The Next Degree of Fragmentation


Have you ever gone to Vegas while the NCAA basketball tournament has just gotten underway? It’s crowded! It’s hard to find a seat at a black jack table, which is a shame if you are in Vegas for the action, and not just walk around the shops at Caesar’s. You wish management would just open up more tables, so you could spend your time trying to win money. Eventually they do.

The equity markets are not so very different.

A decade ago there were many, many games. And the differential between those with the right tools and those without was remarkable. Today this is less so. Loopholes have been mitigated. Many exchange and dark pool practices have been cleaned up. The buyside has become dramatically more educated. Conditions have become better for investors, but they have certainly worsened for the gamers, and frankly a low volatility market has not helped their cause either.

And while the gamers on one hand love fragmentation (they want many destinations to maximize the amount of race conditions, and the “arbitrage” they can play), they also hate the increased technology costs associated with the needed connectivity.

Introducing the New Bats Market Close

This brings us to this morning’s note. BATS has been not-so-quietly pitching plans to open up a new game, and yesterday it was more formally announced. BATS, pending SEC approval, will commence the BATS Market Close auction. This auction will compete with the closing auctions on the primary exchanges – NASDAQ and NYSE. It works like this:

  • Beginning at 6:00am members can enter, and cancel, MOC orders in the BMC, right up until the cutoff at 3:35pm.
  • A Bats Auction Feed will in real time transmit to the market place the matched trade information.
  • At 3:35pm execution reports will be sent to members, with unmatched quantities sent back to members.
  • After 4:00pm prices will be attached to these original executions, once the primary markets close and those prices become available. At this time, the SIP will receive the trade information.

BATS notes the recent trend for less trading participation during the day, and  increased participation at the end of the day. They note that while volume growth over the last several years has been relatively flat, volume growth in the closing auctions has grown nearly 70%; it now exceeds 5% of the day’s trading volume.

BATS also appears ready to argue against those who would argue that this product will harm the integrity of the primary market close:

BMC is designed to avoid fragmentation of liquidity at the primary listing market by only accepting Market On-Close (MOC) orders. Limit orders are the basis from which price formation occurs. Market orders are recipients of that price formation, but do not contribute to the price level.

Success of BMC may lead to a reduction in the number of MOC orders pooled together at the primary listing markets, but the number of matched market order shares will be published in advance of the primary market’s cutoff time for MOC orders. In this way, Bats will provide a transparent tally to reflect the added auction depth from the MOC orders for which it is responsible.

While participants could opt to convert Limit-On-Close orders to Market-On-Close orders in an effort to participate in BMC and reduce their transaction costs, such a decision involves risk and would be unlikely to ever be deployed except in stable securities with little price risk.

Our Thoughts on the BMC Auction

  1. The buyside generally prefers to trade in unfragmented ways, maximizing larger block fills. The closing auctions, while imperfect, have served as an important means to achieve large and low-cost liquidity. Fragmenting these few remaining “liquidity events” seems less in their interests and more in the interests of high speed intermediaries.
  2. The BMC will siphon liquidity from the primary closing auctions, and it will also create a new race condition at 3:35pm. The BATS Auction Feed will transmit information to the market place – additional information than already is transmitted by the exchanges.
  3. Entering a MOC order in the BMC Auction will not guarantee a fill, whereas entering a MOC order in the primary auctions will guarantee a fill.
  4. BATS will likely take market share from the primary exchanges, which will make their market data more valuable.
  5. BATS will likely figure to receive a high portion of the divided market data revenue pie.
  6. If the BMC is successful, NASDAQ and NYSE may alter their closing auction processes. The timing could be moved earlier for example, which might make BATS move its BMC cutoff time earlier as well.  Who knows – perhaps the exchanges will keep moving up their cutoff times to compete with each other so that the market closing auction cutoff time becomes 9:31am!
  7. If BMC is successful, perhaps NYSE and NASDAQ will start new closing auctions to compete with BATS – auctions that target each other’s primary listing auctions just as BATS is doing.
  8. Perhaps IEX, or other ne exchanges, copy what BATS is doing as well, further sliding the market down the fragmentation slope.
  9. Will trading around the close become costlier, as liquidity begins to fragment in new unforeseen ways?
  10. BATS’s attempt to fragment the close is very unique, and especially friendly to high speed trading firms in this way: High speed firms love the new games created by fragmentation, but not the new connectivity costs. BATS is delivering. If approved, BATS will deliver said fragmentation, without creating a new exchange to connect to.

If approved, be ready for a new dimension of fragmentation. Brokers will have to change their execution algos, as well as study the costs and benefits of splitting your closing orders. This is an added complexity with many potential pitfalls, but without much incremental benefit for investors. Did you seek out BATS management and ask them for this market innovation? If you did not, who did?

We are open-minded though, and we are especially anxious to hear your thoughts; please feel free to share them with us!