forced to decrease -- it will lose its edge. That's where market makers made their counter move. Wall Street Journal, January 1997:
Nasdaq officials have asked the SEC to permanently lower the “minimum quote rule” to 100 for all Nasdaq stocks, saying it would help the market accommodate the SEC’s new sweeping order-handling rules.
The “minimum quote rule” above means exactly the opposite. It means lowering the “maximum” number of shares traders could buy and sell through the SOES. The SEC approved their proposal. A few months later, the same newspaper returned to the story:
[SOES] traders also hotly complain about Nasdaq’s 90-day test program that lets market makers trade only 100 shares of certain stocks at a time with SOES traders. The difference is vast; a quarter-point profit on 1,000 shares is $250; on 100 shares its only $25. And the pilot program includes the top-10 Nasdaq-traded stocks like Microsoft, Intel and Cisco Systems.
The Islamic legend has it that Azrael complained to God that it was unfair for him to be singled out as the cause of death on each and every occasion. God answered that he need not worry as there would always be a “cover story”: accident, disease, war, murder. Azrael’s name would never be mentioned!

Everything from the beginning of this post could be told as “human story”: Market makers vs. trades, regulators vs. market makers, or Nasdaq system vs. the NYSE system. But behind the stories is speculative capital whose inexorable logic, as the motivation of humans, drives the events.

Again, put yourself in the position of speculative capital. You have reduced the stock bid/asked spreads to under a penny, which means that the size of the trades have to be very large to be profitable or even make sense. But your size is reduced to a mere 100 shares. What would you do?

Why, you would do what every retailer knows: make up for lower spreads through the volume. That is high frequency trading. En ma fin, est mon commencement, you declare. The small time day trader is dead. Long live high frequency trader!