Lawrence began this presentation by pointing out that market structure impacts all traders and investors. Direct impacts come from the fact that markets are more fragmented than ever before. There are dozens of venues where trades can be executed, including exchanges, direct access networks, and dark pools. In the very distant past, all trades in listed stocks went through centrally located market makers, but the current environment is much more diverse. Now, there are 13 exchanges, more than 30 automated trading systems, and about 200 broker dealer desks where trades can be executed.
That diversity in where trades can be executed does have advantages for the trader. Lawrence identified several key drivers of changes to market structure. Technology has led to market fragmentation, as orders can now be routed simultaneously across different platforms and the central market maker function is no longer the only way to trade. Regulations imposed by the government and the exchanges themselves acting as self regulatory organizations ensure that orders are routed fairly. Competition among the different trade venues has led to smaller spreads, in the most liquid stocks spreads may be less than a penny.
These changes have led to new market participants. High frequency trading (HFT) didn’t exist several years ago but now accounts for a significant amount of trading volume every day, estimated to be about 60-70% of activity. Technology has been the key driver allowing these firms to find their niche. Low cost computers, costing only $5-10,000, allow for the execution of the HFT algorithms. The New York Stock Exchange is at the leading edge of technology, and Lawrence joked that their goal is to develop a system that will fill your orders before you even send them in. Current technology limits execution speed to nanoseconds. A significant amount of HFT is related to traditional market-making activity, Lawrence estimated that about half of their trading was related to this.
Execution is also changing due to globalization. But, regulations differ around the world. In the US, Lawrence said the regulators are trying to increase competition and there are no restrictions on dark pools or internalization of orders by broker-dealers or others. Canada only permits internalization if there is a price improvement offered to the trader. European regulators have limits on dark pools and order internalization based upon the size of the order.
Steve Poser followed Lawrence with some comments directed at technicians. He noted that several academic studies have shown that HFT has a net positive impact for traders. Spreads are narrower and liquidity is generally deeper. Dark pools offer even deeper liquidity for large trades. But there has been a general decrease in the average trade size over the past few years, and now stands at about 300 shares per trade, down from an average of 900 shares per trade as recently as 2002.
HFT, Steve continued, is highly correlated with market volatility, and it tends to be most pronounced in high volatility stocks. Internalization and dark pool activity seems to be concentrated in less volatile stocks. Technicians can model volatility against volume to gain insight into this relationship.
Increased volume and trading activity also has an impact on many indicators. Breadth and tick data will obviously be impacted, and tick data may actually present an overwhelming amount of data given the smaller average trade size. Volume data may not be directly comparable to historic data given the dramatic changes in execution and market activity.
Breadth data can also be impacted by the number if exchange traded products now listed on the exchanges. An operating company only calculation can help with this challenge. Steve also noted that high volatility days will have a large impact on 52-week highs and lows. May 6, 2010 was a great example of a large number of stocks and ETFs hitting new lows, even though they traded at that level for only a few minutes. That was an event driven more by market structure than price trends or the underlying fundamentals of a company.
Market structure has evolved rapidly recently, but market structure has always been changing. It will continue to change in the future, and technicians need to follow these changes closely. Indicators may not retain their historic usefulness in the current market, given the increased volume that occurs for new reasons. Rapid price arbitrage and market making functions are now shown on the tape just like trades driven by the emotions of hope and grade that classical chart patterns are based upon. The impact of all this is unknown.
The future holds still more changes, and perhaps more data. The NYSE is looking at the feasibility of sub-penny quotes. While the possibility is under discussion, there are technology considerations that need to be addressed. There are already one billion quotes a day in the system, and sub-penny quotes would increase that number. Communications and data processing capabilities would need to be expanded to accommodate that, and that would drive even more changes to volume and breadth data.
To view this presentation from the 2011 Annual Symposium, got to: http://go.mta.org/11slp