Sunday, February 13, 2011

High Frequency Trading and Flash Crash – 5: “Discretion to Delay Trades”

It is easier to gain insight into the theoretical principles of a system in its early stages of development. The early stages, whether of a natural, mechanical or social system, include only its defining features, those irreduciably minimum parts which are absolutely necessary for it to become what it is. As such, they are easy to spot and “tie” – through reverse engineering – to the principles that drive the system. As systems become more mature or more “advanced”, the layers of the later additions cover and obfuscate the essentials.

Thus, if you want to understand the principles of mechanical flight – powered, sustained, controlled – you’d have an easier time with a replica of the Wright Brothers’ plane or a 1920s crop duster than an Airbus A380. The latter has way too many components that are not essential or even related to mechanical flying. They get in the way of understanding what makes a machine airborne.

(I was told that within a few years German cars will have no hoods that drivers can open, which is just as well. Even today, under the hood of a BMW, for example, you only see sealed boxes with warning signs not to touch. So, while a layman could understand and thus, fix, a 1950s Ford, few drivers could make head or tail of a modern car, although the principles on which it moves are exactly the same as that of a Model T or even earlier: an engine produces power which is transferred to wheels which move the vehicle.)



The same applies to natural phenomena, which is why scientists constantly “go back” in time – whether through studying fossils, for example, or looking at the outer edges of the universe – in search of the earlier forms of their research subjects

And the same goes for social systems. High school students will hardly suspect it, but the reason we study history is to learn the past conditions that have shaped our societies. Only by knowing the past can we map the course of social development and locate our position in that process. To understand capitalism, for example, we have to go back to the time when the contours of this particular socio-economic system were being formed. Without that historical context, we could tell stories about the City hedge funds or HF traders – there was this “hedgie” who was so razor sharp and so eccentric and loved arts and only drank Opus One and etc – but we would understand nothing about the system.

(The logical equivalent of this “going back in time” is abstraction: peeling away the logical layers of a phenomenon to get back to its reason. “In the analysis of economic forms,” Marx wrote in the opening pages of Capital, “Neither microscopes nor chemical reagents are of use. The force of abstraction must replace both.”

That is why I began the HFT and flash crash series with a historical and historical background. Looking at HFT as is, it is difficult to see what is wrong with the practice. So, many trades are now being executed very quickly; instead of thousands of trades a minute, we now have millions. So what? Don’t they cancel out one another? And how exactly does that bring about a crash – or flash crash? Those were in fact the questions a commentator asked in Part 4 of the series.

Following the course of the development of trading in the stock exchange allowed us to see how speculative capital managed, over time and through various means to:

i) Reduce the bid/asked spread
ii) Increase the trading volume
iii) Increase volatility
iv) Break the monopoly of exchanges and specialist on order execution

These developments are sold to the public as benefiting “all the investors”; who doesn't benefit from a reduction in the bid/asked spreads?

But any benefits to the public, to the extent that they are real, are incidental. These changes are brought about not to benefit society at large but in consequences of the operation of speculative capital. They set the stage for further onslaught of speculative capital on markets.

Speculative capital, though, still remains unsatisfied. Not everything it wants and demands is achieved as a by-product of its operation. So, direct, human intervention also becomes in order.

The nature of such rule changes to the working of the stock exchanges are too technical to be noticed and appreciated even by those who follow the events.

Fortunately, like an anthropologist discovering an intact 30,000-year human fossil, I came upon this Financial Times news story (Nov. 1, 2010) about the Istanbul Stock Exchange that had all the critical elements in one place. The story was about the changes the ISE had adopted to make itself appealing to speculative capital – a replay of the Turkey's EU membership process on a small scale. It said:
Turkey's main bourse, the Istanbul Stock Exchange, is set to become the latest to open up to the rapid electronic trading practices that are sweeping the world’s markets with plans to ease access to foreign investors.

Brokers planning to take advantage of the moves said the change, effective from Monday, would open Turkish equity markets to algorithmic trading and attract quantitative investors.

The ISE will begin reducing tick size, the increments by which prices can move up or down, for stocks and exchange-traded funds, in a step likely to appeal to high-frequency traders.

Other changes, which took effect in October, allow traders, for a nominal fee, to cancel or reduce orders midsession – crucial to deploying algorithms making “passive” trades.
..
A third change the ISE has made means the identity of buyers and sellers will no longer be disclosed until the next session, a move some brokers see as a loss of transparency but others as a level of anonymity normal on European exchanges.
...
Overseas brokers had previously held back from trading in Turkey because “lots of the discretion built into algo trading US and European stock wasn’t available,” said Rob Boardman, Europe managing director of broker Investment Technology Group. “If a machine wanted to delay trading for a while, it couldn’t.”

If a machine wanted to delay trading for a while, it couldn't. Now, it can. This, FT calls “discretion”, i.e., freedom, that US and European stock exchanges offered but the ISE did not. Until now. Turkey, too, is emerging.

Mark that sentence. Therein lies everything you need to know about HFT and flash crash.

And I just realized that I would need more than 5 parts before I am done with the series!