Are today’s securities markets just faster, or are they fundamentally different from markets 10 or 15 years ago?
Increasing automation means securities trades that used to take minutes to clear can now occur in microseconds. Researchers are looking at the implications for investors, professional traders, and regulators and for the structure of markets themselves.
“Traditionally, financial markets appointed specialist ‘market makers’ to supply liquidity and keep markets orderly,” says Professor David Michayluk of UTS Business School, a researcher in market microstructure. “Those market makers played a key role in determining prices, and they accounted for a large percentage of trading.
“This is no longer the case. Over the past decade the work of traditional market makers has largely shifted to automated systems. Computers have taken the place of humans, and markets are trading much faster than ever before.”
Professor Maureen O’Hara, one of the world’s leading authorities in market microstructure research, says technological change has transformed markets, though speed is just one part of the story.
The rules and structure of a market are its microstructure. Microstructure is important because it influences the types of market information traders see and the ease with which they can “learn” from it.
“Markets are not just faster,” Professor O’Hara says of today’s environment. “Markets operate very differently now and I think they operate differently for several reasons – one of them being that markets are so much more strategic now.
New technology and greater speed lead to new strategies, which lead to new methods of trading and in turn to new market design, says Professor O’Hara, the Robert W. Purcell Professor of Finance at the Johnson Graduate School of Management at Cornell University in the United States and now also a member of the Finance Discipline Group at UTS Business School.
“When you trade strategically you don’t just transact one security, you’re transactinging across securities,” she says. “You’re not just transacting in this one asset class, you’re going to be transacting across all the asset classes.”
While traditional market makers set prices using only information about order flow in the particular security they traded, today’s automated liquidity provider uses information about all order flow in the market, Professor Michayluk says.
Financial economist Dr Austin Gerig, formerly a postdoctoral research fellow at UTS, now with the US Securities & Exchange Commission, believes automation does mean increased efficiency.
“When you mechanise any industry, or any part of the economy, you expect there to be efficiency gains – and I think the same thing has happened in financial markets,” he told a recent seminar at UTS Business School.
Today, every trade that occurs in any market is essentially a bet on the future state of the global economy, Dr Gerig says, who was speaking personally, not in his SEC role.
“Blur your eyes a little bit and everything is a trade on the same thing, so you have to take into account all movements across all securities. And in order to be tuned in to signals that are global signals, you have to have machines to synchronise prices, synchronise liquidity pools, across the universe of securities. You simply can’t do that with humans.”
Professor Michayluk and Dr Gerig have conducted modelling that compares traditional market making with an automated market maker that trades in all securities. In their model, automated liquidity providers priced securities better than traditional market makers because of the ability to observe order flow across securities.
Automated liquidity had material effects on investors, they found. Informed investors made smaller profits, because they now had to compete with one another. Uninformed investors lost less money because they were able to transact through the liquidity provider. (Research shows little retail trade goes directly to exchanges these days.) Overall, higher trading volumes meant transaction costs reduced.
But what are the implications of automation for regulators?
The UTS seminar heard of growing concentration of trading, with about 7 per cent of accounts responsible for 93 per cent of all turnover in Australia. Among high frequency traders specifically, the top 10 traders were responsible for 80 per cent of HFT turnover – which itself is about a quarter of overall market turnover.
Dr Gerig says, globally, there are risks when everything is so synchronised and firms are operating in markets internationally. “If something happens in one part of the market that can now very quickly propagate to the universe of other securities,” he says.
For researchers, that means issues such as excess intermediation and venue design come to the fore, says Professor Michayluk.
Professor O’Hara says that, certainly from a US perspective, “we are still in many ways regulating markets the way we used to, and I actually don’t think it’s the way regulation should go”.
Regulatory systems were designed to deal with risk, not to deal with uncertainty, she says. “When you are uncertain you just stop, and therein lies the challenge. When you have uncertainty you need to regulate on an ex ante basis not an ex post basis.
“Closing markets after they’ve fallen is not really good enough. You need to build in regulation that stops things before they happen.”
Another question that researchers are asking is “how fast is too fast?”. Dr Gerig, says there’s a “speed arms race”, and potentially over-investment in technology.
One estimate is that achieving a 3-millisecond decrease in communication time between Chicago and New York markets cost $US500 million.
“We know speed is important but we don’t yet know whether or not microseconds are important,” he says. “Do microseconds really matter? Or, really, could we operate at seconds?"