Patient Capital Outperformance
The Investment Skill of High Active Share Managers
Who Trade Infrequently
Martijn Cremers
Ankur Pareek
University of Notre Dame
Rutgers Business School
1
Overview
1. Active management
– Active Share: Difference in holdings
– Historical Active Shares, by fund type and size
– Active Share and performance (updated and by style)
2. Patience: a rare skill?
– Fund Duration
– Active Share, patience and performance in 3 samples:
• Retail Mutual Funds (net returns)
• Aggregate Institutional Long-only Equity Holdings (13F)
• Aggregate Hedge Fund Long-only Equity Holdings (13F)
3. Conclusion
2
‘Active’ means ‘Different’
 Investors face basic choice between
i)
ii)

Actively managed funds
Passive benchmark (index fund or ETF)
What does ‘active’ mean?
 Basic definition: ‘active’ means different


Outperformance only possible if fund is different
 and fees not too large relative to overlap
How do we measure this?
3
Alternative Active Share formula
4
What is a ‘low’ Active Share?
• How many people can be ‘above average’?
– What percentage of benchmark assets can beat the
benchmark?
• Logic: exactly half the holdings
– Active funds: Active Share (clearly) above 50%
• Benchmark-specific: S&P 500 versus Russell 2000
– 100 random stocks from S&P 500, value-weight
– Result: Active Share of about 80%
– 100 random stocks from Russell 2000: Active Share of 95%
– Benchmark concentration / # of stocks / liquidity / etc.
5
Active Share Distribution of Large-Cap U.S. – Equity Funds
Note: LARGE CAP funds only (% TNA in Active Share deciles)
100%
90%
90%-100%
70%
80%-90%
70%-80%
60%
60%-70%
50%
50%-60%
40%-50%
40%
30%-40%
20%-30%
30%
10%-20%
20%
0-10%
10%
2013
2012
2011
2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
0%
1981
Percent of Total TNA in passive funds
80%
6
Active Share Distribution of Mid-Cap U.S. – Equity Funds
Note: MID CAP funds only (% TNA in Active Share deciles)
100%
90%
90%-100%
70%
80%-90%
70%-80%
60%
60%-70%
50%
50%-60%
40%-50%
40%
30%-40%
20%-30%
30%
10%-20%
20%
0-10%
10%
2013
2012
2011
2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
0%
1984
Percent of Total TNA in passive funds
80%
7
Active Share Distribution of Small-Cap U.S. – Equity Funds
Note: SMALL CAP funds only (% TNA in Active Share deciles)
100%
90%
Percent of Total TNA in passive funds
80%
70%
90%-100%
80%-90%
60%
70%-80%
60%-70%
50%
40%
50%-60%
40%-50%
30%-40%
30%
20%-30%
10%-20%
20%
0-10%
10%
0%
8
Cremers – Petajisto 2009 study
(data in study starts in 1990 and ended in 2003, updated to 2013 here)
For all actively managed all-equity retail mutual funds:
• Calculate Active Share at the end of this year
• Assign benchmark
– Based on self-declared benchmark if available
– Otherwise assigned based on holdings overlap
• Calculate fund’s net returns
– Over the next 12 months
– Benchmark-adjusted: deduct benchmark return
– Net: after deducting all fees & trading costs (except loads)
• Repeat everything at the end of next year
9
Updated results: 1990 – 2013
• At end of year, sort US-equity retail mutual funds into 5 groups
• Calculate benchmark-adjusted net returns over next 12 months
Abnormal performance per year in %
1.00
0.50
0.00
-0.50
Low Active
Share
Q2
Q3
Q4
High Active
Share
-1.00
-1.50
Alpha p/y
Bmk-Adj Return p/y
10
AS Q1
AS Q2
AS Q3
AS Q4
AS Q5
1-May-13
1-Jul-12
1-Sep-11
1-Nov-10
1-Jan-10
1-Mar-09
1-May-08
1-Jul-07
1-Sep-06
1-Nov-05
1-Jan-05
1-Mar-04
1-May-03
1-Jul-02
1-Sep-01
1-Nov-00
1-Jan-00
1-Mar-99
1-May-98
1-Jul-97
1-Sep-96
1-Nov-95
1-Jan-95
1-Mar-94
1-May-93
1-Jul-92
1-Sep-91
1-Nov-90
1-Jan-90
Cumulative benchmark-adjusted net returns
for U.S. equity retail mutual funds
$1.30
$1.25
$1.20
$1.15
$1.10
$1.05
$1.00
$0.95
$0.90
$0.85
$0.80
11
Small-cap versus Large-cap Funds, 1
• Separately sort small-cap and large-cap funds into Active Share quintiles
• Active Share predicts future performance for both samples
Large Cap (Alpha in % per year)
0.25
0.00
-0.25
-0.50
Low Active
Share
Q2
Q3
Q4
High Active
Share
-0.75
-1.00
-1.25
-1.50
12
Small-Cap versus Large-Cap Funds, 2
• Separately sort small-cap and large-cap funds into Active Share quintiles
• Active Share predicts future performance for both samples
Small Cap (Alpha in % per year)
2.5
2
1.5
1
0.5
0
-0.5
Low Active
Share
Q2
Q3
Q4
High Active
Share
13
Why would Active Share
relate to future performance?
• Successful managers need
– Skill, guts and opportunity
– Don’t need skill to have a high Active Share
• Do need guts and opportunity
• Suppose: I have a lot of guts and opportunity
– But no skill…: How long would I survive?
• Skill: hard to measure, attracts flows and is expensive
• Is skill scalable? Time-varying? Transferable? Rare?
• What is a rare skill?
14
Rare skill: patience?
(with high Active Share)
• Fund Duration
– average length of time an average dollar in the
portfolio has been in the portfolio (in the last 5 years)
• Robustness: turnover & holdings-turnover
• 3 samples:
1. retail mutual funds – net returns
2. all institutional portfolios (13Fs) – quarterly holdings
3. hedge funds (13Fs) – (long-only) quarterly holdings
15
Year
2012
2011
2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
Percentage of Assets
Percentage of Assets in Institutions
by Active Share Range
100%
90%
80%
70%
60%
50%
>90%
80%-90%
40%
70%-80%
30%
60% - 70%
< 60%
20%
10%
0%
16
Year
2012
2011
2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
Percentage of Institutions
Percentage of Number of Institutions
by Active Share Range
100%
90%
80%
70%
60%
50%
>90%
80%-90%
40%
70%-80%
30%
60% - 70%
< 60%
20%
10%
0%
17
Evolution of Duration over Time
(Stock) Duration, Institutional Turnover, (Inverse) Share Turnover, and
Institutional Investor Holdings from 1985-2010
1.60
100%
90%
1.20
70%
60%
1.00
50%
0.80
40%
30%
0.60
20%
0.40
10%
0.20
Stock Duration (left axis)
(Inverse) Share Turnover (left axis)
Institutional Investor Holdings (right axis)
Institutional Turnover (right axis)
2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
0%
1985
Stock Duration (in years) and
(Inverse) Share Turnover (in years)
80%
Institutional Turnover and
Institutional Investor Holdings (in %)
1.40
18
How short-term are institutions?
Fund Duration for Selected Investors
(Weighted Average Duration Across all Stocks of an Investor)
5
4
3.5
3
2.5
2
1.5
1
0.5
Berkshire Hathaway
Fidelity
STRS Ohio
U Chicago
2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
0
1985
Institutional Duration (in years)
4.5
19
Percentage of mutual funds
by Fund Duration
100%
> 3 years
90%
80%
70%
2- 3 years
60%
50%
1 - 2 years
40%
30%
6 mon - 1 year
20%
10%
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
0%
< 6 months
20
Examples of Mutual Funds with High/Low Active
Share and Fund Duration
21
Abnormal returns of duration-quintile
portfolios of mutual funds by Active Share
3.00%
2.00%
Mutual Funds:
Annualized 5-factor alphas
1.00%
0.00%
Q1 Duration Q2 Duration Q3 Duration Q4 Duration Q5 Duration
-1.00%
-2.00%
Q1 Active Share
-3.00%
Q5 Active Share
22
Average Cumulative Abnormal Net Returns for $1
Investment in Portfolios of Mutual Funds
$1.50
$1.30
$1.20
$1.10
$1.00
$0.90
$0.80
$0.70
Low Active Share/Low Fund Duration
Low Active Share/High Fund Duration
High Active Share/Low Fund Duration
High Active Share/High Fund Duration
2013
2012
2011
2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
$0.60
1994
Performance of $1 Initial Investment
$1.40
Year
23
Explaining the outperformance of patient mutual
funds with high Active Share
Annualized alpha of Q5-Active-Share & Q5-Fund-Duration portfolio
2.50%
2.00%
Annualized Alpha
1.50%
1.00%
0.50%
0.00%
Net
benchmarkadjusted
returns
Series1
1.92%
5-factor alpha
of net bmkadj return
add only BaB
add only QmJ
add both BaB
and QmJ
2.30%
1.91%
0.80%
0.59%
24
Percentage of institutions by Fund Duration
100%
> 3 years
90%
80%
70%
2- 3 years
60%
50%
1 - 2 years
40%
30%
6 mon - 1 year
20%
10%
< 6 months
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
0%
25
Percentage of institutional assets
by Fund Duration
100%
90%
> 3 years
80%
70%
2- 3 years
60%
50%
1 - 2 years
40%
30%
6 mon - 1 year
20%
10%
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
0%
< 6 months
26
Examples of Institutional Portfolios with High/Low
Active Share and Fund Duration
27
Abnormal returns of duration-quintile
portfolios of institutional portfolios by Active Share
1.50%
1.00%
Institutional Portfolios:
Annualized 5-factor alphas of 13F holdings
0.50%
0.00%
Q1 Duration Q2 Duration Q3 Duration Q4 Duration Q5 Duration
-0.50%
-1.00%
Q1 Active Share
-1.50%
Q5 Active Share
28
Average Cumulative Abnormal Holdings-based Returns
for $1 Investment in Portfolios of Institutional Holdings
$2.00
$1.60
$1.40
$1.20
$1.00
Low Active Share/Low Fund Duration
High Active Share/Low Fund Duration
Low Active Share/High Fund Duration
High Active Share/High Fund Duration
2012
2011
2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
$0.80
1985
Performance of $1 Initial Investment
$1.80
Year
29
Cumulative Abnormal Holdings-based Returns for $1 Investment
in Portfolios of the Holdings of Active and Long-term Institutions
$2.00
$1.50
$1.25
$1.00
$0.75
BERKSHIRE HATHAWAY
EARNEST PARTNERS
GARDNER RUSSO & GARDNER
RUANE CUNNIFF & GOLDFARB
SPO PTNR & CO.
COLUMBIA WANGER ASSET MGMT
2012
2011
2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
$0.50
1985
Performance of $1 Initial Investment
$1.75
Year
30
Percentage of hedge funds
by Fund Duration
100%
> 3 years
90%
2- 3 years
80%
70%
1 - 2 years
60%
50%
6 mon - 1
year
40%
30%
< 6 months
20%
10%
2011
2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
0%
31
Examples of Hedge Fund Portfolios with High/Low
Active Share and Fund Duration
32
Abnormal returns of duration-quintile
portfolios of Hedge Funds by Active Share
4.00%
3.50%
Hedge Funds:
Annualized 5-factor alphas of 13F holdings
3.00%
2.50%
2.00%
1.50%
1.00%
0.50%
0.00%
-0.50%
-1.00%
Q1 Duration Q2 Duration Q3 Duration Q4 Duration Q5 Duration
T1 Active Share
T3 Active Share
33
Average Cumulative Abnormal Gross Holdings-based Returns
for $1 Investment in Portfolios of Hedge Funds Holdings
$2.50
$2.00
$1.75
$1.50
$1.25
$1.00
2012
2011
2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
$0.75
1994
Performance of $1 Initial Investment
$2.25
Year
Low Active Share/Low Fund Duration
Low Active Share/High Fund Duration
High Active Share/Low Fund Duration
High Active Share/High Fund Duration
34
Results robust to…
• Alternative patience proxies
– turnover (mutual funds)
– holdings-based turnover (for institutions and hedge
funds)
• Controlling for flows / size / year f.e. / etc.
– In pooled panel regressions
• Controlling for longer-time changes in turnover
• BaB (Betting against Beta) and QmJ (Quality minus Junk)
– Cannot explain results for institutions & hedge funds
35
Conclusion
• Active Share
– Low Active Share funds: tended to underperform
– High Active Share funds: tended to outperform
• Large cap: high Active Share did not underperform
• Small cap: high Active Share outperformed
• Patient funds with high Active Share outperformed
– Mutual funds
– Institutions (holdings-based, long-only, 13Fs)
– Hedge funds (holdings-based, long-only, 13Fs)
• Impatient funds (had holdings which) underperformed
– Low Active Share funds: More underperformance with
more frequent trading
36
How active is your fund?
And how patient?
37