24 September 2012
With the robots, markets are back to 1930
A new generation of tools has been improved over the last five years, changing the entire financial
Wall Street (Afp)
From calls to clicks, the reversion of financial markets is served. The old good times of floor-dealings are gone, the amphitheatre par excellence where it was possible to sell anything among the calls has been replaced by machines. Computers revolutionized finance, as well as financial markets revolutionized trading. Normal computers are already outdated: the microsecond age has come. The HFT (High-Frequency Trading) era has started.
A new generation of tools has been improved over the last five years, changing the entire financial world, thanks to the unprecedented levels reached by technology – impossible to conceive some years ago. Practically, financial markets became plutocratic – again. As in Wall Street between 1920 and 1930, the financial universe is now governed by money: ninety years ago, the wealthier you were, the more privileged information you bought. Nowadays, the wealthier you are, the faster IT system you buy. The most remarkable evolution concerns HFTs, high-frequency trading systems, information systems based on algos molded on the final client's needs which are able to substitute hundreds of traders. Born between 1998 and 1999, at the end of last century, HFTs have replaced the already existing traditional negotiation tools. From Telerate or Quotron to enormously more complex systems: “Installed on supercomputers, their number crunching capability doesn’t make a poor impression when compared with recent machines adopted lately by NASA,” a Tradeweb analyst told Linkiesta. They operate in microseconds, trading huge balances, generally with short positions (often intraday) - even though this is a declining trend. HFTs are everywhere in the market: from stocks to bonds, through commodities and metals: “You can immediately identify them when trading – the Tradeweb analyst tells – a certain kind of operation on an index or another, as deep as quick, can be due to them only.”
Several HFT systems have been developed during the years. Labeling them “robot” is both naïve and wrong: customization is total. Cisco Systems, IBM, Oracle: the big names of the IT environment created programs which can be customized as much as needed. Moreover, there are several minor companies able to create HFT systems – at considerable costs – and teams of financial mathematicians ready to modify each single system. “We're talking about loads of money: some systems can be worth millions,” explains an analyst at Nomura. HFTs can be based on market volumes, market-to-market valuations, volatility, index performances, stock performances…on whatever can be used as benchmark. Almost the all of them come with a newsfeed reader implemented, making the machine able to decide whether to open or close a position. An instance: a news release on the FTSE MIB reaching a certain index mark appears on a Bloomberg terminal. The HFT system, which previously had bought futures and is programmed to sell them on the FTSE MIB, starts to sell. Automatically. “I’ve seen incredible scenes, with traders on HFTs working half than we do with our traditional systems. Sure, there are more risks, but you can leave the system work by itself,” a commodity trader at a primary Italian bank told Linkiesta.
There aren’t precise numbers on the volumes traded by these systems. Usually, who owns an HFT system trades on OTC (over-the-counter) markets – which are unregulated, with no official data. The most plausible estimates were given by the IOSCO (International Organization of Securities Commission) last May: HFTs accounted for a portion between 55% and 65% of the stock trading volume in the European financial markets. Dealing with US markets, this portion rises between 70% and 80 percent. Similar figures for Asia – even though it is necessary to take into account the slowness of the path to regulation, subject to more delays than in Europe or in the US. HFTs find their perfect environment in the currency market, with daily trades amounting to about $4,000bn: quick trades, remarkable carry trade, high leverage, high volatility. And no mention of how much these HFTs can really move. An explanation can be found in the dark pools – the anonymous version of the MTFs (Multilateral trading facilities), where HFTs often trade. Dark pools are sort of parallel markets, where FESE (Federation of European Securities Exchanges) estimated trades amounting to $5,500bn in alternative liquidity. Anonymous. We don’t know who enters, who owns what, how that is traded, how the price is generated. (Barely) seen, (price) approved, (asset) bought. The ideal environment for HFTs.
Considering the continuously increasing volumes, it would be easy to say that if you don’t own HFTs you’re out. Not an adage: banks, hedge funds, investment funds – investments in HFTs are increasing. Among the heaviest users, some American hedge funds like Getco LLC, Knight Capital and Citadel LLC, some European funds like Vega or Brevan Howard. And then, investment banks, like Goldman Sachs, Morgan Stanley, UBS, J.P. Morgan. Almost the all of them own this kind of trading systems. The reason is explained to Linkiesta by a Morgan Stanley trader: “The first time I saw an HFT I was fascinated. We didn’t have one at that time, we owned a simple CBT (Computer Based Trading) system, and I wasn't even about to finish my trade when that HFT finished its one.” Microseconds versus seconds. No chance. The first institutions with this kind of trading systems were some American hedge funds, like Renaissance. Then, a progressive diffusion – until HFTs became an integral part of financial markets.
Why is it worth to use HFTs? Omen nomen, first of all. High frequency trading grants an unprecedented competitive advantage on normal traders. As underscored by the US SEC (Security Exchange Commission) in a 2010 report on HFTs, “higher frequency means higher stability of the systems.” Theoretically true, as shown in the Flash Crash of May 6th 2010, the United States stock market crash in which the DJIA (Dow Jones Industrial Average) plunged about 1000 points (about 9%) only to recover those losses within minutes. As explained by the American CFTC (Commodity Futures Trading Commission) in a report dating back to the end of 2011, “the execution of the orders through HFTs reached optimal levels.” In other words, no problems for the traders. If A sells an asset to B, using HFTs there is just a minimal possibility of an error to be done in the order. Furthermore, technology has cut the costs of the transitions. As explained by the UK Government’s Foresight Project on The Future of Computer Trading in Financial Markets, the total cost of trading has fallen dramatically in the period 1975- 2010. “When it comes to trading, liquidity is enough now,” writes Sir John Beddington. This is perhaps the most important advantage in implementing HFTs. In such a fast and interconnected financial world, go short in liquidity means being set up. LTCM, or Long-Term Capital Management, can be taken as the outstanding example of this concept. Also known as “Nobel Prize Fund” in 1998 it lost $4.6 billion in less than four months following the Russian financial crisis. Would it have been different with an HFT? “Probably, giving that it is possible to trade liquidity directly in a dark pool, o in more than one, anonymously and immediately,” as Andy Haldane, head of the BoE Financial Stability Division, explained at the beginning of 2011. Finally, another remarkable advantage: the more items are listed into a public consultation, the easier it will be to get to an optimal solution. Similarly, both SEC and CFTF underscored, together with their British counterpart FSA (Financial Services Authority), that HFTs grant a better price generation. “Together with traditional trading systems, dark pools and MTFs in general, these tools granted benefits to traders, who are now able to find the optimal price for assets faster than ever,” as stated by FSA in 2011. After the analysis of a dataset composed of 26 HFT firms (total value amounting to $2.8bn, editor's note), Jonathan Brogaard, economist at the Northwestern University, suggested that HFTs tend to improve market quality. In his paper “High Frequency Trading and Its Impact on Market Quality” he highlighted that a stock price is "considerably higher, even by 5.6%" when HFTs are not used. That's huge, especially considering that each trader's need is to reach the best price. Above all, the main virtue of HFT systems is their speed – but the faster, the riskier.
HFT systems entail numerous risks – they are moved by several algos, hence can be touched by several categories of risks. As the Flash Crash highlighted, vulnerability is the issue. In case of HFTs based on volatility, when the VIX (Volatility Index) marks a certain value there are HFTs ready to sell their open positions on VIX. A trader at Morgan Stanley disagrees: “It’s kind of a walking on an iced lake by night: you don’t have a clue of what’s under your feet. If your computer crashes, if there’s an HFT faster than yours, if a bug closes all your positions…well, you can’t do anything at all.” The obvious example is Knight Capital, which few weeks ago lost $440 million in a software glitch. The more you can earn, the more you can lose.
The next constant is going to be a war: man vs. machine. And so far, we’ve been the losers of all the battles. Traditional traders are losing, because they can’t negotiate profitably against HFTs. Regulators are losing, because they’re too busy in analyzing little savers to understand that the world has changed. Banks are losing too. In a world where liquidity, not solvency, is the parameter to evaluate counterparty risk, go short in the intra-day means being attacked without a single possibility of defense. “Money are transferred continuously. And HFTs are making financial markets smaller and smaller,” a trader at Goldman Sachs underscores. How? With barriers to entry – the price of these trading systems, the margin of single trades are just examples. HFTs are not for everyone: only a good purchasing power grants the possibility to buy them. But everyone can be influenced by their still precarious stability.
Furthermore, HFT systems amplify the Butterfly Effect: a hurricane's formation in the US can be contingent on a butterfly flapping its wings in Australia. Practically, a small change at one place in a system can result in large variations into that system and into all its correlated systems. Similarly, if an HFT system closes a position on a safe bank, it is possible for the concerned stock to decline sharply: the butterfly effect can be dramatic anyway then. Two years ago HSBC issued a report analyzing the relation between “Black Swans (unexpected events of large magnitude and consequence, editor’s note) and HFTs.” If the probability of a Black Swan is added to the error probability connected with HFTs, risks increase exponentially. “Better for you not to play with this systems, which already represent the future, if you are afraid of risks” the trader at Morgan Stanley says. Well, as every trader knows, there is no risk free investment either.
*Translation by Marion Sarah Tuggey