Anda di halaman 1dari 8

TRADING STRATEGY

Ana Avramovic
+ 212 325 2438
Phil Mackintosh
+ 1 212 325 5263

Inside the NBBO: Pushing for Wider and Narrower! - Spreads


Market Commentary 15 May 2013

Key Points Spreads Are Tighter Than Ever


Bid-ask spreads, currently averaging less than 3 bps for S&P 500
Many traders would agree that tighter
stocks, are at the lowest levels ever. And yet, market participations
spreads are good thing.
continue to question the quality of our market structure. To complicate
Indeed, our data show that tighter spreads the debate, a current regulatory proposal would widen bid-ask
translate to lower costs and less trading spreadswhile another measure was approved last year that allows for
off-exchange. narrower spreads.
And yet, regulators are discussing Exhibit 1: Average bid-ask spreads over time (S&P 500 stocks)
100
widening spreads.
90
Interestingly, regulators have also recently 80
approved plans to narrow spreads. 70
In this report, we examine the merits of
basis points

60
each plan, consider whether they would 50
work, and highlight potential problems. 40
30
We also look at todays record low spreads
and whether they could be even tighter. 20
10

0
2001

2007
1994
1995
1996
1997
1998
1999
2000

2002
2003
2004
2005
2006

2008
2009
2010
2011
2012
2013
Source: Credit Suisse Trading Strategy

Spreads are closely related to volatility. Since volatility is sitting near


record lows, similarly record low spreads should not be too surprising.
But when we take a longer view at the history of spreads, two themes
emerge:
Exhibit 2: Spreads vs volatility in 3 regimes 1. Although the VIX was at comparable levels to today in 2005-2007,
Spreads (S&P 500 -stocks)
vs Volatility S&P 500 stocks spreads were almost twice as expensive (see box on page 2). They
11
get even more expensive when we go farther back.
10 2. In the past year, spreads have remained fairly stable despite the
9
falling volatility (Exhibit 4, page 2).
We look at both of these below.
8
Electronic trading & competition compressed spreads
Spreads (bps)

7
Our market has experienced a number of structural changes since the
6 mid-90s: tick sizes went from eights to sixteenths in 1997, and from
sixteenths to decimals in 2001. Electronic and algorithmic trading grew
5
through the 2000s. Ultimately, these also led to the emergence of a
4 new market player: the high frequency trader (HFT).
Pre-crisis (2004-07)
3 Crisis (2008-09) HFT is often blamed for hurting market structure and adding unnecessary
2010-Now complexity to our market. Without debating that, its hard to argue with
2 data, and the data clearly show a decrease in transaction costs (see How
0 10 20 30 40 50 60 70 80
Much is Market Structure Hurting Investors? for the Credit Suisse
Volatility (VIX)
Transaction Cost Index). This suggests that competition for top-of-book
Source: Credit Suisse Trading Strategy position by HFT and electronic market makers may be responsible for
lowering costs.

(212
(
90
Exhibit 3: Global bid-ask
Global Bid-Ask Spreads spreads
80
Developed
Emerging
TRADING STRATEGY
70

60
Exhibit 2 (on page 1) showed spreads and volatility since 2004 and
suggests that there are three distinct periods with spreads getting
Avg Spread (bps)

50

40
progressively tighter each time. Most important, in the current, post-
30
crisis period (2010-now), spreads are lower than they ever were for every
20 volatility level.
10
Recently, spread compression has stalled
-
We should all appreciate how tight spreads are, both compared to prior
Czech
Netherlands
India

Mexico

Korea

Greece
Malaysia
Canada

Israel
Australia
Hungary
Sweden
Portugal
Poland
Austria

Hong Kong

Indonesia
Germany

Italy
South Africa

Ireland

Turkey
Russia

Belgium
France
UK
China
Brazil
Denmark

Spain
United States

Taiwan
Singapore
Finland
Chile

Thailand
Egypt
New Zealand
Norway

Japan

Switzerland
years, and to the rest of the world (see Exhibit 3). But recent data
Source: Credit Suisse Trading Strategy suggest we should be seeing even tighter spreads. As the VIX has
Exhibit 4: Spreads vs. Spreads
Avg Bid-Ask VIX vsrecently
VIX diverging
trended lower in 2012 & 2013, spreads, especially for smallcaps, have
6 48 not kept pace (Exhibit 4).
Could falling HFT be behind this divergence?

Avg R2K Spread (bps) & VIX Level


5 40
Avg S&P Spread (bps)

4 32
We mentioned above how HFT were part of the major structural changes
that increased competition and led to narrower spreads. However, it
3 24 seems the field became saturated and compressed spreads to the point
2 16
that when volumes and volatility came down, there wasnt enough margin
to support the enlarged industry. Making matters worse, costs continued
1 8 to mount. Faced with these challenges, a number of HFT firms are
- 0
making cuts, switching to other asset classes, and, increasingly, shutting
down entirely over the same period as spreads have widened.
May-11

May-12
Jan-11

Mar-11

Jan-12

Mar-12

Nov-12

Jan-13

Mar-13
Jul-11

Sep-11

Nov-11

Jul-12

Sep-12

VIX SPX Spreads (left axis) R2K Spreads (right axis)

Source: Credit Suisse Trading Strategy

The relationship between spreads and volatility


Bid-ask spreads & volatility share a direct relationship. In the volatility of the stock, and the time it takes to get hedged
chart below, the darker color corresponds to a higher VIX level. (based on order book depth). The resulting payoff profile is
It is easy to see that the peaks are darker. similar to a sold call option.

This makes sense as higher volatility calls for wider spreads as Consequently, we can use a simple Black-Scholes options
market makers demand greater compensation for the additional model to proxy for spread pricing dynamics: time on the bid is
risk volatility brings. Conversely, when volatility & risk decline, we similar to Theta; the half spread you earn is the in-the-money-
should expect to pay lower spread costs. ness of the option.

Average bid-ask spreads by volatility regime


This simplified1 approach provides a mathematical way to look
10 at the fair value of spreads. Importantly, we see that spreads
9 should increase as volatility, price, or time to execute increase.
8
We also see that the fair spread for a stock may be well
Volatility
7
<15 above 1-tick especially for high priced stocks.
Bid-ask spread (bps)

6 15-20
5 20-30
30+
4

0
Jan-04

Jan-05

Jan-06

Jan-07

Jan-08

Jan-09

Jan-10

Jan-11

Jan-12

Jan-13

Source: Credit Suisse Trading Strategy.

Posted orders are like written options


Posted orders work a lot like options. The best a bid can do is
collect the spread (premium). But if the price moves up and
takes out the offer, the buyer will have to cover his legged
position. This will cost at least one spread, depending on the Source: Credit Suisse Trading Strategy.

1
The real world is even more complex, as execution time varies depending on liquidity at the time, order book priority, and the market makers covering strategy.
Maker-taker rebates and order book imbalances might also affect the actual moneyness of the spread. But the concept is the same posting gives a free option
to other participants, and that option has a probability of positive return based on volatility, among other things.

2
TRADING STRATEGY

Regulators Call for Wider spreads


Exhibit 5: Avg
Average shortfall
shortfall increases
by bid-ask with stockstrades)
spread (medium-sized spread Most of us would naturally think that tighter spreads are better. After all,
0 - 1 bps 1 - 10 bps 10 - 20 bps 20+ bps real investors are liquidity demanders and need to cross the spread to build
0
(avg=0.8bps) (avg=4bps) (avg=14bps) (avg=50bps)
a position.
-5
Wider spreads increase cost
-10
Real executions seem to support this view - according to our ExPRT TCA
avg shortfall (bps)

-15
-20
data, stocks with wider spreads face consistently increasing shortfall costs
-25
(Exhibit 5).
-30
But might they increase liquidity enough to offset this?
-35
Yet, regulators are examining the potential for wider spreads to increase
-40
liquidity in more thinly traded small cap names.
-45
-50 As part of the JOBS act, passed in April 2012, Congress authorized the
SEC to increase tick sizes for emerging growth companies under the
Note: medium sized trades shown; pattern is consistent across sizes belief that it would promote liquidity and encourage IPOs. The SEC is
Source: Credit Suisse Trading Strategy
currently holding roundtables to evaluate the impact of changing tick sizes.

Why would wider spreads increase liquidity?


The idea is based on the belief that a larger tick size translates to wider spreads, and that wider spreads will encourage more
market participants to try to capture the spread.
Most importantly, this also assumes that posted size is what defines better liquidity. The JOBS act does not consider depth
beyond the top-of-book.
The following example shows the how wider spreads should make the NBBO larger:
In the 1-tick market below, there are only 100 shares on both the $10.12 bid & $10.13 offer. It is likely that the stock that
would be bid at $10.11 and $10.10 would not be displayed as that creates unnecessary signaling.
With 5-cent spreads, the $10.12 bid would join the $10.10 bid. Also, the $10.11 and $10.10 bids that are currently
undisplayed would become lit at $10.10. In total there would now be 700 shares on the bid.
If the same happened on the offer, the spread would be 5 cents wide, and the offer size would be 900 shares.

Source: Credit Suisse Trading Strategy

Exhibit 6: Spreads naturally increase for small cap stocks


Small cap stocks already do have wider spreads
Based on our EDGE pre-trade, small cap stocks already have wider
spreads (Exhibit 6). On average, small cap stocks are more than 8x wider
than large cap stocks (Exhibit 4, page 2). Given their lower turnover and
typically higher volatility, this follows the intuition discussed in the box on
page 2, and supports the view that the market is naturally achieving wider
spreads on small stocks.

3
Source: Credit Suisse Trading Strategy
TRADING STRATEGY
Another path to wider spreads: lower prices
The JOBS act would achieve wider spreads by fixing larger tick sizes.
Exhibit 7: Lower priced stocks naturally have wider
However, there is another way to get to wider spreads: lower stock prices.
Spreads (basis points)
Falling price makes spread relatively larger
A basis-point spread is defined as the spread in cents / stock price. If the
spread in cents is constant, the basis-point spread will increase as the
price falls (Ex 7). This is typical of large, liquid stocks that already trade 1-
cent wide, since 1 cent is the minimum allowed tick size in the US.

Companies can self-serve!


Now that weve seen that lower prices will also widen spreads, we should
also point out that there is a way for companies who think wider spreads
would be preferred to accomplish it themselves: using stock splits.
Do the companies themselves want wider spreads?
Despite the fact that companies can increase spreads by cutting their
share price, companies have performed very few stock splits in the past
Note: dark lines represent the value of each tick (in bps) for each stock price.
1 tick results in a wider spreads (in bps) for lower priced stocks. few years (Exhibit 8). This suggests that, for whatever reason, companies
Source: Credit Suisse Trading Strategy like to have high-priced stocks. Undoubtedly, with US commissions paid
as cents-per-share, investors appreciate the savings. But how does it
actually affect execution quality?

Exhibit 8: Number of stock splits declining over time High priced stocks trade more than 1-tick wide anyway
Number of Stock Splits by Year
(S&P 500
(S&P stocks)
500 stocks) While low-priced stocks have wider spreads, Exhibit 7 shows that spreads
140
start to widen again when prices get very high (supporting our option
120 pricing model from page 2).
Number of Regular Stock Splits

100 So although companies resist stocks splits, the market still prices spreads
according to supply-demand and return dynamics. Exhibit 9 shows that
80
half of all stocks priced below $100 trade 1-penny wide, but for expensive
60 stocks, half trade 7 or more ticks wide.
40 Exhibit 10: $Value posted at the NBBO is larger for cheaper stocks
20

0
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013

Source: Credit Suisse Trading Strategy

Exhibit 9: Expensive stocks have a larger avg tick size


Average tick size of S&P 500 stocks
(S&P 500 stocks)
1
2 Source: Credit Suisse Trading Strategy
"regular"
stocks
1
50%
2
33%
3 4 More
510
8% 5%2%
68
1%
79
0% 3 Should we force wider spreads?
4 Weve seen that, in practice:
5
Smallcaps already have wider spreads
6
7 Lower priced stocks have wider spreads
stocks 2 3 4 5 6 7 8 9 10 More
priced 11% 7% 16% 11% 4% 9% 2%7% 11% 22% 8 Not all stocks trade 1-tick wide anyway
>$100
9
Companies have the ability to widen spreads themselves, if they so
10
choose
0% 20% 40% 60% 80% 100% More

% of stocks These facts raise the question of whether we need to force a process that
Source: Credit Suisse Trading Strategy seems to already happen naturally. Bid-ask spreads are meant to reflect
market conditions. Arbitrarily fixing them would create distortions and may
unnecessarily complicate the market.

4
TRADING STRATEGY

Exhibit Average
11: Stocks with wider
% of shares crossedspreads have
by bid-ask a higher
spread
Will It Work?
% crossed
(medium-sized trades) Wider spreads means it costs more to cross the spread and take liquidity.
25%
This should make traders more reluctant to pay the spread and encourage
them to post resting orders on the near touch instead. Hence the logic
20%
that wider spreads should increase posted size.
However, wide spreads may come with other undesirable consequences:
% of shares crossed

15%

1. Attracts HFT: If the spread is wide enough to be outside the fair


10% value discussed on page 2, it may become a haven for HFT who will
crowd the book in order to pick up spreads.
5%
2. Longer time to execute: With more traders and HFT now
trying to post on the near touch, the queue will increase and your
%
0 - 1 bps 1 - 10 bps 10 - 20 bps 20+ bps time to execute will go up. This is illustrated dramatically by
(avg=0.8bps) (avg=4bps) (avg=14bps) (avg=50bps)
Citigroup on the next page.
Note: medium sized trades shown; pattern is consistent across sizes
Source: Credit Suisse Trading Strategy 3. Increased signaling and market risk: A longer waiting time
means youre more exposed to risk both market risk (that the
market will move against you), and signaling risk (since your order
Exhibit 12: Wider spreads in HK cause volume to will be displayed for a longer time).
concentrate in the MOC when spreads are zero
4. More trading goes dark: To minimize signaling risk and also to
reduce transaction costs (dark pools often provide mid-point
executions), traders may then turn to dark pools. This is one of the
reasons dark pools were invented in the first place. Using our
ExPRT data, we find that stocks with wider spreads do trade more in
dark pools (Exhibit 11).
5. Less intraday liquidity: Volume curves in Asian markets
demonstrate how wider spreads affect trading patterns in the real
world. In Hong Kong, for example, tick sizes are so large and
queues so long that nearly 20% of all activity avoids regular
continuous trading and instead concentrates in the closing auction
(when spreads are zero. See Exhibit 12).

Source: Credit Suisse Trading Strategy


So the ironic consequence of an attempt to increase posted size may
actually be reduced intraday liquidity and a drive to the dark!

Is there another way to promote liquidity?


Some have suggested other possible ways to increase posted size but all of these have potential downsides as well.
Increasing the size of a round lot: The chart at right shows that the median Median Posted Size on the Bid/Ask (in round lots)
8
S&P 500 stock only shows around 5 lots on each side. For Russell 2000 stocks,
its only around 2-3 lots. With such small posted size, even a modest increase in 7
the minimum lot size from, say, 100 shares to 500 shares would force posted 6
size to increase for the many stocks that currently show less than 5 lots. 5
However, there are (at least) 3 potential problems with the approach: 4
1) This would adversely affect retail investors who already need over $80,000 3
to buy one lot of stocks like GOOG. 2
2) Market makers may be reluctant to commit more capital, so they may restrict 1
S&P 500
their quoting to fewer, more liquid names. Russell 2000
-
3) Given the higher costs from capital commitment, market makers may widen 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
spreads as compensation. This raises costs for investors. Source: Credit Suisse Trading Strategy

Creating a separate institutional market: In theory, if institutions feel comfortable knowing that theyre interacting with a defined
group of like-minded traders, they may be more willing to show larger size. In practice, these restricted communities have not panned
out. To cite one example, NYSEs New York Block Exchange recently shut down for lack of interest. This seems to imply that theyre
not the ideal solution either. Aside from not gaining traction, such a scheme would increase fragmentation and complexity as well.

5
TRADING STRATEGY

Exhibit 13: Citis posted size ballooned when price fell Citigroup: A case study in widening spreads
and reverted post-reverse-split
Citigroup provided a dramatic example of what happens in the real world
when spreads increase. In a highly-publicized collapse, the stock sank
from $50 to $1 in just two years. This caused the bid-ask spread to blow
out from 2bps to 100bps. Although the JOBS act would widen spreads
by increasing the minimum tick size rather than lowering prices, it is
reasonable to assume that the results would be the same. The fallout:
1. Posted size increased (even in $value): Just as the JOBS act
intends, posted size ballooned as spreads increased. Despite the
falling share price, NBBO size increased materially even in $value
terms (Exhibit 13).

Source: Credit Suisse Trading Strategy


2. Execution risk & signalling increased as waiting time went up:
Comparing the posted size to the daily value trading in C shows that it
took considerably longer for new bids to get to the top of the book as
Exhibit 14: Time to get through the Citigroup queue the spread expanded (Exhibit 14).
jumped to over 3 minutes
Before Citis low, it took around 5 seconds to get to the top of
600
the queue.
500 From May 2009-May 2011, with the book growing 100-fold,
the average wait time jumped to over 3 minutes, reaching as
400
high as 10 minutes at times.
seconds

300 This exposed investors to more execution risk and longer signalling.
200 3. C allegedly became a haven for HFT: A 100bps spread + 30mil
maker rebate (each side) offered by many exchanges made C a very
100 profitable reversion trade. And HFT, with their low latency, was best
equipped to achieve the price-time priority necessary to make the
-
trade work. This exaggerated the problems discussed in #2 above.
May-09
May-07

May-08

May-10

May-11

May-12
Jan-07

Jan-08

Sep-08
Jan-09

Jan-10

Jan-11

Jan-12

Jan-13
Sep-07

Sep-09

Sep-10

Sep-11

Sep-12

4. Dark trading increased: We also found that with a lower stock price
Source: Credit Suisse Trading Strategy and wider spread, volume shifted from the NYSE and Nasdaq to
crossing venues (represented by the TRF Exhibit 15). This makes
sense since a mid-point execution becomes more attractive as the
spread gets more expense. It also likely explains the fact that we saw
little change in average shortfall despite wider spreads (Exhibit 16).
5. All this reverted after the reverse split: When C completed their
1-for-10 reverse split, the spread reverted to 2 bps, book size and
queue length collapsed back to pre-crisis levels, and TRF volumes
receded.

Exhibit 15: More Citi volume crossed off-exchange Exhibit 16: Average C shortfall costs relatively constant
as price fell/spreads widened as more C: Average Shortfall
crossing Cost spreads
offset wider
20
10 for 1
15 reverse split
10

-5

-10

-15
Price hits $1
-20
Apr-09
Jan-08
Apr-08

Jan-09
Oct-08

Oct-09

Apr-10

Oct-10

Apr-11

Oct-11

Apr-12
Jan-10

Jan-11

Jul-11

Jan-12

Jul-12
Oct-12
Jan-13
Jul-08

Jul-09

Jul-10

Source: Credit Suisse Trading Strategy Source: Credit Suisse Trading Strategy
6
TRADING STRATEGY

Regulators ask for smaller spreads too!


Our study is complicated by the fact that there are other market
participants who also have strong arguments for switching to sub-penny
ticks that would narrow spreads.
Eliminating the latency advantage
One argument for narrower spreads is that it will remove the advantage
that low-latency traders have. Anyone wanting priority will only need to
offer stock inside the latest spread to create a new best offer.
Of course this argument doesnt consider potential system-wide issues
that non-integer prices might create, including needing more space to
show the best bid/offer prices and upgrading technology to be able to
handle the increased complexity. It is also possible that exchanges might
need to recode systems to accept more decimals in bids and offers.
Exchanges want it both ways
Ironically, exchanges support the JOBS act to widen tick sizes but
theyve also successfully lobbied the SEC to modify regulations that allow
them to use smaller ticks (sub-penny).
The motivation behind the new Retail Liquidity Program is to level the
playing field with dark pools and internalizers who are able to execute
trades between the NBBO. .
Exhibit 17: Japans tick sizes have artificially created Japan Exchange goes sub-yen
among the widest spreads globally Interestingly, the Japan Exchange Group recently announced the
implementation of a pilot program to reduce tick sizes to a tenth of current
levels. Proprietary trading systems akin to ATSs in the US already
employ these smaller increments, so the change would make them more
competitive with PTSs.

Beware of Unintended Consequences


There are compelling arguments to both widen tick sizes and reduce them.
However, it is extremely difficult to introduce targeted regulation given how
complex our market is.
Weve also shown that investor costs are at all-time lows, and that the
market already does a pretty good job of appropriately assessing risks and
pricing them accordingly.
Source: Credit Suisse Trading Strategy Regulation that forces different spreads for an arbitrary class of companies
might result in unintended consequences and increasing market
complexity. It may increase the percentage of stocks that trade in dark
pools or attract more HFT fighting for rebates.
Perhaps we should be more concerned with why bid-ask spreads have
started to widen, even as volatility continues to decline. Our data shows
that this is more likely to increase trade costs for real investors and at
the end of the day, isnt that what matters most?

7
TRADING STRATEGY

Disclaimer:
Please follow the attached hyperlink to an important disclosure: http://www.credit-suisse.com/legal_terms/market_commentary_disclaimer.shtml . 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 Corporations 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://creditsuisse.com/researchandanalytics/

2013, CREDIT SUISSE