The buck
stops here:
Active-Passive
Debate:
Vanguard
money market
Market Cyclicality
and funds
Leadership Volatility
Vanguard research
Christopher B. Philips, CFA; Francis M. Kinniry Jr., CFA; David J. Walker, CFA
■■
In this update to our original 2009 paper, we reexamine the active-versus-index debate
from the perspective of market cyclicality and provide context for the changing nature
of performance leadership.
■■
We show that when evaluating the performance of active managers versus a benchmark
index, investors should be acutely aware of the differences in the managers’ strategies
involving factors such as size (market capitalization), style (price/earnings ratio and
price/book ratio), and relative positioning.
■■
We show that the market environment can have a greater impact on relative performance
than manager skill or even cost differences.
■■
Most important, we show that during periods of significant performance deviation
between opposing market segments (for example, large- and small-capitalization, or
growth and value), active managers will produce a wider distribution of returns showing
more pronounced performance differences relative to the market.
July 2014
Setting the stage: Is ten years long enough?
Today, a majority of investors would probably consider
ten years to be a long-term investment horizon. However,
even over that length of time, historical trends don’t
always hold true. For example, although stocks have
outperformed bonds and cash over the very long term,
they have lagged bonds in 17 of 79 rolling ten-year
periods since 1926 and even trailed cash 12 times.
The performance of active funds relative to a broad
market benchmark index can be similarly volatile.
As Figure 1 demonstrates, in the ten years ended
December 31, 1999, 71% of active managers underperformed the U.S. stock market. But during the decade
ended December 31, 2008, 37% lagged, a change of
34 percentage points over nine years. We see a further
shift when looking at the ten years ended 2013, when
55% of active managers underperformed. This volatility
not only clearly implies that ten years is not long enough
to be considered “long term,” it also raises the question
of what exactly may be contributing to these swings in
performance leadership.
Notes about risk and performance data: Investments are subject to market risk, including the possible loss of the money
you invest. Bond funds are subject to the risk that an issuer will fail to make payments on time and that bond prices will
decline because of rising interest rates or negative perceptions of an issuer’s ability to make payments. Past performance
is no guarantee of future returns. Prices of mid- and small-cap stocks often fluctuate more than those of large-company
stocks. Diversification does not ensure a profit or protect against a loss in a declining market. Note that hypothetical
illustrations are not exact representations of any particular investment, as you cannot invest directly in an index or
fund-group average.
2
Figure 1. Performance leadership can shift over ten-year periods
a. Distribution of active manager net excess returns versus benchmark: Ten years ended December 31, 1999
90
Underperformed:
Number of funds
80
Outperformed:
U.S. stock market return
452 funds (71%)
188 funds (29%)
70
60
50
40
30
20
10
0
< –8% –8%
to
–7%
–7%
to
–6%
–6%
to
–5%
–5%
to
–4%
–4%
to
–3%
–3%
to
–2%
–2%
to
–1%
–1%
to
0%
0%
to
1%
1%
to
2%
2%
to
3%
3%
to
4%
4%
to
5%
5%
to
6%
6%
to
7%
7%
to
8%
> 8%
Annualized excess returns
b. Distribution of active manager net excess returns versus benchmark: Ten years ended December 31, 2008
350
Number of funds
300
Underperformed:
Outperformed:
U.S. stock market return
1009 funds (37%)
1682 funds (63%)
250
200
150
100
50
0
< –8% –8%
to
–7%
–7%
to
–6%
–6%
to
–5%
–5%
to
–4%
–4%
to
–3%
–3%
to
–2%
–2%
to
–1%
–1%
to
0%
0%
to
1%
1%
to
2%
2%
to
3%
3%
to
4%
4%
to
5%
5%
to
6%
6%
to
7%
7%
to
8%
> 8%
Annualized excess returns
c. Distribution of active manager net excess returns versus benchmark: Ten years ended December 31, 2013
900
Number of funds
800
Underperformed:
Outperformed:
U.S. stock market return
2057 funds (55%)
1684 funds (45%)
700
600
500
400
300
200
100
0
< –8% –8%
to
–7%
–7%
to
–6%
–6%
to
–5%
–5%
to
–4%
–4%
to
–3%
–3%
to
–2%
–2%
to
–1%
–1%
to
0%
0%
to
1%
1%
to
2%
2%
to
3%
3%
to
4%
4%
to
5%
5%
to
6%
6%
to
7%
7%
to
8%
> 8%
Annualized excess returns
Sources: Vanguard and Morningstar.
Notes: This comparison evaluates active funds after costs against a market benchmark index that incurred no costs. It excludes sector funds, specialty funds, and real estate funds.
See Appendix for list of benchmarks used.
3
Figure 2. Size and style factors can play a role in leadership shifts
Cumulative performance spread
The historic spreads between large- and small-capitalization and between growth and value
80%
60
Growth outperforms value;
large outperforms small
40
20
0
–20
Value outperforms growth;
small outperforms large
–40
–60
1988
1992
1996
2000
2004
2008
2012
Growth-value differential: Ten years
Large-small differential: Ten years
Source: Vanguard. Data through December 31, 2013.
See Appendix for list of benchmarks used.
4
Figure 3. Hypothetical impact of style box dispersion
on market returns
10%
10%
5%
5%
Market: 17.59%
Large
35%
Growth
22%
17%
Mid
35%
Value
18%
15%
Small
Returns
Growth
Large
Weights
Value
Mid
The extreme performance spread and reversal in segment
dominance in 2000 are important for two main reasons.
First, large-caps typically account for close to 70% of the
capitalization of the aggregate market; mid-caps make
up approximately 20%, and small-caps 10%. So it is no
surprise that the market will realize a total return most
similar to that of large-caps.
When returns are so widespread, large-caps play
an even greater role in market and fund manager
performance. For example, Figure 3 shows a
hypothetical scenario in which large-cap value stocks
outperform large-cap growth stocks by 500 basis points.
Small
One widely noted change from the ten-year period
ended 1999 to the decade ended 2008 was the nearly
simultaneous shift in performance leadership from
growth stocks to value stocks and from larger stocks
to smaller. As the large-cap growth bull market of the
late 1990s ended, small-cap and value stocks began
to outperform. Figure 2 demonstrates the magnitude
of this change. The orange line, representing the cumulative
ten-year performance spread between value and growth
stocks, shows a reversal from 2000 to 2008 of the trend
established in the 1990s, when growth dominated value.
The yellow line representing the cumulative ten-year spread
between large-cap and small-cap stocks shows a similar
shift. By the end of 2008, value had outpaced growth
over the previous ten years by a cumulative 35%, and
small-caps had outpaced large-caps by a cumulative 43%.
Interestingly, in the subsequent five years, although
growth has rebounded, small-caps have continued to
outpace large-caps.
16%
13%
Market return: 18.40%
This hypothetical example does not represent the return on any particular investment.
Source: Vanguard.
Figure 4. Relative ranking of style box total returns
Mid
13.82%
17.32%
18.96%
14.04%
13.40%
11.81%
Market: 17.59%
0.68%
–0.97%
–3.20%
4.57%
4.00%
–0.43%
4.94%
3.15%
2.86%
Market: –0.63%
Value
Blend
Growth
Large
20.60%
Growth
7.53%
7.85%
8.09%
Mid
18.13%
Blend
Ten years ended 2013
10.66%
10.13%
9.31%
Small
Large
15.36%
Value
Large
Growth
Mid
Blend
Ten years ended 2008
Small
Value
Small
Ten years ended 1999
9.67%
10.27%
10.66%
Market: 8.11%
Sources: Vanguard, MSCI, and CRSP.
See Appendix for list of benchmarks used.
Because of the distribution of weights across the six
style boxes, large-cap value is the only segment that
beats the overall market. In such circumstances, active
managers in this segment would probably find it easier
to also outperform the market. Those in the remaining
style boxes would face a headwind beyond their control—
an environment in which their style is decidedly out of
favor and lags the market.
As Figure 4 shows, in the late 1990s, the performance
of large-cap growth stocks far exceeded those of largecap value and most others. As a result, the market itself
outperformed most individual segments—including
large-cap value—and, by extension, a larger portion of
active managers.
Of course, the opposite can happen as well, as we
saw during the ten years through the end of 2008.
In this period, large-caps in general underperformed
smaller stocks. In direct contrast to 1999, large-cap
growth dramatically lagged every other style, including
large-cap value. This very poor performance depressed
the return of the overall market to the point that even
large-cap value outperformed it by more than 130 basis
points. In contrast to the headwind they faced in 1999,
small-cap value managers benefited from a significant
tailwind. In the most recent period, although the market
still marginally beat each of the large-cap segments,
the return spreads across styles were much tighter,
so manager performance experienced fewer heador tailwinds.
5
Figure 5. The relative performance of all managers depends on the relative performance of market segments
a. Distribution of active manager net excess returns versus market benchmark: Ten years ended December 31, 1999
120
Number of funds
100
80
60
Underperformed (#, %):
Large blend (119, 74%)
Large growth (32, 32%)
Large value (123, 95%)
Mid blend (43, 83%)
Mid growth (36, 49%)
Mid value (35, 95%)
Small blend (17, 89%)
Small growth (21, 51%)
Small value (26, 96%)
Outperformed (#, %):
Large blend (41, 26%)
Large growth (68, 68%)
Large value (7, 5%)
Mid blend (9, 17%)
Mid growth (38, 51%)
Mid value (2, 5%)
Small blend (2, 11%)
Small growth (20, 49%)
Small value (1, 4%)
U.S. stock market return
40
20
0
< –8% –8%
to
–7%
–6%
to
–5%
–7%
to
–6%
–5%
to
–4%
–4%
to
–3%
–1%
to
0%
–2%
to
–1%
–3%
to
–2%
0%
to
1%
1%
to
2%
2%
to
3%
3%
to
4%
4%
to
5%
6%
to
7%
5%
to
6%
7%
to
8%
> 8%
Annualized excess returns
Number of funds
b.
Large blend
Mid blend
Small blend
Distribution
of growth
active manager
net excess returns
versus
Large
Mid growth
Small growth
Large value
Mid value
Small value
Underperformed (#, %):
400
300
200
market benchmark: Ten years ended December 31, 2008
Large blend (301, 60%)
Large growth (436, 75%)
Large value (88, 20%)
Mid blend (15, 11%)
Mid growth (99, 29%)
Mid value (0, 0%)
Small blend (3, 2%)
Small growth (66, 23%)
Small value (1, 1%)
Outperformed (#, %):
Large blend (199, 40%)
Large growth (143, 25%)
Large value (360, 80%)
Mid blend (126, 89%)
Mid growth (237, 71%)
Mid value (81, 100%)
Small blend (190, 98%)
Small growth (224, 77%)
Small value (122, 99%)
U.S. stock market return
100
0
< –8% –8%
to
–7%
–6%
to
–5%
–7%
to
–6%
–5%
to
–4%
–4%
to
–3%
–2%
to
–1%
–3%
to
–2%
–1%
to
0%
0%
to
1%
1%
to
2%
2%
to
3%
3%
to
4%
4%
to
5%
7%
to
8%
6%
to
7%
5%
to
6%
> 8%
Annualized excess returns
Large blend
Mid blend
Small blend
Large value
Mid value
Small value
c. Distribution
of growth
active manager
net excess returns
versus market benchmark: Ten years ended December 31, 2013
Large
Mid growth
Small growth
900
Number of funds
800
700
600
500
400
Underperformed (#, %):
Large blend (537, 81%)
Large growth (596, 67%)
Large value (502, 81%)
Mid blend (70, 44%)
Mid growth (125, 30%)
Mid value (36, 19%)
Small blend (85, 29%)
Small growth (82, 22%)
Small value (24, 16%)
Outperformed (#, %):
Large blend (123, 19%)
Large growth (290, 33%)
Large value (118, 19%)
Mid blend (88, 56%)
Mid growth (291, 70%)
Mid value (149, 81%)
Small blend (211, 71%)
Small growth (284, 78%)
Small value (130, 84%)
U.S. stock market return
300
200
100
0
< –8% –8%
to
–7%
–7%
to
–6%
–6%
to
–5%
–5%
to
–4%
–4%
to
–3%
–3%
to
–2%
–2%
to
–1%
–1%
to
0%
0%
to
1%
1%
to
2%
2%
to
3%
3%
to
4%
4%
to
5%
5%
to
6%
Annualized excess returns
Large blend
Small blend
Large growth
Small growth
Large value
Small value
Mid blend
Sources: Vanguard and Morningstar.
See Appendix for list of benchmarks used.
6
Mid growth
Mid value
6%
to
7%
7%
to
8%
> 8%
This brings us to the second reason to note the extreme
performance spread and the shift in dominance between
large-cap growth stocks and small-cap value: the equal
weighting methodology for calculating the percent of
funds out- or underperforming. To illustrate the impact
of the equal weighting methodology, Figure 5 breaks
the distributions in Figure 1 into the nine style-box
components (large-cap, mid-cap, and small-cap, and
growth, blend, and value) for the three ten-year periods.
In line with the trends shown in Figure 2, we see a
wholesale shift in the style of fund that outperformed
the market from the ten years ended 1999 to the decade
ended 2008. Indeed, although 96% of small value funds
underperformed over the first period, only 1% lagged
over the second.1
Because we are counting each fund to calculate the
percentage that outperformed the market, the actual
number of funds in each style box in each time period is
just as important as the performance of the style itself.
For example, as of 1999, 61% of all funds were large-cap,
25% were mid-cap, and 14% were small-cap. In 2008,
the percentages had shifted to 56% large-cap, 21% midcap, and 23% small-cap. As a result, although 63% of
active managers beat the broad market over the ten years
ended December 31, 2008, most of that success can be
attributed directly to the performance of value and smallcap stocks combined with the outsized growth in the
number of small-cap funds. Likewise, the significant
underperformance of active managers over the decade
ended December 31, 1999, was largely due to the
performance of large-cap growth stocks and the large
portion of funds in the large-cap blend and large-cap
value style boxes. The takeaway, then, is that during
periods of notable deviation in performance between
opposing market segments (such as large and small,
or growth and value), the distribution of returns among
active managers will be much more pronounced. This
was the case in the 1990s and has been the case in
the 2000s. During a prolonged period of less severe
deviations, we would expect fund styles to have much
less of an impact and costs to be a major component
of relative returns.
1 In analysis not shown here, we also examined markets outside of the United States and found a similar pattern. Through December 2008, 70% of small- and mid-cap funds outperformed a
broad-market benchmark index, but only 25% of large-cap funds did the same.
7
Figure 6. Volatility of outperformance also occurs within style boxes
Percentage of active managers that underperformed style benchmarks
Mid
76%
83%
55%
81%
58%
7%
48%
55%
40%
36%
69%
31%
42%
27%
67%
Value
Blend
Growth
Large
76%
Growth
67%
77%
66%
Mid
79%
Blend
Ten years ended 2013
% underperforming style benchmark
93%
87%
51%
Small
Large
78%
Value
Large
Growth
Mid
Blend
Ten years ended 2008
% underperforming style benchmark
Small
Value
Small
Ten years ended 1999
% underperforming style benchmark
64%
78%
87%
Sources: Vanguard and Morningstar.
See Appendix for list of benchmarks used.
Digging deeper into style box performance cyclicality
In addition to influencing the perception of how active
managers perform in comparison with the overall market,
the relative performance of one style versus another may
also affect how we evaluate active managers against
their style benchmark index. Even within more narrowly
defined market segments, we have seen significant
volatility in the distribution of returns for active managers
around a given benchmark.
Figure 6 evaluates the relative performance of active
managers versus their style-specific benchmark index
for the ten-year periods ended 1999, 2008, and 2013.
Clearly, the volatility in the percentage of outperforming
funds observed in Figure 5 is also present within each
style box. For example, over the decade ended 1999,
76% of large-cap growth managers lagged their bench­
mark. But over the ten years ended 2008, only 40%
lagged. Although it is possible that these managers
suddenly became more skilled at picking stocks, it is
far more likely that the volatility reflects differences in
how they built their portfolios and the dynamics of the
large-cap growth market over time.
To generate a return higher than that of a benchmark
index, an active manager’s portfolio must differ from
that benchmark in some respect. A manager may choose
to hold more or fewer stocks, to hold them at different
8
weights, or, more likely, to resort to some combination
of the two. For example, a large-cap value manager
holding 50 stocks in equal proportions (2% per stock)
may be measured against a market cap-weighted largecap value benchmark with more than 300 names. The
degree to which the fund’s holdings and weightings
differ from those of the benchmark and the distribution
of winners and losers within the benchmark will dictate
the fund’s relative performance. As shown in Figure 7,
regardless of time period, active managers in any market
segment are a heterogeneous group, characterized by
wide differences in basic valuation metrics such as size
(market capitalization) and style (price/earnings ratio).
This analysis provides several key takeaways. First, and
most critical to the importance of cyclicality to relative
performance, the median statistics for funds differ in
some respects from those for their benchmarks. For
example, in 1999, median market capitalization was
$24.23 billion for large-cap value funds and $31.76 billion
for the benchmark index. More than 50% of large-cap
value funds had a smaller market cap than the benchmark
median. All else being equal, then, during periods of largecap outperformance such as 1999, the benchmark index
will tend to beat a majority of active managers simply
because of the difference in median market capitalization.
Across the time periods studied, the median for each
fund style consistently differed from that of its benchmark
index in P/E ratio, market cap, or both.
Figure 7. Peer-group fundamentals show wide dispersion
Fund statistics as of December 31, 1999
P/E ratio
Market
capitalization
($B)
Percentile
P/E ratio
Market
capitalization
($B)
P/E ratio
Market
capitalization
($B)
95%
median
5%
24.68
19.58
14.82
$52.20
24.23
12.55
20
$31.76
Large
95%
median
5%
49.53
36.87
26.01
$110.88
51.32
16.29
52
$125.57
95%
median
5%
19.42
15.62
12.31
$11.68
6.61
2.21
NA
NA
Mid
95%
median
5%
49.47
31.30
20.12
$10.17
4.34
2.17
NA
NA
95%
median
5%
14.56
12.66
10.23
$1.31
0.57
0.16
17.4
$0.66
95%
median
5%
40.60
23.14
16.86
$1.72
1.13
0.39
62.4
$1.10
Percentile
P/E ratio
Capitalization
Market
capitalization
($B)
Small
Market benchmark
Large
Growth funds
Mid
Market benchmark
Small
Capitalization
Value funds
Fund statistics as of December 31, 2008
95%
median
5%
14.27
10.66
8.27
$52.32
30.81
10.68
9.9
$39.47
95%
median
5%
13.81
10.35
6.78
$10.77
4.44
1.51
12.4
$4.15
95%
median
5%
14.64
11.60
7.68
$1.52
0.81
0.21
13.7
$1.00
P/E ratio
P/E ratio
Market
capitalization
($B)
Large
Percentile
Market benchmark
Market
capitalization
($B)
95%
median
5%
18.93
13.86
10.67
$45.33
27.26
9.88
13.2
$31.46
Mid
Large
P/E ratio
Market
capitalization
($B)
Capitalization
P/E ratio
Growth funds
95%
median
5%
21.59
15.56
10.56
$9.26
4.22
1.67
11.7
$4.80
Small
Percentile
Mid
Market benchmark
Market
capitalization
($B)
Small
Capitalization
Value funds
95%
median
5%
22.86
15.62
10.96
$1.90
1.06
0.33
17.8
$1.04
Fund statistics as of December 31, 2013
95%
median
5%
18.66
16.24
13.77
$79.56
45.65
18.78
16.8
$71.89
95%
median
5%
20.43
17.79
14.04
$12.47
7.70
3.88
20.6
$9.36
95%
median
5%
21.53
17.22
13.44
$2.86
1.49
0.30
23.1
$2.94
P/E ratio
P/E ratio
Market
capitalization
($B)
Large
Percentile
Market benchmark
Market
capitalization
($B)
95%
median
5%
29.42
22.46
18.05
$72.53
42.94
17.56
24.7
$56.07
Mid
Large
P/E ratio
Market
capitalization
($B)
Capitalization
P/E ratio
Growth funds
95%
median
5%
29.77
24.99
18.25
$14.54
7.93
3.75
31.2
$10.69
Small
Percentile
Mid
Market benchmark
Market
capitalization
($B)
Small
Capitalization
Value funds
95%
median
5%
33.23
25.22
19.08
$3.27
1.99
0.63
41.8
$3.11
Sources: Vanguard and Morningstar.
See Appendix for list of benchmarks used.
9
Second, each style box contains a substantial difference
in medians between the funds with the largest market
cap and highest P/E ratio and those with the smallest
market cap and lowest P/E ratio. For example, in 1999,
the market-cap spread among large-cap value funds was
$40 billion. In other words, the top 5% of funds were
exposed to stocks with a median market capitalization
more than four times as large as that for the stocks the
bottom 5% were exposed to. In a period when the largest
stocks outperformed (all else being equal), funds with
significant exposure to those stocks would outperform
those with exposure to smaller stocks, even within the
same market segment.
Third, the top 5% of large-cap value funds in 1999
also maintained a larger profile than the index’s. Their
median market cap was $52.20 billion, while the market
benchmark’s was $31.76 billion. Again, all else being
equal, at a time when the largest stocks outperformed,
those funds with more exposure to large stocks would
beat the index, and those with less exposure would lag.
The differences between the funds and their benchmarks
are not as notable for the median P/E ratio as for the
market cap. We do see some differences, however,
particularly among small-caps. In 1999, 2008, and 2013,
the median P/E ratio for small-cap managers was lower
than for the benchmark index, particularly within small-cap
growth. If this scenario representing a tilt toward value
exists during a period when growth dominates value, we
would expect a majority of the funds to underperform.
10
The spread between the 5% of funds with the highest
median P/E ratio and the 5% with the lowest is also
significant. In fact, among large-cap value stocks in 1999,
the spread was nearly double: 24.68x versus 14.82x.
Naturally, this metric would be of interest to a value
investor, as low and high P/E stocks would be expected
to perform quite differently in this environment. Because
the highest 5% of funds were exposed to stocks with
higher P/E ratios than the benchmark median, in a growthdominated market, these funds would be expected to
outperform (all else being equal). On the other hand, the
lowest 5% of funds, with a median P/E lower than the
benchmark median, would be expected to outperform
in a value-dominated market. As with market cap, the
valuation dispersion within a particular style box can
lead to very different performance among similar funds.
We have looked at the large-cap value segment to this
point, but based on the statistics presented in Figure 6,
large-cap growth stocks may be of particular interest.
In 1999, not only did this segment show very significant
dispersion across market caps and P/E ratios, the bench­
mark index also was characterized by a larger market cap
and P/E than those for virtually all the funds in the sample.
This situation, combined with the fact that 1999
represented the peak of the bull market dominated by
large-cap growth stocks, makes it no surprise that the
benchmark outperformed 76% of large-cap growth
managers over the ten years ended 1999. On the other
hand, it’s easy to see how, when small-cap and value
subsequently dominated, a majority of active managers
were able to outperform the large-cap growth benchmark
over the decade ended 2008.
Figure 8. Measuring the cyclicality of an equity style box: Rolling distributions of actively managed
large-cap growth funds versus large-cap growth index, ten-year excess returns
100%
Percentage of funds
80
60
40
20
0
–20
–40
–60
–80
–100
Dec.
1996
Dec.
1997
Dec.
1998
Dec.
1999
Dec.
2000
Dec.
2001
Dec.
2002
Dec.
2003
Dec.
2004
Dec.
2005
Dec.
2006
Dec.
2007
Dec.
2008
Dec.
2009
Dec.
2010
Dec.
2011
Dec.
2012
Dec.
2013
Percentage outperforming a large-cap growth index
Percentage underperforming a large-cap growth index
Sources: Vanguard and Morningstar.
See Appendix for list of benchmarks used.
Figure 8 shows the constantly shifting relationship
between active managers and their target indexes over
time, with peaks and valleys that seem to coincide with
the long-term relative outperformance of one style or
another. In 1998 and 1999, when large-cap growth
stocks predominated, large-cap growth managers found
it difficult to consistently outperform their benchmark
index. (Managers who beat the benchmark are above
the x-axis; those who underperformed are below.)
However, in 2007 and 2008, when their style lagged
the others, large-cap growth managers found it easier
to outperform. As we discussed, this cyclicality is
probably driven by differences in weighting between
the managers’ portfolios and the benchmark. During
periods when a larger percentage of stocks out- or
underperform, the aggregate performance differences
between funds and the benchmark will be magnified.
As we have seen, particularly in the large-cap growth
segment from the late 1990s through 2008, active
managers’ weighting schemes have appeared to result
in substantial performance volatility in a given period.
The cumulative effect was significant underperformance
when the index was strongest and notable
outperformance when it was weakest.
11
Cyclicality in the fixed income market
Market cyclicality also has an impact on the aggregate
performance of active managers in the fixed income
market. Of all the risk factors that affect their returns
relative to their benchmarks, duration is the most
important. Indeed, in one of the most famous studies
on this topic, Litterman and Scheinkman (1991) found
that a change in interest rates is the most important
determinant of the returns of a bond or bond portfolio.2
As a result, whether to maintain a more aggressive
(longer) or less aggressive (shorter) duration than the
benchmark will tend to overwhelm an active manager’s
decisions about other factors such as sector exposure
and credit quality.
The duration of a fund (or a group of funds) therefore
allows the investor to understand the likelihood of
success relative to a benchmark index in various
interest rate scenarios. For example, if the fund has
a shorter duration than the benchmark and interest
rates fall, the benchmark will rise in value more than
the fund, all else being equal. If a majority of fund
managers expect interest rates to perform one way
(and position their funds’ durations accordingly) and
they in fact go in another direction, a majority of active
fixed income funds will likely underperform.3
Figure 9 shows how the percentage of active managers
who outperformed the Barclays U.S. Government Bond
Index over the previous 12 months has fluctuated over
time with respect to changes in intermediate-term interest
rates. We find a reasonably strong relationship, with a
correlation of 0.7. This weakens somewhat to 0.3 when
using short-term rates. The most obvious seemingly
direct relationship occurred during 2007-2008, when
the percentage of managers beating the index over the
previous 12 months approached 0% as intermediate-term
interest rates plunged. Of course, one possible explanation
is that active managers collectively had a shorter duration
than the benchmark in anticipation of interest rate
increases. Another possible explanation is that they held
significantly more lower-quality agency securities just as
these securities underperformed as a result of the creditand mortgage-market turmoil. Finally, the general flight
to quality in late 2007 and 2008 may have contributed.
We also examined how corporate fund managers
performed in relation to the credit spread and found a
positive though weaker relationship, with a correlation
of 0.4. This is likely due to the fact that coping with
credit spreads in addition to duration and yield curve
makes the relationship noisier over time.
12-month change in
10-year yields
5%
100%
4
90
3
80
2
70
1
60
0
50
–1
40
–2
30
–3
20
–4
10
–5
0
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
12-month change in yield (left)
Percentage of managers that outperformed the benchmark over previous 12-months (right)
Sources: Vanguard and Morningstar.
See Appendix for list of benchmarks used.
2 The findings of Litterman and Scheinkman were confirmed and expanded upon in a number of later studies, including but not limited to Bliss (1997), Golub and Tilman (1997),
and Matzner-Løber and Villa (2004).
3 Of course, outperformance also depends on other factors, such as whether credit spreads widened or tightened, how the slope and curvature of the yield curve may have shifted,
and the magnitude of these changes.
12
Percentage of active managers
that outperformed the benchmark
Figure 9. The relationship between interest rate changes and active manager performance
Conclusion
Appendix: Benchmarks used in this analysis
This analysis demonstrates that the volatility in the
percentage of funds outperforming a benchmark index is
directly related to underlying market trends. The number
of active managers in each style box and the differences
in the style and size characteristics of their portfolios will
explain a significant portion of out- or underperformance.
Deciding whether active management or indexing is a
“better” strategy requires an investor to focus on the
rationale for active management. Active management
offers the opportunity to outperform a given benchmark,
but at the cost of higher average expenses, potentially
significant tracking error, and the risk of underperforming.
Indexing does not offer the ability to outperform a
benchmark, but because of its relatively low expenses
and tracking error, the strategy has outperformed many
similarly positioned active managers over the long term.
Equity benchmark indexes are as follows. Broad market:
Dow Jones Wilshire 5000 Index through April 22, 2005,
MSCI US Broad Market Index through June 2, 2013,
CRSP US Total Market Index thereafter; large-cap blend:
Russell 1000 Index through November 30, 2002, MSCI US
Prime Market 750 Index through January 30, 2013, CRSP
US Large Cap Index thereafter; large-cap growth: S&P
500/Barra Growth Index through May 16, 2003, MSCI US
Prime Market Growth Index through April 16, 2013, CRSP
US Large Cap Growth Index thereafter; large-cap value:
S&P 500/Barra Value Index through May 16, 2003, MSCI
US Prime Market Value Index through April 16, 2013,
CRSP US Large Cap Value Index thereafter; mid-cap
blend: S&P MidCap 400 Index through May 16, 2003,
MSCI US Mid Cap 450 Index through January 30, 2013,
CRSP US Mid Cap Index thereafter; mid-cap growth:
Russell Midcap Growth Index through November 30,
2002, MSCI US Mid Cap Growth Index through April 16,
2013, CRSP US Mid Cap Growth Index thereafter;
mid-cap value: Russell Midcap Value Index through
November 30, 2002, MSCI US Mid Cap Value Index
through April 16, 2013, CRSP US Mid Cap Value Index
thereafter; small-cap blend: Russell 2000 Index through
May 16, 2003, MSCI US Small Cap 1750 Index through
January 30, 2013, CRSP US Small Cap Index thereafter;
small-cap growth: Russell 2000 Growth Index through
December 31, 1993, S&P SmallCap 600/Barra Growth
Index through May 16, 2003, MSCI US Small Cap Growth
Index through April 16, 2013, CRSP US Small Cap Growth
Index thereafter; small-cap value: Russell 2000 Value
Index through December 31, 1999, S&P SmallCap 600/
Barra Value Index through May 16, 2003, MSCI US Small
Cap Value Index through April 16, 2013, CRSP US Small
Cap Value Index thereafter. Government bond benchmark:
Barclays U.S. Government Bond Index.
References
Active Vs. Passive Statistical Redux. Journal of Indexes,
indexuniverse.com/index.php?option=com_content&view=
article&id=3066&Itemid=285.
Bliss, Robert R., 1997. Movements in the Term Structure of
Interest Rates. Federal Reserve Bank of Atlanta. Economic
Review, 82(4): 16–33.
Davis, Joseph H., Glenn Sheay, Yesim Tokat, and Nelson Wicas,
2007. Evaluating Small-Cap Active Funds. Valley Forge, Pa.:
The Vanguard Group.
Golub, Bennet W., and Leo M. Tilman, 1997. Measuring Yield
Curve Risk Using Principal Components Analysis, Value at Risk,
and Key Rate Duration. Journal of Portfolio Management,
Summer, 72–84.
Litterman, Robert, and Jose Scheinkman, 1991. Common Factors
Affecting Bond Returns. Journal of Fixed Income, June, 54–61.
Matzner-Løber, Eric, and Christopher Villa, 2004. Functional
Principal Component Analysis of the Yield Curve. Working Paper.
Philips, Christopher B., Francis M. Kinniry Jr., Todd Schlanger,
and Josh M. Hirt, 2014. The Case for Index Fund Investing.
Valley Forge, Pa.: The Vanguard Group.
13
Connect with Vanguard® > vanguard.com
For more information about Vanguard funds, visit vanguard.com or call 800-662-2739 to obtain
a prospectus. Investment objectives, risks, charges, expenses, and other important information
about a fund are contained in the prospectus; read and consider it carefully before investing.
CFA® is a trademark owned by CFA Institute.
The funds or securities referred to herein are not sponsored, endorsed, or promoted by MSCI, and MSCI bears no liability
with respect to any such funds or securities. For any such funds or securities, the prospectus or the Statement of
Additional Information contains a more detailed description of the limited relationship MSCI has with The Vanguard
Group and any related funds.
Vanguard Research
Russell is a trademark of The Frank Russell Company.
© 2014 The Vanguard Group, Inc.
All rights reserved.
Vanguard Marketing Corporation, Distributor.
Standard & Poor’s 500 is a trademark of The McGraw-Hill Companies, Inc., and has been licensed for use by The
Vanguard Group, Inc. Vanguard mutual funds are not sponsored, endorsed, sold, or promoted by Standard & Poor’s,
and Standard & Poor’s makes no representation regarding the advisability of investing in the funds.
P.O. Box 2600
Valley Forge, PA 19482-2600
ISGACT 072014