This letter went out to BATS users on Thursday October 23rd.
The last few months have presented the U.S. securities markets with unprecedented volatility. A simultaneous convergence of several drivers, including frozen credit markets, housing decline, mortgage defaults, presidential campaigns, a dramatic 12 month bear market decline, mark-to-market write down accounting, and rule making under pressure, have all contributed to our industry’s current state of turmoil.
The recent financial chaos has been responsible for emergency rule making and regulatory reaction. Some of the rule making has expired, and some remains in effect going forward. In addition to emergency rule making, political and economic pressures are also responsible for a series of industry discussions on “what should we do next?”
I am seeing lines drawn between four different camps: 1) Those who prefer that nothing else be done besides what’s already been done; 2) Those who think a stock-by-stock circuit breaker should be put in place; 3) Those who want to institute a price or tick test; and 4) Those who think more research should be done, with careful consideration of unintended consequences.
BATS is generally supportive of the circuit breaker idea, a concept that has been openly discussed amongst exchanges and sell-side firms. While not perfect, at least this approach wouldn’t get in the way of normal every-day trading and would only be activated in the case of extreme distress on a stock-by-stock basis. In the absence of immediate pressure to act, however, BATS falls into the “more research is required” camp.
We are big fans of the recently enacted rules and enforcement regulation regarding fails to deliver. From all the data we have seen, this is by far the most rational and most needed form of regulatory intervention, and it goes right to the heart of the matter. If naked short selling activity contributed to the rapid and dramatic decline in certain stocks, then addressing this illegal behavior with more stringent delivery requirements and follow through enforcement, including prosecution, will likely prove to be the best industry response.
We believe that a price or tick test is the least desirable direction, since this approach could have a serious negative impact on liquidity in the markets. There is a small constituency arguing for reinstitution of the tick test. Proponents of the tick test argue that 1) It’s easy to [re]implement; and 2) It will help slow the decline in stock prices by limiting short selling activity. We generally disagree with both arguments.
A recent study, performed by the NYSE, surveyed CEOs and CFOs of publicly traded firms to find out whether they “wanted to see the tick test turned back on”. Maybe it was done to give issuers a chance to voice their opinion, or maybe it was intended to be used as quantitative evidence. An interesting data point maybe, but hardly a surprising outcome.
Market structure should not be driven by stakeholders who have a vested interest in the direction of the market, but should be instead driven by market centers and regulatory authorities who can hold themselves out as neutral to the ultimate direction of the markets. Said another way, market structure should support a system that allows for fair and accurate discovery of the fundamental value (prices) of the assets traded on them. The litmus test should be “whether a particular market structure will be effective at addressing an underlying issue and still provide a framework for fair and orderly markets in support of free trade”, not whether a particular structure appeals to the issuers or any other constituent, especially if that structure has been proven in the past to be ineffective.
A common misconception about short selling is that it is primarily used by speculative traders who will profit when stock prices drop. Yes, that is one use, and it’s the one that gets the most media attention. What most non-industry practitioners don’t understand is that a significantly larger volume of short selling activity results as a consequence of legitimate market making activity. Automated market making firms typically make continuous two-sided markets in their securities (posting both bids and offers simultaneously), and at any given point during the day a trading firm might be net-long or net-short in each stock they are active in. If they happen to move into a net-short position as a result of their market making activity, then any new sell orders they send to the market will be marked “short”. That is until their overall company position for that stock happens to change back to net-long. This kind of automated market making activity accounts for, by most estimates, more than half of the resting liquidity in today’s U.S. equity markets.
Hence, a poorly designed price or tick test that is intended to limit when, and at what prices, short sell orders can be executed will most likely have a major impact on liquidity provision. Limiting the ability for market makers to “make markets”, thus limiting the opportunities for others to find liquidity and trade against it, is an outcome our industry should try to avoid.
Most will recall that our industry was subject to price tests once before. These tests were formally removed, however, in July 2007. The industry conducted pilots and performed extensive analysis before removing the long held bid and tick tests. Those studies showed that the tests were no longer effective, and as such were officially removed from the rule books.
In an attempt to collect additional empirical evidence around the impact of short selling activity, and to help determine whether a price/tick test would potentially be effective, we ran an analysis across all trading activity at BATS over the last several months. The results support a position that short selling is not a significant driver of a downward direction in the price of securities. This conclusion suggests that the implementation of a price test would not be effective in limiting the progression of a downward trend. The analysis also reinforces the notion that a price test would have a material adverse impact on liquidity provision.
The tables below summarize, for each period analyzed, what percentage of selling was long versus short. We also broke down the selling activity to identify what percentage of volume was executed: 1) Above the last national execution price; 2) Equal to the last national execution price; and 3) Below the last national execution price. The sum of values in each row represents 100% of all activity during the period. Said another way, the sum of all long sales plus all short sales accounts for 100% of all executed volume, since a seller (either long or short) is on one side of every transaction in the market.
As an example of how to read the table, you can see that only 12% to 13% of all executions were the result of a short seller trading at a price less than the last execution price on the consolidated tape. This was true in each of the periods measured and did not vary significantly between any of the periods measured. You can also see, just as often, short sellers participated in executions above the last execution price on the consolidated tape. In fact, short sellers accounted for 30% to 35% of all transactions equal to or higher than the last national execution price. That’s a lot of liquidity that ends up supporting and/or raising prices. Another observation is that long sellers accounted for more shares sold at prices below the last national execution price than did short sellers.
Analysis results for selling activity during five (5) different periods:
￼For reference purposes, a summary of the S&P 500 during the periods analyzed:
The big take away here is that short selling is a necessary aspect of market making activity, i.e. liquidity provision. The short selling mechanism does not appear, by itself, to be responsible for undue downward pressure on stock prices. Short selling activity is a natural consequence of making a continuous two-sided market, as well as an important tool used in fundamental price discovery. Attempts to arbitrarily restrict short selling activity can have a material impact on legitimate market making activity and natural price discovery. If we aren’t careful, a poorly designed rule aimed at short selling could have a long lasting adverse impact on the quality of the U.S. equity markets.
BATS Prepares for Exchange TransitionThe next several weeks will bring major changes for BATS, both in the United States and in Europe. In the U.S., the big news surrounds our final push to convert into a registered Exchange.
On Friday of this week, BATS Exchange will go live for the first time. Friday begins a nine day transition from our current status as an ECN to being fully operational as an Exchange. The transition begins with one symbol and progresses through the full range of U.S. equity securities, finalizing November 6th.
￼On Tuesday November 11th, with the transition process successfully behind us, BATS will formally celebrate our grand opening as a registered National Securities Exchange. We are working out the details to have live media coverage of our opening bell that morning. More details will be available on our web site soon.
BATS Readies European Launch … be afraid, be very afraid :)As if converting to an exchange in the U.S. wasn’t enough to keep us busy, we are also preparing for the official launch of the BATS European Markets Division on Friday, October 31st. BATS Europe received its MTF authorization October 14th, and all systems are go for trading to begin on Halloween (we told the competition they should be scared).
BATS started the project to create a market center (i.e., MTF) in Europe just a little more than six months ago, and we are now within a week of going live. This is a testament to our focus and determination to move at “BATS speed” and the drive to simply get things done. Our team of professionals in London pushed this effort forward over the last few months with remarkable success. Heading up the London team are BATS Europe CEO Mark Hemsley, COO Paul O’Donnell, and Ken Conklin, Global Head of Business Development. As a result of tireless effort and hard work from the entire BATS Europe team, now 23 employees strong, we will be entering the competitive landscape at a pivotal time in the evolution of Europe’s financial markets.
Trading on the BATS Europe platform will begin with 10 securities on October 31st, and will then open up to all LSE, Euronext, and Xetra securities during the next few weeks. We have more than 20 clients already signed up with many more in the pipeline. Based on all the new venues now offering competitive trading services, as well as the ever changing settlement and clearing environment, the next several months are shaping up to be an exciting time for the securities industry in Europe.
New Record Volume and Market Share in SeptemberThe BATS Growth story continues to unfold month after month. Since July when I reported BATS matched market share had reached 10.10%, we have grown our volume-based market share to 11.49% for the month of October. As a result of highly volatile markets, and continued organic growth, our average volume for October is close to 1.7 billion shares daily, with a recent high water mark of over 2.7 billion shares handled on October 10th.
Not-So-New Way to Measure Market Share … Notional Value"Notional Value", which is also referred to as "Dollar Value Traded", is calculated by multiplying the execution price of each transaction by the total number of shares executed in each transaction. This method of calculating market share, as opposed to simply using the total number of shares that changed hands, can be a better representation of the actual "size" of a market. The larger the Notional Value traded, the more risk that actually changed hands. By way of example, 100 shares of stock ABC at $350.00 per share is a much larger transaction than 100 shares of stock XYZ at $2.00 a share. Other assets classes, such as futures and options, and most other equity markets outside of the U.S., historically have used Notional Value as an index into how much activity takes place on those markets.
BATS is leading the way once again with a new dynamic display showing notional value traded for all registered equity markets in the U.S. (i.e., those reflected in the consolidated tape). The BATS Market Volume Summary page (https://www.batstrading.com/market_volume.php) has become the benchmark for streaming daily market share analysis. One of the most useful web pages to the securities industry, thousands of visitors come to this page on a regular basis to see where trading is taking place.
We have now added an additional tab to the Market Volume Summary page. In addition to being able to see the total number of shares traded, by tape and by market, you can now see the same matrix by Notional Value as well. As before, the markets in the table are ranked in descending order by overall 5-day average market share. If you get a minute, take a look, you might be surprised by the results.
Thank You We would like to thank our subscribers for their continued support. As always, your comments and feedback are welcome. Sincerely, Joe Ratterman Chairman, President and CEO BATS