Why do listed firms pay for market making
in their own stock?
Johannes Skjeltorpa
and
Bernt Arne Ødegaarda,b,c
a Norges Bank
b University of Stavanger (UiS)
c Norwegian School of Economics (NHH)
Jun 2012
Topic of paper
We study:
I
Designated Market Makers (DMM) at Oslo Stock Exchange
I
DMM (brokerage house/bank) hired by the firm against a fee.
I
ensure a liquid secondary market in firm’s stock
Main question in this paper:
I
Why are firms willing to pay a cost to improve the secondary
market liquidity of their issued shares?
Contribution:
I
Earlier studies ⇒ DMMs have a significant effect on market
quality
I
We look at DMMs from the firm’s perspective: what
determines the hiring choice?
Background - appearance of DMMs
Evolution towards electronic Limit Order Markets
I
No market makers, liquidity supplied by patient traders
I
Problem: low liquidity supply for some stocks (e.g. small-caps)
Response: Designated Market Makers (DMM)
I
I
listed firms can hire a financial intermediary (DMM) to
provide liquidity in its stock.
typical contract:
I
I
I
practice market making at least 85% of the day
maximum bid/ask spread of 4%
ensure minimum number of shares available at best quotes
But:
I
Why is it the listed firm that is paying?
Motivation: Firm decision
Corporate finance view:
Firm value: PV future cashflows X , discounted at cost of capital r :
Firm Value = V =
X
r
Hire a DMM: Pay an annual fee:
V =
X − Annual cost of DMM
r
Previous studies: Value of firm increases.
I This increase in V must come from either of
I
I
Cash flows
Cost of capital
New firm value calculation:
V =
X − Annual cost of DMM + Other cash flow consequences
r + Change in cost of capital
Firm’s perspective: Source of value of DMM?
Firm benefits from liquidity only when interacting with market
Potential direct effects on cash flow (X)
I
Reduced equity issuance costs (fees) [Butler et.al, 2005]
I
Lower costs of stock repurchases [Brockman et.al, 2008]
I
Lower direct cost of debt issuance [Butler/Wan, 2010]
Potential effect on discount rate (r)
I
Liquidity risk priced [e.g. Pastor/Stambaugh (2003),
Acharya/Pedersen (2005)]
I
DMM reduces liquidity risk loading ⇒ lower r .
But: this argument more from stock owner perspective...
Wedge between firm and owners: Transaction costs
Value of firm
V =
X
r
Value of position to an individual owner:
Value =
X
× Fraction of firm owned − Transaction costs
r
Improved liquidity following DMM
I
Lower transaction costs.
Not clear why this is relevant for the firm
I
short term traders benefit most
I
why subsidize short term trading/speculation?
However: One group of important owners for whom the firm may
care: Founders/venture capitalists.
May want to lower costs of exit at IPO lockup...
Our innovation relative to the literature
Previous studies on DMM initiations
I
Stockholm [Anand et.al, 2008], Deutsche Börse
[Hengelbrock, 2008], EuroNext [Menkveld/Wang, 2009],
Paris Bourse [Venkataraman/Waisburd, 2007]
By hiring a DMM, market quality can be directly influenced by the
firm that has issued the stock.
We treat the hiring of Designated Market Makers as endogenous
decisions, and ask:
I
What influences this decision?
The data
Oslo Stock Exchange (OSE) – Electronic limit order market, main
market for trading Norwegian stocks
I
DMMs allowed at the OSE from October 2004
I
look at DMM hirings from 2004 through 2010
the DMM is paid by the firm to “maintain an orderly market”
I
I
I
I
I
little info on actual costs
Anand et.al [2009], average fee ∼ USD 40k per year (Sweden)
Norway ∼ USD 30k per year
OSE monitors stocks with DMM to ensure that the DMM
fulfills obligations
Describing DMM deals at the OSE
Year
Total active stocks
Active DMM contracts
% of firms with DMM
2004
207
7
3%
Firm size (MCAP) groups
N(firms) in size quartile:
1 (small)
0
2
2
3
3
4 (large)
2
2005
238
30
13%
2006
258
42
16%
2007
292
50
17%
2008
286
57
20%
2009
263
47
18%
2010
235
48
20%
4
16
5
5
11
19
8
4
17
15
14
4
24
18
11
4
32
9
6
0
32
8
6
2
What is the effect on the market of hiring a DMM?
First, check that effect of DMM initiations is similar at the OSE as
other markets
1. Does liquidity improve?
I
I
liquidity significantly improves (6 month/1 year)
turnover and #days traded increases
2. Does the market react?
I
about 1% excess 5 day event return around announcement
Similar results in our sample as other studies on DMMs for other
markets.
Changes to average spread following DMM introduction
year after
Year before
DMM Stocks
DMM Stocks
16
30
14
25
12
20
10
8
15
6
10
4
5
2
0
0
0
0.02
0.04
0.06
0.08
0.1
0.12
0
0.02
0.04
0.06
0.08
0.1
0.12
Analyzing decision to hire DMM
What determines the decision to hire a DMM?
I
Decision theoretic empirical analysis (Probit)
I
Pr (Hire DMM)=f (likelihood of accessing market)
Determinants of the hiring decision:
I Likelihood of capital needs
I
I
I
Likelihood of stock repurchase (cash distribution)
I
I
I
ex-ante: growth potential (Tobin’s Q)
ex post: actual stock issuance
ex-ante: repurchase program announcements
ex-post: actual repurchases over next year
Other variables
I
I
I
insider transactions (ex-post exit motivation)
<2 years since listing (ex-ante exit motivation)
Pre-DMM liquidity
Probit analysis: Ex-ante variables
Model
Liquidity (RelSpread)
Q last year
Repurchase program
Listed < 2 years
Constant
N (firm/years)
Pseudo R2
New DMM hire
next year
(1)
(2)
-7.40∗∗
(0.03)
0.30∗∗∗
(0.00)
0.15
(0.49)
0.40∗∗
(0.03)
-1.33∗∗∗
(0.00)
368
0.12
.
.
0.30∗∗∗
(0.00)
0.15
(0.49)
.
.
-1.62∗∗∗
(0.00)
425
0.09
Probit analysis: Ex-post variables
Model
Liquidity (RelSpread)
Issue equity
Actual repurchase
Insider trades (sells)
Constant
N (firm/years)
Pseudo R2
New DMM hire
next year
(1)
(2)
-6.73∗
(0.06)
0.41∗∗
(0.03)
0.42∗∗
(0.03)
0.02
(0.34)
-1.26∗∗∗
(0.00)
329
0.07
.
.
0.36∗∗∗
(0.01)
0.35∗∗
(0.02)
.
.
-1.54∗∗∗
(0.00)
559
0.03
Asset pricing implications of DMMs
Asset pricing with liquidity risk
I
E.g. Pastor/Stambaugh (2003), Acharya/Pedersen (2005)
I
Norway: Næs/Skjeltorp/Ødegaard (2008,2009)
DMM effect on liquidity risk
I
Traditional market makers control their risk and inventory by
changing spread and/or price
I
DMMs unable to do this ⇒ take on liquidity risk that
otherwise would be present in the market
Implications:
I
reduction in liquidity risk loading for stocks that hire DMM
I
i.e. lower risk premium for DMM stocks
Does hiring a DMM affect liquidity risk?
Estimate a two-factor model adding liqudity factor LIQ to market
eri,t = αi + βim erm,t + βiliq LIQt + εi,t
DMM stocks, mean
DMM stocks, median
Liquidity beta (β liq )
Pre DMM Post DMM
0.114
-0.062
0.044
-0.022
Test for difference
change
p-value
-0.176∗∗∗
0.002
∗∗
-0.157
0.014
⇒ Significant drop in liquidity risk loading (βiliq )
⇒ liquidity risk loading closer to larger/liquid stocks after DMM
Economic significance of liquidity risk reduction
Is there a liquidity risk premium at the Oslo Stock Exchange?
Factor
erm
LIQ
GMM J-test
I
I
CAPM
λ[k] p-val.
0.014 (0.00)
J(χ2 (8))
24.47
p-val.
(0.00)
Two factor model
λ[k]
p-val.
0.0113 (0.00)
0.0119 (0.00)
J(χ2 (7))
9.26
p-val.
(0.16)
significant risk premium associated with liquidity
reduction in liquidity beta following a DMM hiring implies
I
I
2.5% lower expected returns for DMM stocks (annualized)
suggest an economically significant effect on cost of capital
Conclusion
Why pay for a DMM?
I
Secondary market liquidity matters to the firm because of the
market’s role when new capital is raised or distributed
I
Firms pay to improve liquidity when they plan on accessing
the stock market in the near future
Implications for asset pricing
I
liquidity risk loading drops, liquidity risk transferred to DMM
I
2.5% lower expected return (annualized)
⇒ suggest an economically significant effect on cost of capital
⇒ likely to cover the cost of having a DMM