Do HFT firms pilot WWII bombers?

As a follow-up to my recent “Air Raid” commentary on the OPR and Locked Market Ban, there has been a lot of comments sent to me both public and private.  The Themis Blog speculated that calls to eliminate the OPR and Locked Market Ban was a reaction to declining HFT profits, which prompted the title of this note.  To my knowledge, HF trading firms, for the most part, don’t like either rule, but they have adapted to them.  In any case, even if the de-regulation I suggested did make such firms more money, that does not invalidate the proposal.  Market structure debates should be decided on what is best for investors, and the added complexity, higher infrastructure costs, higher spreads, and difficulty to innovate under the OPR means that it should be eliminated.

My favorite analogy is that the market structure, under a dominant NYSE was broken and the OPR and Locked Market Ban was a cast applied to fix it.  It is long past the time that we sawed off the cast and let the market breathe.

Getting back to the comments, one of them was that the most recent research, done in 2005 concluded that liquidity providing orders were disadvantaged by trade-throughs.  This, unfortunately, means that the SEC will not have data to justify changing the regulation.   What is interesting about that comment is that it is precisely why I wrote the “Air Raid” commentary in the first place; the cause of those findings no longer exists and the best approach is to ask if there is any research that supports those rules continuing. Absent such data, we should not keep rules with such apparent costs, just because they already exist…  My favorite comment said basically the same thing:  We should consider Reg NMS as a 10 year pilot program and simply terminate it based on the evidence.

 

In addition, there were comments from people who think that the OPR and Locked Market Ban somehow help individual investors compete better with HFT firms.  Other than the fact that some people see HFT traders as the “boogeyman”, there is no reason to think that HFT traders will ignore quotes, except for the simple fact that they do take costs into account when trading.

I have attempted to summarize the comments and questions into to following Q&A, which goes into a bit more depth on some of the key issues:

 

Q:  Why shouldn’t the SEC be concerned that eliminating the OPR will disadvantage retail and institutional posted orders?

 

A: “Best Execution” obligations are a better method of requiring brokers to access liquidity than Rule 611, as that obligation is not limited to the “top of book” and the OPR forces the broker to route orders that they believe will not be in their client’s interest.

The OPR only requires that brokers send ISOs to the displayed quantity at the top of the book of other exchanges, which means that when exchanges or brokers satisfy their Rule 611 obligations, they may not interact with displayed retail and institutional orders as they “walk their book” after sending ISOs.  (It is worth noting that, retail orders are often not placed at the top of the book as those investors tend to pick one price level and leave their orders at that level).  As a result, it is arguable that when brokers or exchange routers rely on satisfying Rule 611 as their de facto standard, that neither best execution nor order protection/trade through protection is satisfied.

In addition, the OPR forces brokers to route orders to access displayed liquidity in at least two circumstances that are not in their client’s interest:

  1. The analytics of that broker concluded that the liquidity is not likely to be available when they try to access it.
  2. The analytics of that broker concluded that accessing that liquidity is likely to harm their client by creating too much market impact, too much information leakage or from the opportunity cost of being forced to trade with a slow venue.

 

Q:  Aren’t locked markets a distortion of true trading interests?

A:  No.  Locked markets in most situations represent a spread of over half a penny in real terms.  The passive orders would receive just under 0.3 cents rebate while an aggressive order would incur a cost of 0.3 cents.   More important, however, is the fact that the current “ban” on locked markets doesn’t ban them; it creates an incentive for competitive “ship and post” activity.   Consider this simple example.

Two orders entered on behalf of institutional clients that receive the gross price plus a commission.  The first is a buy of 20,000 with 400 showing (The balance is in reserve) on exchange X. Second is a sell of 20,000 with 400 showing on exchange Y.  Instead of an immediate cross of 20,000, here is what happens:  400 gets “shipped” by Y to X and Y posts 400 visibly.  X prints 400, routes 400 to Y and posts 400 visibly, Y reciprocates and so on, with the result that it takes fifty routed orders and quote changes to complete the orders

Q:  How does the OPR interfere with the provision of block liquidity?

A:  The OPR requirement harms capital commitment by creating a “Heads you win, tails I lose” situation for brokers.  When clients want the immediacy of capital commitment, they used to be free to negotiate with market makers and agree on a net price either on the phone or via a system, but they no longer can.   This is because, the broker must first send ISO orders to all displayed quotes first and wait for those to be either filled or rejected before completing the client execution.  It is also worth noting that many brokers have interpreted the “best execution” requirements as meaning that they should sweep the full book (including any reserve quantity that “refreshes”).  In situations where this activity uncovers excess liquidity on the book, brokers have to give that improved price to their client, but, when there is none, they are “allowed” to commit capital under the rule.  This asymmetrical return profile means that brokers charge more for providing liquidity to all clients.  In addition, Rule 611 forces brokers who are willing to commit capital outside of the spread to immediately show their trading interest to the entire street when they sweep the book to enable the trade.   Before this rule, a broker who agreed to buy or sell a large position to a client could delay the trade report, keeping the activity quiet while they worked out of the position.  Today, however, all the proprietary trading firms that subscribe to market data feeds can reconstruct precisely what happened when a block trade is executed, including how much of the trade was likely provided as a principal fill.  This information leakage adds to the cost of block trading and is particularly important when institutions are seeking liquidity in smaller capitalization stocks that trade less volume.

Q:  Why should the SEC be concerned about the high levels of message traffic caused by the OPR and Locked Market Ban?  Hasn’t modern technology made this issue unimportant?

A:  The consideration is not that extra message traffic is difficult to manage per se, but it is an issue in the aggregate.  Consider two points:

  1. A significant amount of the message traffic under the OPR and Locked Market Ban derives from prop trading firms sending millions of post only orders (non-ISO) to exchanges in situations where they EXPECT to be rejected. They do so, because in the very small percentage of cases where they are NOT rejected, their order is typically near the top of the exchange’s order book queue on a price time priority basis.  Such queue placement, in thick order books often results in profitable trades.  Market makers call that activity “spoofing”, but while the regulatory definition of spoofing is illegal, this is not.  That is because regulators use that term for placing orders without intent to execute and these orders, if not rejected, are intended to execute.  If the OPR was eliminated and locked markets allowed, the exchanges would no longer be required to reject those orders and, therefore, this activity would stop.
  1. While it’s not difficult for market participants in any individual stock to manage the message traffic, the aggregate volumes are an issue. This is a contributing cause to the significant increase in infrastructure and data costs for trading firms.  In addition, it puts a very high burden on derivative product market makers.  I was recently told by a prominent options market maker, that they routinely have over 1 MILLION active quotes at any point in time, and they need to manage all of them relative to movements in the underlying stocks.  That is obviously harder to do with exponentially higher message traffic.

To summarize the discussions, one thing that was generally agreed upon was that THE key issue is that the SEC should decide if the OPR and Locked Market Ban would be rules they would pass today, if they did not already exist.   Unless a compelling reason can be determined to keep these rules, eliminating them to decrease complexity, reduce costs, and spur innovation is the best course of action.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s