The Exorbitant NMS Speed Game

Summary

• The National Market System (NMS), with more than 12 separately-located exchange transaction engines processing identical transactions, is expensive for no good reason.
• And there is a quick fix – colocated exchanges.
• Once this change is made, the door will open for further improvement.

Photo by powerofforever/DigitalVision Vectors via Getty Images

“The time has come,' the Walrus said,

To talk of many things:

Of shoes — and ships — and sealing-wax —

Of cabbages — and kings —

And why the sea is boiling hot —

And whether pigs have wings.'”

• Lewis Carroll

I question the common sense of the National Market System (NMS) – separately located computerized transaction engines processing identical transactions, all to achieve an achievable goal in an impossible way – ever faster computation and distribution of a single value, the National Best Bid/Offer spread (NBBO), collected at multiple separate locations.

Multiple locations are a relic of a time when people, not machines, traded in exchange locations. Now that computers determine location, there is no good reason to locate exchanges in multiple places. To locate them in a single place would eliminate many of the excess costs of retail trading identified after the GameStop frenzy.

Most obviously, colocating the exchanges would reduce the number of colocation fees paid by each broker-dealer from more than 12 to one.

The NMS problem in physics.

SEC rules require customer fills be completed within the NBBO. But the NBBO, the way it is calculated today, is the non-answer to an answerable question. The question: What is the NBBO at this instant in time?

With high-speed processors and data transmission today, in a physical world governed by the Theory of Relativity, there is no way to measure one NBBO for two exchanges at separate locations. Distance and time are not independent in Relativity. A number measured at one location and time is not comparable to a number measured at a different location at the same time.

This sounds academic, but it is a very practical, highly profitable, reality for High-Frequency Traders (HFTs). It is also the source of a major share of the profits of the Gang of Three (Intercontinental Exchange Inc. (ICE), Nasdaq Inc. (NDAQ), and CBOE Global Markets (CBOE)) and of their recently SEC-approved competitors, IEX and MEMX.

But the calculation of NBBO from multiple locations has nothing to do with quality customer fills. The SEC knows NBBO is nonexistent but proposes no commonsense solution to the problem although an obvious solution exists. Why not just explain the issue? Who could possibly have a stake in producing prices at the maximum achievable speed from 15 or more remote source locations?

A single location.

One way to resolve the physical conundrum is to define NBBO at a single location. I know. You are probably thinking “My basement.” But seriously, the location decision belongs to experts. The point is that many problems are resolved by a single location. For example:

• The end of payments for order flow.
• The end of inter-market arbitrage.
• The end of multiple exchanges all charging fees for a single transaction price.

But not a single exchange. Once the first change is made, the second change is inevitable. This third change is more problematic. How will the different business plans of the various exchanges reach their customer constituency? The exchanges would still want to attract different customer constituencies. And why not?

But a single location would put an end to pointless interlocation arbitrage among otherwise identical exchanges. There would be fewer exchanges. Speed differentials would be identifiable by a look at the exchange rules on file with the SEC. Thus, the exchange rules affecting order speed and the trading constituencies the rules benefit would be transparent.

Trading costs after the location adjustment.

The Gang of Three seeks to cover the entire customer waterfront. Their rules, once upon a time circa 1970, benefited broker-dealers more than customers (retail and institutional) but with the change to electronic trading and publicly held for-profit status, exchanges suit themselves. Their somewhat inflated fee structure inspired broker-dealers to create their own exchange (MEMX) and institutions to create theirs, IEX.

The Gang of Three. Their fees are the highest this oligopolistic market will bear. Wall Street broker-dealers and high-frequency traders (Citadel, Virtu)) are the primary beneficiaries of differing fees charged by the Gang of Three. According to a recent lawsuit successfully brought against the SEC's proposed pilot study of maker-taker fees charged by exchanges, exchange fees may not be regulated by the SEC now.

MEMX. MEMX fees are an attempt to undercut these high Gang of Three fees, to reduce the exchange share of the customer dollar, but they do not disturb the equation that balances the cost of trading for retail customers and institutions vs Wall Street's broker-dealers.

IEX. Built to favor institutions, IEX's speed bump coupled with its discretionary peg (D-PEG) is designed to give its primarily institutional resting orders a speed advantage over Wall Street speed traders by allowing pegged orders to read and react to incoming prices before outside orders can get to them.

In summary, the cost to retail customers of trading has been buried in the bid-ask spread since Robinhood introduced zero commission trading in 2013. There has been a hot debate about the share of the cost of trading that should go to exchanges – with action in the form of low-exchange-fee MEMX – and some interest in reducing the costs of institutional trading – with IEX and numerous alternative trading systems (ATS) designed to lighten the institutional load. Yet no broker/exchange/investment fund has gone after the billions in excess retail costs of trading that the GameStop (GME) frenzy revealed.

A retail-directed securities production and distribution institution.

Once the easy, single-location fix happens, are there any costs of transactions and portfolio management that remain available to be returned to the retail traders whose investments create them? There are four underlying costs of investing in securities.

• Clearing.
• Transactions.
• Inventory.
• Risk management.

Each of these costs can be minimized by putting the securities trading and portfolio management process under a single transparent roof. A schematic diagram of a transaction between settlements in this system looks like this.

Source: Author

A schematic diagram of settlement looks like this.

Source: Author

Clearing costs. The combination of an exchange with an investment fund puts all investors’ funds under a single roof, dramatically reducing clearing costs. Interestingly, an ETF of another investment fund adds another layer of clearing costs, giving the exchange investment fund ETFs a competitive advantage.

Transaction costs. By settling purchases with a single seller, transaction costs and credit risk are minimized.

Inventory. By placing inventory under a single roof, interest earned by securities lending is transparent.

Risk management. Brokers no longer hold customer inventories. The exchange investment manager serves that function. This minimizes the capital cost of margins while assuring a single exchange-determined standard level of margin protection based on market conditions.

Summary.

An SEC plan to colocate the exchange computers, eliminating the need for customers to pay fees for colocation at 15 or more separate exchange locations, would easily save more in reduced trading costs than it would cost in one-time operating expenses. An exchange proposal to the SEC to colocate exchanges would show the trading world a willingness to reduce customer costs while improving customer service substantially.

Other improvements. A different form of transactions service firm would be a one-stop-shop. By eliminating inefficient aspects of the current market structure, the real cost of trading to retail and institutional traders could be dramatically reduced.

Major inefficiencies would be eliminated or reported to customers transparently.

1. Current clearing through the DTCC for each individual transactor – to be replaced by a single exchange investment management firm that clears all transactions.
2. Reducing transaction costs by eliminating transactions directly between buyers and sellers.
3. Securities lending revenues that go to brokers and institutions now would be transparently reported by an exchange with an affiliated investment fund, and perhaps rebated to customers.
4. Risk management would be made both less costly and safer by permitting exchanges to set margins in line with evolving market conditions.

The myth that retail trading is free did not survive the GameStop frenzy. Will there be a lasting improvement in the balance between Wall Street costs and customer costs of trading financial instruments now that the unbalanced cost reality is clearer?

Analyst’s Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.