Posted on April 28, 2014 by David Laidlaw
Michael Lewis’ most recent novel, Flash Boys, has drawn attention to the impact of high frequency (HF) trading on markets and traditional investors. Lewis’ thesis is that this trading has so distorted costs that the market is rigged in favor of these high speed traders against other stock market participants. While the system could benefit from greater transparency, we believe that high frequency trading is a net positive that increases liquidity and lowers transactional costs for investors by efficiently performing the market-making function.
High frequency traders, by definition, trade very quickly. Orders can be placed in milliseconds through a combination of high bandwidth fiber optic links, powerful servers co-located with the stock exchanges and sophisticated trading software. Over one-half of the volume of the 6 billion stocks traded every day is executed by high frequency traders. More amazingly, these traders offer over 300,000 orders per second. The majority of these orders, which are used by the traders as a price discovery mechanism, are cancelled.
This speed advantage allows them to buy securities at slightly lower prices than other investors and sell for higher prices. Though quite narrow, making money on the spread between these prices allows for a number of firms to operate very profitably. For example, a high frequency firm might be able to buy shares in Microsoft for $39.295 and sell them for $39.305 while a pension fund manager may only be able to buy and sell the same shares for $39.30. Therefore, the HF Trader is making $0.01 for every share traded. However, if the HF Trader is able to buy and sell 10 million shares at this spread, he or she can earn $100,000.
Market exchanges have always required a middleman to buy from sellers and sell to buyers for orderly functioning. In the equity markets, this function was historically performed by “specialist” firms such as Spear, Leeds and Kellogg or LaBranche & Company. These specialist firms had unique relationships with the exchanges such as the New York Stock Exchange where they were allowed to profit handsomely from their market-making activities, especially since prices were quoted in 1/8ths of a dollar.
HF Traders are now serving as market makers. The volume of their trades and offers allows for more accurate pricing of securities and narrower spreads between bids and asks. Canada’s stock market regulators increased fees on trades, order submissions and cancellations in 2012 and spreads widened. Studies showed that the the bid-ask spread rose 9% on the Toronto Stock Exchange as a result of the added fees.
While we believe that HF trading adds to liquidity and is a net positive, there are areas that would benefit from greater transparency. For example, the exchanges charge market participants for access to trading data. HF traders pay for the highest tiers of exchange data which provides an advantage over other traders.
The sheer volume of trading and offers through a more fragmented market has caused more glitches than in the past. It has been debated whether or not the Flash Crash of 2010, where the market dropped about 9% in a matter of minutes, was the results of HF trading. The market was also not able to process Facebook’s IPO last year as thousands of orders were cancelled and the stock didn’t trade for hours. Regardless of whether or not the traders were the cause of these problems, their presence in the market probably contributed to the market’s inability to function in both of these circumstances. Efforts should be undertaken to make markets stronger and more foolproof, rather than more fragile.
The structure of the stock market has become much more fragmented recently. While the New York Stock Exchange dominated trading volumes for decades, many more trades are routed over smaller exchanges. Large volumes of trades also take place outside of the exchanges in “dark pools” where buyers and sellers are matched.
If the stock markets were truly rigged, exchanges competing for business should be able to present differentiated venues to attract higher volumes. This process is happening. In his book, Lewis promotes the IEX founded by Brad Katsuyama as a more level playing field. The IEX intentionally slows down order processing so that HF trades do not have an advantage. The exchange markets itself as one where there is no unnecessary intermediation and information is not leaked to insiders. Fidelity Investments is also exploring creating another trading venue to limit predatory trading behavior. The emergence of these competing exchanges and venues is a very positive development and one that should blunt the negative impact of HF trading.
One of the most passionate arguments against HF trading is that it erodes investor confidence. The argument contends that if investors, especially less sophisticated participants, believe the market is rigged, then they will not participate causing our system of market capitalism to be degraded. Based on fundamental characteristics such as revenues, cash flows and sales, stocks are trading in line with their historical averages since the middle of the 20th century. Therefore, it does not seem as though there is widespread lack of confidence since, if there were, valuations would be much lower and the market for initial public offerings would be much less robust.
In our firm, we pay for live quotes from an institutional trading firm that allows us to see trades as they are being executed in real time. We see both the high bid plus those offers below the current high bid. This presentation allows us to visualize the depth of the market and understand the true liquidity of a particular security. Market orders are almost always executed for the price that we expect. For limit orders, we often bid at a level we hope will get filled, but sometimes we have to adjust our prices to execute the trades. Our institutional trader allows us to execute these trades through different liquidity sources (dark pools) and use different trading algorithms.
From our perspective, trading costs have declined incredibly over the past couple of decades and the trends are still positive. Trades executed through discount brokers are now executed for commissions of $10 or less. For bank custodied assets, we typically trade for $0.01 per share. When I started in this business in the mid-1990s, there were brokers that were still charging $0.50 or more per share with minimum ticket charges of $200. For liquid stocks, spreads of $0.01 between the bid and ask are not uncommon. Before the advent of decimalization when stocks began tradings in $0.01 increments, stocks generally traded in $0.125 increments or higher. Therefore, trading spreads were oftentimes 12.5 times greater than now just because of the increments into which the quotes were divided. For buy and hold investors, the impact of trading costs in most securities is negligible. If we buy 1,000 shares of a stock for $20/share, we might pay $8 commission for the trade plus $5 in opportunity costs for the frictional trading costs to HF traders or others in the market-making capacity. If the the position is sold 8 years later for $40.00 share, the investor would make $20,000 and only have to pay about $25 maximum for all trading costs associated with the purchase and sale of the security. These trading costs are far lower than available in other arenas such as fixed-income securities or real estate.
Overall, the market for common stocks would benefit from greater transparency. The regulators could also address some of the structural issues so that the markets can adjust to rapid price swings and high volumes without freezing. However, the system functions much better than it did in the “good old days” and competition should spark continued innovation which benefits all investors.