Is the National Best Bid or Offer (NBBO) being Ignored?
|
Back to
Table Of Contents
Executive Summary
The NBBO
lies at the heart of
Regulation NMS (Reg. NMS) and is the key concept that assures investors are
getting the best price when buying or selling stocks. However, due to the recent industry trend
that emphasizes speed at all costs, the NBBO, in practical terms, no longer exists.
There is no audit trail that can show definitively whether an investor received the
best price on their trade. The regulators do not appear to understand the root of
the problem because they continue to promote new regulation, when a simple and effective
solution exists: enforce Reg NMS.
Likewise, getting rid of quotes that only serve to manipulate others is
as easy as enforcing Section 9
of the Securities Exchange Act of 1934.
If any new regulations are needed
after enforcing existing ones, we think a minimum quote life of 50 ms makes the most
sense. Just the discussion of implementing such a rule would expose how deeply flawed
the system is today and would be sure to raise a lot of eyebrows. But to be clear, we do not advocate new regulation.
The Death of the NBBO
As the lifetime of a quote approaches zero, the arguments for and against a minimum-quote-life
rule become more interesting. Let’s suppose, for example, that the
day has arrived where a few of the top
HFT systems are able to send and cancel quotes
in 1
nanosecond (ns). For reference, light travels 30 cm (1 foot) in 1 ns. At this
rate, we could see 1 billion quotes per second
per stock (qpss). A thousand active stocks trading
at this rate would generate 1 trillion quotes per second (qps).
We admit that was extreme, but given the hyperbole from some HFT marketing
groups, we couldn't resist. So let’s slow things down by a factor of 1,000 and imagine
a world where HFT systems can send and cancel a quote in 1
microsecond (us). At this speed we could expect 1 million qpss and a thousand
stocks would generate 1 billion qps. Put another way, an active market with a speed
limit of 1 microsecond would generate 1 billion quotes per second, requiring everyone
receiving CQS
to spend about 1,000 times more for telco and equipment.
Still extreme? Let's make things 1,000 times slower again and imagine a world where
HFT systems can send and cancel a quote in 1
millisecond (ms). At this speed, we could expect 1,000 qpss and a thousand stocks
would generate 1 million qps. Even at this rate, many quotes will expire before leaving
exchange networks. Today we frequently see qpss rates of 2,000, with peaks in the 5,000 - 30,000 range.
These numbers are growing at alarming rates, and within a year,
if left unchecked, recipients of CQS will need to upgrade their telco to 10 gigabit.
It's helpful to keep in mind that section I.C.4 of
Reg NMS (page 30) states:
Accordingly, one of the Commission's most important responsibilities is to preserve
the integrity and affordability of the consolidated data stream.
And from the same document, page 410:
But in those limited contexts where the interests of long-term investors conflict
with short-term trading strategies, the conflict cannot be reconciled by stating
that the NMS should benefit all investors. In particular, failing to adopt a price
protection rule because short-term trading strategies can be dependent on millisecond
response times would be unreasonable in that it would elevate such strategies over
the interests of millions of long-term investors – a result that would be directly
contrary to the purposes of the Exchange Act.
Perhaps we need to modernize the definition of a quote, or maybe we need
a new name for what used to be called a quote. Not long ago, when a trader
(or auto-trading software) received a quote marked auto-execute, there
was a reasonable
expectation of hitting (trading at) that quote — the only real exception being that another trader
might beat them to it. Today, under the same circumstances, there is a significant chance that the same auto-execute
quote would have
already expired in transit and not be honored; the speed-of-light just isn't fast
enough.
|
"No one uses the SIP for the NBBO anymore"
|
The change in semantics becomes most important when we look at the definition of the
NBBO. Per Reg NMS, the NBBO for a stock is defined to mean the best
bid/offer sent by a market center
to the Security Information Processor known as the SIP (CQS,
UQDF). But no one
uses the SIP for that anymore, we are often told, which would seem to violate the
letter and spirit of Reg NMS. We'd like to note that
last year, 2.5 million
subscribers spent over $450 million
to receive and process CQS.
So what do they use for the NBBO if not from the SIP?
Each exchange computes the NBBO internally from direct connections to other
exchanges. As the speed of trading increases, the likelihood of any two exchanges having
the same NBBO decreases. Most of this is because of the pesky speed-of-light limitation.
So how does a trader know whether a trade was routed properly to the exchange with
the best price?
He doesn't. It is impossible.
You see, each exchange’s view of the
other exchange prices only exists in memory on
that exchange's machines. It is not
recorded. There is no audit trail. Sure, each exchange provides book-level data, but
that only includes prices for that one exchange
— not the prices that existed on the other
exchanges at the time of each order.
|
"It is impossible to verify that a trade received the best
price"
|
If we are going to allow machines to trade faster and faster, then at the very minimum,
there must be audit trail data that includes each exchange’s view of top-of-book
quotes for every linked exchange trading that stock. In other words, what now only exists in an exchange routing computer’s
RAM, needs to be captured and made available. The exchange routing simulation video below may help you visualize this. In the video,
each box represents one exchange and the price information it has from the other exchanges -- that
is the information which needs to be recorded to assure trade through price protection.
Essentially, each interlinked exchange would end up recording its view of every exchange's
top-of-the-book prices -- exactly what the SIP (the box at the bottom) does now.
Reg NMS has already addressed this issue: page 32:
If the
benefits of a fully consolidated data stream are to be preserved, each consolidator
would need to purchase the data of each SRO to assure that the consolidator's data
stream in fact included the best quotations and most recent trade report in an NMS
stock.
|
"The claim that aggregation causes the SIP to run slower is
absurd"
|
So why not use just use the SIP? Auditing, maintaining, and improving one system is much
better
than 14 (one for each exchange). Well, because as some, including the SEC, have said: The SIP is not as fast because it has to aggregate information
from the other exchanges.
Wait a minute.
How can an exchange ensure every order receives the best prices if it doesn't aggregate
information from all the other exchanges? Take a close look at the simulation video
and note that every exchange must essentially replicate the SIP's aggregation function.
There is no way around this. Which means the claim that aggregation causes the SIP
to run slower is absurd. Quote rate overload is the primary cause of latency in the
SIP and direct feeds.
Besides, there are numerous places where Reg NMS language is expecting that order
routing would reference the SIP such as this (emphasis ours) on
page 314:
For instance, modifications to order routing and execution systems will need to
be made to route and execute orders in compliance with the requirements of the Rule
to prevent trade-throughs of protected quotations (which include, for instance,
the ability to recognize quotations identified in the consolidated quotation system
as manual quotations on a quotation-by-quotation basis).
And on page 423, the use of the adjective displayed referring to prices in the SIP:
Intermarket sweep orders must, by definition, be routed to execute against
the full displayed size of protected quotations...
There is more.
A computer that aggregates information from other computers uses an aggregation policy
that details how it selects between messages coming in at the same time from multiple
computers (things like time-slice period, scheduling quantum, size of input queues,
overflow behavior, etc.). Let's just say it's very complicated and has many dependencies: it's easy to make mistakes (or hide bias).
The aggregation policy is important because it affects the 3rd criterion in NBBO selection:
price, size, time.
The SIP's aggregation policy is something that is fairly easy to verify. How easy is it to obtain and verify the 14 aggregation policies used by the 14 other exchanges? How do you know if exchange X's aggregation policy treats other exchanges equally?
It gets worse.
We have discovered a few anomalies when analyzing aggregation characteristics of CQS, which is
the SIP for stocks listed on NYSE, AMEX, and
NYSE Arca (home of many
ETFs). One disturbing anomaly is that the
SIP appears to be ignoring the timestamp in the quotes it receives from an exchange, using instead the actual time the SIP receives the quote, even if the timestamp
in the quote is over 5 minutes late. That is, we found an example of the SIP treating 5 minute old data as real-time, affecting
the NBBO in thousands of
stocks. We will be publishing the results of our analysis shortly. The main point, is that this type of detection is impossible to carry out for individual exchange
aggregation behavior; and if it was, it would take 14 times
more work.
|
"We found an example of the SIP treating 5 minute old data
as real-time, affecting the NBBO in thousands of stocks"
|
Even more disturbing, the SIP applies the timestamp to quotes and trades after
these messages have been throttled and queued, so it is impossible to detect or measure
latencies occurring within the system. We have
published numerous
studies on this
problem because it was one of the prime factors in the Flash Crash on May 6, 2010.
Recipients couldn't detect significant delays until the system suddenly became overloaded.
If timestamps reflected the actual time a quote was generated, we would have known
how dangerously close to saturation the SIPs had become weeks earlier. Sophisticated
trading firms with direct exchange feeds had to be aware of this.
We understand how difficult it can be to grasp these problems. Understanding complex
networked systems where the speed-of-light is the dominant source of latency is hard. When
the fate of your nation's financial infrastructure is at stake, we think regulators should have a solid understanding of these issues.
|
|
Exchange Routing
The video below shows how orders flow between linked exchanges trading one stock. Orders
are represented by triangles, and begin from the edge of the display and feed into one exchange. That exchange then updates and checks its internal table to determine
whether to route the order to another exchange with a better price, execute it locally,
or simply update the top of the book for that exchange. Only orders that affect the
top of the exchange book are included to reduce clutter. The order is then transmitted
to every other exchange making a market in the stock (via premium direct exchange
feeds) as well as CQS which is shown at the bottom. When other exchanges receive the
order, they update their internal tables in order to keep track of the other exchanges'
top-of-the-book quotes.
The price data within each box therefore represents how that exchange views the prices
on other exchanges trading that stock. Unlike this simulation which assumes a perfect
world: due to the variations in distance and connection
quality, system load, and
other real-world imperfections, exchanges won't always have identical NBBO information.
As update rates increase, the percentage of time that
all exchanges have identical NBBO information will rapidly drop to zero.
|
You can download the exchange routing simulation application,
which allows you to speed-up, slow-down, and pause the update rate.
Simply unzip to an empty directory, and
drag-and-drop one of the text files over the
executable. There is no installation,
and removal is as simple as deleting the files.
Inquiries: pr@nanex.net
Publication Date: 07/13/2011
http://www.nanex.net
| This report and all material shown on this
website is published by Nanex, LLC and may not be reproduced, disseminated, or
distributed, in part or in whole, by any means, outside of the recipient's
organization without express written authorization from Nanex. It is a
violation of federal copyright law to reproduce all or part of this publication
or its contents by any means. This material does not constitute a solicitation
for the purchase or sale of any securities or investments. The opinions
expressed herein are based on publicly available information and are considered
reliable. However, Nanex makes NO WARRANTIES OR REPRESENTATIONS OF ANY SORT
with respect to this report. Any person using this material does so solely at
their own risk and Nanex and/or its employees shall be under no liability
whatsoever in any respect thereof. |
|
|
|
