TRADING STRATEGY
Ana Avramovic
Market Structure
We’re All High Frequency Traders Now
Market Commentary
Key Points
High frequency traders have dramatically
altered US equity market structure over the past
decade. Among the changes:
- Highly skewed intraday volumes and volatility
- Wider dispersion in liquidity among the most
liquid and least liquid stocks
- Greater pricing efficiency – for largecaps
Many of these changes – particularly the
contrasts between the most and least liquid
securities – seem to actually be intensifying as
HFT increasingly congregates in the most liquid
names and leaves the more challenging
smallcaps alone.
But just as brokers and other industry
participants have had to adapt to these
changes, they’ve also learned from them.
Brokers are now introducing dynamic algorithms
– such as Credit Suisse’s Dynamic VWAP –
that apply HFT-style strategies in order to
compete with HFT and reduce costs.
Our report also includes a list of recent
regulatory initiatives related to HFT.
Ana Avramovic *
Ana.avramovic@credit-suisse.com
(212) 325 2438
* primary author
(
212-325-2438
15 March 2017
How HFT Has Altered Market Structure
Is HFT good or bad for the market? While that can be a loaded question
that we’d rather not address, we can instead examine its impact on the
market.
HFT really became a dominant force in 2008/2009 on the back of the
financial crisis when volatility spiked, spreads widened, and investors
were clamoring for liquidity, which HFT stepped in to provide.
In the nearly decade since, they’ve firmly established their place in the
market ecosystem, primarily serving as a facilitator connecting buyers and
sellers through time, but also frequently criticized in that role for being
superfluous, or worse, predatory.
Whatever your view, their impact has been wide and, likely, lasting. We
find that they have:
Dramatically increased overall market volumes, but especially in
the most liquid stocks. This has created a vast – and growing –
disparity in liquidity between the most and least liquid stocks, in
both volumes and spreads.
Skewed intraday volume distributions. Where HFT is actively
serving as a market maker, in the most liquid largecaps, there is
a more even distribution of trading throughout the day (though it
still remains skewed to the end of the day). As you go down the
liquidity spectrum, where there are fewer market makers and
more naturals, liquidity concentrates more and more around the
auctions.
Made prices more efficient. There are fewer instances of prices
gapping in stocks that generally have a larger HFT presence.
HFT arbitrage also serves to keep ETF prices very closely inline
with their NAV.
But just as HFT has altered so much of how our markets work (spurring
numerous regulatory actions, see page 7), the brokers who serve them
and have to interact with them on nearly every trade have also learned
from them. Brokers are now introducing dynamic algorithms that apply
HFT-style strategies in order to reduce costs (speak with your CS sales
rep to discuss specifics).
As institutional investors avail themselves of these sophisticated
algorithms, and discount brokers fill retail trades through HFT
wholesalers, we’re all high frequency traders now.
Victor Lin Meera Krishnan
Victor.lin@credit-suisse.com Meera.krishnan@credit-suisse.com
(415) 836 7643 (212) 325 5613
 TRADING STRATEGY

How We Got Here

Exhibit 1: High Frequency Trading as a % of all
US Trading
High Frequency EquityasTrading
a % of all US Equity Trading “High frequency trading” may encompass a range of different strategies
70% (see box, page 3), but the majority of them – including passive market
60%
making and arbitrage – thrive in conditions of high volatility and wide
spreads. These strategies also depend on a stock having good liquidity
50% which allows HFT to enter and exit a trade very quickly and avoid carrying
40%
risk.

30%
The peak of HFT activity came in 2009 when the S&P bottomed, the
VIX topped 80, and average daily volume was regularly more than 10
20% billion shares/day. After that point, as the market recovered, volatility,
10%
volume and spreads came down, and there was less demand for a
market maker’s services. At the same time, the field had become more
0% crowded. Less demand and more supply of market making services
2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016
reduced profit opportunities. Meanwhile, costs continued to mount as a
Source: TABB Group
technological arms race forced firms to continually shell out for the latest
and fasted equipment. To continue to stay in business, one had to be
Exhibit 2:Avg
Average Posted
posted NBBO $valueNBBO $Value
of Smallcaps of Smallcaps
vs Largecaps faster than the competition. Data fees, which are a mandatory expense
$180,000
vs Largecaps byprice
by stock Stock Price to ensure adequate risk management and regulatory compliance, have
$160,000
LARGECAPS also climbed year after year.
SMALLCAPS
This combination – more competition for fewer profit opportunities –
avg $value posted on NBBO

$140,000

$120,000 caused some HFT firms to leave the market. Some switched to other
$100,000 asset classes; some turned their attention to other equity markets
$80,000 abroad. Those that remained in US equities largely concentrated on the
$60,000 most liquid, largecap securities, leaving the more challenging smallcaps
$40,000 alone. We can now see the dramatic impact this has had as the market
$20,000 has become highly skewed between the most and least liquid securities.
$0
<$20 $20-$50 $50-$100 >$100 Impact of HFT
Source: Credit Suisse Trading Strategy 1. Higher Volumes
Possibly the largest, longest lasting, most visible impact of HFT is higher
total volumes. Considerably higher. We estimate that volume from
money managers and investors, both active and passive, has remained
fairly consistent for at least a decade (between about 3 and 4 billion
shares per day). Total US volumes today, however, are more than
double what they were in the pre-crisis, largely pre-HFT years. The
difference is mainly due to HFT and high speed trading strategies.
Exhibit 3: Breakdown of US Volume
Breakdown of USby Source (Active vs Passive)
Volume
10
HFT
9
Passive Funds
Avg Daily Volume (billion shares)

8 Active Funds
7

-
1998

2003

2012
1996
1997

1999
2000
2001
2002

2004
2005
2006
2007
2008
2009
2010
2011

2013
2014
2015
2016

Source: Credit Suisse Trading Strategy

For methodology, please see Surprise! Active Managers’ Volumes are Looking Up
2

What is High Frequency Trading?
In 2010, the SEC issued a Concept Release on
Equity Market Structure that explored a range of
topics and solicited feedback on how regulators
should approach market structure changes.
Regarding HFT, the Concept Release acknowledged
that the term is “not clearly defined”. Therefore,
“rather than attempt any single, precise definition of
HFT”, the SEC instead outlines four broad strategies
they believe are commonly associated with high
frequency trading:
1. Passive market making
2. Arbitrage
3. Structural (exploiting latency advantages)
4. Directional – further broken up into:
 Order Anticipation
 Momentum Ignition
In a subsequent SEC review of economic literature
and publicly available information (see
https://www.sec.gov/marketstructure/research/hft
_lit_review_march_2014.pdf), the SEC concludes
that “the level and nature of HFT activity can vary
greatly across different types of stocks.”
Furthermore, “different strategies can have quite
varying effects on market quality.”
Ratio (average / median)
TRADING STRATEGY
Facilitating Trading or Just Noise?
There are both good and bad sides to the increased volume. The most
common complaint is that most of this volume is unnecessary, simply a
shuffling of shares back and forth without any intention to take a position,
or worse, manufactured activity simply designed to take advantage of
slower investors on the other side. While that can be true, we showed in
Valuing the Liquidity Provider that a majority of HFT activity serves to
connect those natural buyers and sellers and reduce waiting times, often
substantially so.
2. Greater Dispersion in Bid-Ask Spreads
Bid-ask spreads for largecaps and smallcaps generally move in the same
direction, meaning they both widen or narrow in accordance with volatility.
However, we find that the dispersion in spreads between the most liquid
and least liquid stocks has grown since mid-2009.
Exhibit 4 shows the average spread compared to the median spread of
both largecaps and smallcaps and how that ratio has changed over time.
For both of them, the ratio increases in volatile periods (meaning the
average is farther away from the median; there is more dispersion among
the stocks). However, beginning in about 2010, these differences have
grown larger for smallcap stocks – there is a greater difference between
the median and average spread, a larger difference between the widest
and narrowest smallcap spreads.
Recall that 2009 is roughly when HFT peaked and many of those
remaining focused their efforts on the largecap space.
Exhibit 4: Dispersion in Bid-Ask Spreads
Ratio of Average Spread to Median Spread in Smallcaps and Largecaps
Source: Credit Suisse Trading Strategy
3
 TRADING STRATEGY

3. More Intraday Skews (Volume and Volatility)

Another area where largecap and smallcap liquidity has grown
increasingly skewed due to HFT patterns is in intraday volumes and
volatility.
Longer Price Discovery for Smallcaps; More Flickering Quotes for
Largecaps
In The Trader’s Trade-Off: Volume vs. Volatility we showed that intraday
volume and volatility are much more skewed in smallcaps compared to
largecaps. At the beginning of the day, smallcaps tend to be more
Exhibit 5:Intraday
Average Intraday
Volume Volume vs. Volatility
v Volatility
Each point represents
(average Sept 10 min bucket
2015-March during trading day
2016)
volatile as they take a bit longer to establish fair price. But, at the end of
0.6% the day, they actually have slightly lower volatility than largecaps.
S&P
0.5%
open
How is that possible? Smallcaps do seem to experience larger price
R2K
gaps (see section 4), yet they tend to have less of the small fluctuations;
largecaps, by contrast, with their large presence of market makers, may
Volatility

0.4%
experience something like “flickering quotes” as the price bounces rapidly
0.3% between the (very narrow) bid and ask, particularly at the end of the day
(see The Trader’s Trade-Off: Volume vs. Volatility for details).
0.2%
3:40pm last 10 min Both of these phenomena may be attributed to HFT – their absence in
3:30pm
0.1% the case of smallcaps and busy back-and-forth trading in the case of
largecaps.
0.0%
0% 2% 4% 6%
Volume
8% 10% 12% 14%
More Naturals in Smallcaps Mean Volumes Concentrate at EOD
Source: Credit Suisse Trading Strategy The relative absence of market makers in smallcap stocks has also
caused smallcap volumes to become increasingly skewed towards the
end of the day. The mostly-natural (non-HFT) trading community has
come to expect that there will be substantially more volume at the end of
the day, so many of these traders simply wait until then to trade. That
creates a self-reinforcing cycle whereby more and more smallcap volume
shifts to the EOD (see How to Get an Edge Trading Smallcaps).

Exhibit 6: % of Day’s Volume in First/Last 30 Minutes

% ofNote: excludes
Day's Volume the MOC30auction
in First/Last Minutes Exhibit 7: Russell 2000 Volume Curve: 2007 vs Today
22 Russell 2000 Volume Curve: 2007 vs Today
14%
last 30 min 2007
20
12% 2016
% of Day's Volume

18
10%
% of day's volume

16
8%
14
6%
12 first 30 min
4%
10
2%
8
2006- 2008 2009 2010 2011 2012 2013 2014 2015 2016 0%
2007
9:30

10:00

10:30

11:00

11:30

12:00

12:30

1:00

1:30

2:00

2:30

3:00

3:30

4:00

R2K S&P500
R2K - Last 30 S&P - Last 30 Source: Credit Suisse Trading Strategy

Source: Credit Suisse Trading Strategy

4
 TRADING STRATEGY

4. Pricing Efficiency
We can see the dramatic impact that HFT as arbitrageurs has had on
pricing efficiency by observing the rate that price gaps occur. The
precise role of an arbitrageur is to seek out inefficiencies and mispricings
and step in to correct them. This process generally prevents prices from
moving too far out of line.
Arbitrageurs Minimize Price Gaps – For Largecaps
When we look at how often prices gap1 among smallcap stocks as
compared to largecap stocks, we find that it happens much more
frequently for smallcap stocks. This should not be surprising. What is
more dramatic, though, is when we follow how this rate has changed
over the years as HFT (arbitrageurs) have become a larger part of the
market – for largecaps.
Exhibit 8: Number of Price Gaps (at least 1%) For several years going into 2009, when HFT hit its peak, the ratio of
Number ofin R2K
Gaps in Stocks vs vs
R2K Stocks S&PS&PStocks
Stocks (1% level) gaps in smallcaps to largecaps remained around five gaps in a smallcap
50
stock for every one in a largecap.
45
40 Beginning in 2010, though, and largely continuing through today, as we
35
saw some HFT leaving the market and others congregating in largecap
land, that ratio increased considerably. Smallcap stocks now experience
30
about 40 times as many price gaps of at least 1% as largecap stocks do,
25
an 8-fold increase compared to 2009.
20
15 We should note that 2015 saw a sharp decline in this ratio, meaning that
10
largecaps became relatively more “gappy” than they had been. In 2015,
5
largecaps were impacted by troubles in China and overall fears of a
global slowdown more so than smallcaps were because largecaps tend to
0
be large multinational companies. Smallcaps are generally more exposed
2004
2005
2000
2001
2002
2003

2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016

to the domestic economy, which was actually showing surprising signs of

Source: Credit Suisse Trading Strategy
strength in 2015. See Largecap Traders Rush To Avoid Overnight Risk
for more.
Arbitrage in ETFs Helps Maintain NAV
The HFT as arbitrageur also facilitates the efficient pricing of index
products like ETFs and futures. In fact, this is the reason we estimate
that HFT has a larger presence in ETF trading compared to stock trading
(we estimate as much as 70% in ETFs, compared to about 50% in
stocks. See Are ETFs Hurting Stock Volumes?).
This arbitrage function ensures that ETFs trade in line with their NAV.
However, we can also see what can go wrong when HFT – who are not
registered as market makers; see box on regulations on page 7 – cannot
get the information they need to appropriately price their securities. On
August 24, 2015, due to a confluence of events, about 325 ETFs hit
circuit breaker trading halts (for over 1000 halts total) when a number of
underlying securities opened tens of minutes late (a NYSE rule allowing
for a delayed manual open that may have contributed to the problem has
been repealed and replaced).

1
A price gap is defined as a move of at least 1% within 1 minute.

5
 TRADING STRATEGY

High Speed Trading Goes Beyond HFT

Exhibit 9: Number of NBBO Quote Changes As brokers started to recognize that the changes brought about by HFT
Per quote
Number of NBBO Million Shares
changes Traded
per million shares traded would not likely revert, and as they acknowledged that they had to
25,000
upgrade their technologies in order to compete, we’ve started to see
quote changes / shares traded in millions

more HFT-style techniques from brokers and other sophisticated traders.

20,000
Therefore, because of their frequent order cancelations (which are
required to maintain queue priority and manage risk), such brokers will
15,000 look very similar to an HFT shop when one looks at the tape.

10,000
This is why we find that quoting rates – a measure of how often quotes
are updated during the day – peaked in 2011, while HFT activity peaked
in 2009: brokers learned from HFT and began developing their own high
5,000
speed strategies that generated frequent cancellations and quote
updates.
0
HFT-Style Algos Designed to Have Negligible Temporary Impact
2003

2006
2004

2005

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016
One example of this is Credit Suisse’s Dynamic VWAP (D-VWAP). In
Source: Credit Suisse Trading Strategy
contrast to the traditional VWAP algorithm which has been a broker
staple with little innovation for over a decade, the new D-VWAP uses
HFT market making logic and near-real-time reactions to market
changes to attempt to maximize liquidity provision over liquidity taking.
By avoiding crossing the spread, the algo should have minimal impact
cost. Similar logic and the ultra-high-speed infrastructure on which it is
built – including microwave technology – can then be applied to other
tactics.
Even retail traders are able to benefit from these advancements. Since
retail traders make their trades with discount brokers who in turn execute
through retail wholesalers, who are often large brokers or even HFT firms
themselves, retail traders can indirectly benefit from the same
sophisticated technology as hedge funds and other institutional investors.

A Lasting Impact
HFT has had a dramatic impact on market structure, including increasing
volumes and quote messaging, skewing liquidity patterns, reducing the
time to find a counterparty, and helping to price securities more
efficiently.
While some of these changes have been positive and some have been
negative, it is safe to say that even with any potential regulatory changes
(see discussion on page 7), many of these changes are likely here to
stay.
In order to adapt, brokers and other industry players are learning from
these HFT strategies and in turn offering tools to their clients (like
dynamic algorithms) that take advantage of the same sophisticated
technology and logic.
We’re all high frequency traders now.

6
 TRADING STRATEGY

Can Regulations Catch Up to HFT?

Our current market framework was largely created in 2005, with the SEC’s passage of Regulation National Market
System (Reg NMS). HFT was only a bit player back then, and as we’ve seen, a lot has changed since then. A number
of industry participants have therefore called for regulators to revisit Reg NMS, in whole or in part to reflect the changed
times.
Here is a list of some of the issues that might affect HFT which have arisen in recent years:
Access fee pilot. The SEC’s Equity Market Structure Advisory Committee (EMSAC) has recommended a
pilot program that would lower the cap on access fees that exchanges can charge for trading on their venue. A
related issue is the practice of rebating access fees, known as maker-taker pricing. While EMSAC specifically
recommended not to test an outright ban on maker-taker in the pilot, reasoning that lowering the fee cap to
different levels will provide similar information on how fees serve as an incentive to provide liquidity, others have
called to address (or ban) maker-taker fees directly. Those individuals claim that rebates are an unnecessary
inducement for HFT to trade more. No action is immediately pending from the SEC regarding a pilot.

Speed bumps. IEX, NYSE, and the Chicago exchange have recently introduced or proposed versions of a
“speed bump” that would delay orders or information. The intent is to reduce any speed advantage that HFT
currently enjoys over other investors.

Inspired by the idea of reducing any HFT advantage, Nasdaq has also recently proposed introducing an
“Extended Life Order”, aka ELO, that would give retail traders priority on the book by requiring them not to
cancel their orders for a given amount of time.

HFT registration. The SEC proposed requiring proprietary HFT firms that meet certain criteria to register with
the SEC in early 2016, but nothing has happened with the proposal since then (comment period ended June 1,
2016). Note that currently, HFT acting as market makers are not required to register as such, and they carry
no obligations to continue to provide liquidity in times of stress.

Taxes on messaging traffic or volumes. Various politicians have floated the idea of a financial transaction
tax in some form for many, many years. The most obvious target of such proposals is those who trade most
heavily, HFT. The idea has never gained traction among US regulators, though some versions exist in other
countries, and the EU has been attempting an EU-wide transaction tax for several years.
A related idea is a fee on excessive messaging traffic based on quote-to-trade ratios. Nasdaq actually does
employ such a measure, though it rarely kicks in. DirectEdge briefly introduced a fee in 2012, but it lasted for
only three months (they ended the program due to minimal impact).

Co-location. Co-location was a major topic in 2009-2010, but it is rarely discussed anymore. It refers to the
practice of HFT (and brokers) paying exchanges for the privilege of placing their servers in the same physical
location as exchange infrastructure in order to reduce latencies as much as possible. The practice is legal, but
in 2010, some wanted to outlaw it.

7
 TRADING STRATEGY

Disclaimer:
Please follow the attached hyperlink to an important disclosure: https://www.credit-suisse.com/legal/en/ib/market_commentary.jsp. Structured securities, derivatives and options
are complex instruments that are not suitable for every investor, may involve a high degree of risk, and may be appropriate investments only for sophisticated investors who are
capable of understanding and assuming the risks involved. Supporting documentation for
any claims, comparisons, recommendations, statistics or other technical data will be supplied upon request. Any trade information is preliminary
and not intended as an official transaction confirmation. Use the following links to read the Options Clearing Corporation’s disclosure document:
http://www.cboe.com/LearnCenter/pdf/characteristicsandrisks.pdf

Because of the importance of tax considerations to many option transactions, the investor considering options should consult with his/her tax advisor as to how taxes affect the
outcome of contemplated options transactions.

This material has been prepared by individual traders or sales personnel of Credit Suisse Securities Limited and not by the Credit Suisse research department. It is provided for
informational purposes, is intended for your use only and does not constitute an invitation or offer to subscribe for or purchase any of the products or services mentioned. The
information provided is not intended to provide a sufficient basis on which to make an investment decision. It is intended only to provide observations and views of individual traders or
sales personnel, which may be different from, or inconsistent with, the observations and views of Credit Suisse research department analysts, other Credit Suisse traders or sales
personnel, or the proprietary positions of Credit Suisse. Observations and views expressed herein may be changed by the trader or sales personnel at any time without notice. Past
performance should not be taken as an indication or guarantee of future performance, and no representation or warranty, expressed or implied is made regarding future performance.
The information set forth above has been obtained from or based upon sources believed by the trader or sales personnel to be reliable, but each of the trader or sales personnel and
Credit Suisse does not represent or warrant its accuracy or completeness and is not responsible for losses or damages arising out of errors, omissions or changes in market factors.
This material does not purport to contain all of the information that an interested party may desire and, in fact, provides only a limited view of a particular market. Credit Suisse may,
from time to time, participate or invest in transactions with issuers of securities that participate in the markets referred to herein, perform services for or solicit business from such
issuers, and/or have a position or effect transactions in the securities or derivatives thereof. The most recent Credit Suisse research on any company mentioned is at http://credit-
suisse.com/researchandanalytics

© 2017, CREDIT SUISSE