|
|
|
|
The Mid Cap Gap: Filling in the Missing Piece of the Asset Allocation Pie
Alfred S. Bryant, CFA: Portfolio Manager
David P. Kalis, CFA: Managing Director
Originally published: 2/14/01
From Segall Bryant & Hamill, Investment Counsel. Reprinted with Permission.
he art and science of asset allocation has continued to increase in complexity over the last seventy-five years. The original dilemma hinged on whether stocks were a worthy addition to a portfolio of bonds and other hard assets or if they were a speculative asset class appropriate only for high-risk investors. As investors became more comfortable with the risk/return characteristics of stocks, further refinements, such as large cap stocks versus small cap stocks,international versus emerging markets, and growth versus value emerged. The last ten years have led to the inclusion of additional variables into the equation,including mid cap stocks. Although mid cap stocks have been part of the u.s. market for as long as large and small cap stocks, when it comes to asset allocation, mid cap stocks have failed to achieve due recognition. Two factors that have started to reverse this neglect are the attractive investment charac-teristics of mid cap stocks relative to large and small cap stocks and the creation of mid cap stock indices by both Standard & Poor’s and The Frank Russell Company. For the purposes of this paper, mid cap stocks will be defined as companies that have market capitalizations in the range of $1.3 to $10.3 billion.
Our research concludes that mid cap stocks are an attractive investment class as judged by several different criteria:
Mid cap stocks offer attractive risk-adjusted returns relative to large cap and small cap stocks.
Investors that exclude mid cap stocks from their equity portfolios are clearly placing themselves on a less optimal risk/return spectrum.
There is a distinct mid cap gap being created by investors who rely solely on large cap and small cap managers to create mid cap exposure.
Currently, mid cap stocks trade at a discount to large cap stocks and may offer superior relative performance over the next several years.
Investing in mid cap stocks through a core strategy provides superior returns and reduces the volatility of the portfolio.
Other considerations include: mid cap companies have higher relative revenue and earnings growth rates versus large companies and more stability and liquidity compared to smaller companies.
These positive factors strongly suggest that investors, plan sponsors and consultants can more efficiently optimize their portfolios by including mid cap stocks as a stand-alone asset class.
The objective of this paper is to empirically evaluate the merits of mid cap stocks as
an asset class relative to large and small cap stocks. The evaluation utilizes two his-torical
data sets. The first is a historical pricing series from the University of Chicago’s
Center for Research in Security Prices. This data series will be referred to as “crsp,”
and was provided to Segall Bryant & Hamill by Prudential Securities. The second
series of data involving index returns was provided by The Frank Russell Company.
The issues that will be addressed are as follows: the return and risk characteristics of
mid cap stocks versus large and small cap stocks, guidelines for the allocation of mid
cap stocks to an equity portfolio, current valuations of mid cap stocks, and finally, a
comparison of core, growth and value strategies within the mid cap sector.
Table 1: Return and Risk, 1950–2001: Large, Mid, Small, and Micro Cap Stocks
Source: University of Chicago, Center for Research in Security Prices;and Ibbotson Associates, SBBI 2001.
The data used for the first portion of this study is based upon the University
of Chicago’s crsp data trend. This is a series dating back to 1926 that breaks security
returns out by decile of market capitalization.While the data series dates back to
1926, our study uses the series from 1950 through 2001. The reason for this decision
is that the volatility of returns prior to 1950 appeared to be abnormally high when
compared to the data after 1950, and that the exclusion of the earlier data would provide
a data set more consistent with current market conditions. Both sets of data led
to the same conclusions regarding the attractiveness of mid cap stocks.
Our study uses the first decile to represent the large cap proxy. A market
capitalization weighted combination of deciles two through four create the mid
cap proxy and a market capitalization weighted average of the fifth through eighth
deciles represent the small cap proxy. Deciles nine through ten are considered micro
caps, and are excluded from the study. Basic characteristics of the deciles can be observed from table 1.
Our decision to base the mid cap proxy on deciles two through four was
based upon several factors. The starting point for the process was the default cate-gories
that the University of Chicago uses with the crsp data series to define large
cap, mid cap and small cap equities. The university’s large cap classification includes
deciles one and two.The mid cap group includes deciles three through five, and the
small cap proxy equals deciles six through eight. It was our determination that decile
two, with a market cap range of $10.3 billion to $4.1 billion was clearly in the mid
cap range. As a result,this decile was merged with our mid cap grouping. Additionally,
decile five, with a market cap range of $1.3 billion to $840 million seemed to better
represent small cap equities. For this reason, this decile was merged with our small
cap proxy. The final decile in question was the fourth.This group has a market cap
range from $1.3 billion to $2.2 billion. We studied the market cap, return, risk and
correlation characteristics of this group relative to the others and determined that,in
our best judgement, this decile should be included in the mid cap proxy. These mar-ket
capitalization ranges are consistent with the current ranges for the Russell Mid
Cap Index. It should be noted that if decile four were to be merged with the small
cap group rather than the mid cap group, the insights and conclusions drawn from
this study remain the same.
Attractive Risk-Adjusted Returns
An analysis of the data for the past fifty years reveals that the large, mid and small cap
proxies produce three distinct return/risk combinations that support the basic tenets
of modern portfolio theory. This theory holds that as market capitalization decreas-es
from large to small, the risk of the representative companies increases, and there-fore,
the return must also increase. While this chart depicts the relationships for the
last fifty years, the relationships remain similar when presented on a seventy-five, forty
and thirty-year horizon.
Table 2: Return and Risk, April 1950–2001: Large, Mid, and Small Cap Stocks
Source: University of Chicago, Center for Research in Security Prices;and Ibbotson Associates, SBBI 2001.
The encouraging conclusion drawn from this data is that mid cap stocks
have provided investors with 85.7% of the excess returns that small cap stocks offer
over large cap stocks, while taking on only 37.1% of the additional risk. On a simple
return per unit of risk measurement,large cap stocks lead with 0.83, followed by mid
cap stocks at 0.82. Small cap stocks fall into last place using this measure at 0.71, due
to their higher standard deviation.
Table 3: Correlation Matrix, 1950–2001:
Large, Mid, and Small Cap Stocks
To construct and study portfolios built from the various equity asset class-es,
the correlation among the categories yields important information. As the table
at left illustrates, mid cap stocks have carried a 90.9% correlation to large cap stocks
over the last fifty years while small cap stocks have behaved in a slightly more inde-pendent
nature to large caps with an 81.0% correlation. Mid cap and small cap stocks
have a fairly high level of correlation to each other at 95.1%.
Mid cap and small cap stocks offer an attractive addition to a large cap port-folio
due to their higher return characteristics and their less than perfect correlation.
Modern portfolio theory holds that by adding a higher return asset such as mid cap
stocks to a large cap portfolio, the expected return of the portfolio increases. The standard
deviation of the portfolio will be less than the weighted average standard deviation
of the two assets,due to the lack of perfect correlation between the asset classes.
The Cost of Excluding Mid Cap Stocks
With the return, standard deviation and correlation data presented, various com-binations
of the three asset classes can be combined to study the return and risk
characteristics of the resulting portfolios. The chart below shows three separate com-binations:
the line on the right containing combinations of large cap and small cap
stocks, the center line representing combinations of large, mid and small cap stocks,
and the line on the left consisting of only large cap and mid cap stocks.These various
lines can be considered the beginnings of an efficient frontier, on which the investor
looks for the point to maximize return while minimizing risk.
Table 4: Return and Risk Efficiency: Large, Mid,and Small Cap Stock Combinations
Source: University of Chicago, Center for Research in Security Prices;and Ibbotson Associates, SBBI 2001.
There are several key points to take away from this chart. First, the least
efficient line presented here contains combinations of only large and small cap stocks.
By the least efficient line, it is meant that for a particular rate of return, the portfolio
takes on an unnecessary level of risk compared to other combinations. Notice that
every rate of return that the line on the right provides can be matched by the center
line with less volatility. Furthermore, the line on the left outperforms the center line
on a risk-adjusted return basis for all return levels less than 12.9%. From the previous
section containing return, risk and correlation, it appeared that when adding small
and mid cap stocks to an equity portfolio, the small cap stock’s higher volatility would
be offset by lower relative correlation to the large cap asset class, thereby favoring
small cap stocks. The efficient frontier graph offers evidence to the contrary, indicating
that the return/risk advantage of adding mid cap stocks to an equity portfolio
of large cap stocks is actually superior to the inclusion of small cap stocks.
Asset Allocation
The previous charts demonstrated the attractive features of mid cap stocks as an equi-ty
asset class. The chart below offers guidance regarding what portion of an equity
portfolio should be allocated to mid cap stocks. The data presented illustrates the per-cent
of total u.s. market capitalization that each asset class has contributed during the
last fifty years. During this period, large cap stocks averaged 61% of the u.s. market,
mid cap stocks accounted for 27% and small cap stocks 10%. Note that with the late
1990’s large cap rally, the large cap proxy jumped up to 75% of the total u.s. market
capitalization.
Table 5: Percent of Total US Market Cap, 1950–2001: Large, Mid, and Small Cap Stocks
Source: University of Chicago, Center for Research in Security Prices;and Ibbotson Associates, SBBI 2001.
This gain, which was unprecedented over the last fifty years, came largely
at the expense of the mid cap group. The mid cap group fell to an all-time low of 19%
in April of 2000. This trend has recently begun to revert back to the fifty-year mean.
If this reversion continues, it is certainly a bullish signal for the asset class.
During the last decade, the notion of indexing became mainstream. The
theory behind such an investment style was that active investment management,and
more specifically active security selection, would yield inferior results compared to a
strategy of simply buying the entire market.
Indexing can also be thought of in terms of asset allocation in the equity
markets. For example, one could allocate active managers within large, mid and small
cap categories in the same weightings that the market provides. Such a strategy would
make bets on active management, but remain neutral for market capitalization. One
problem with such an approach would be the yearly volatility in the annual weight-ings.
Balancing a portfolio using such unstable numbers would not be prudent.
However, using the same concept, but basing the weightings on historical averages
may prove to be a valuable strategy. Using the fifty-year averages, this would suggest
that an allocation of 61% large cap, 27% mid cap and 10% small cap would be a mar-ket
neutral bet for each of the asset classes. Using the fifty-year historical return and
standard deviation data presented earlier, such a portfolio would have yielded a com-pound
average return of 12.2% with a standard deviation of 14.4%. This is a fairly
attractive point on the efficient frontier that lies just above the large, mid and small
cap combination line on the return/risk graph shown earlier.
The Large Cap/Small Cap Solution
While the merits of investing in mid cap stocks are clear, the question remains whether
a separate allocation to mid cap stocks is necessary. The case for not including a sep-arate
mid cap manager argues that the dedicated large and small cap managers will
cover mid cap stocks with their overlapping styles. This may or may not be a good
argument depending on the tendencies of the large and small cap managers in ques-tion.
For example, if the total mid cap allocation derived from incorporating one large
and one small cap manager happens to sum up to 27% of the total portfolio (the mid
cap fifty-year average of total u.s. market capitalization), the addition of a separate
mid cap manager is unlikely to add much incremental value to the portfolio. However,
if the total mid cap exposure derived from this scenario is only 10%, exclusion of a
separate mid cap manager is effectively making a bet against the asset class, thereby
threatening the overall efficiency of the portfolio.
The ever-increasing importance of consultants in the investment man-agement
business and the notion of tracking error further bolster the argument for
separate mid cap allocations. Tracking error refers to the difference in performance
between a portfolio and its relevant benchmark. This differential may be attributa-ble
to any number of variables including security selection, sector weightings and
market capitalization. By nature, tracking error can be a positive or a negative. For
example, when a consultant selects a large cap manager for a client, the goal is that
the manager will provide positive tracking error (i.e. outperformance versus the index)
due to superior security selection of companies with market capitalizations in excess
of $10.3 billion. Conversely, the consultant may not be pleased to find that the large
cap manager outperformed the index due to a heavy concentration of mid and small
cap stocks. Despite the outperformance, a problem arises because the large cap man-ager
is not investing the client’s assets in the consultant’s intended category. This sit-uation
is commonly known in the investment management industry as style drift.
The presence of style drift will greatly diminish the value of the work that went into
the formulation of the client’s asset allocation strategy. The end loser in this scenario
is the client,who will have suffered not from a poorly developed asset allocation strat-egy,
but from poor asset allocation execution by the portfolio manager.
The important question for the typical investor who already has a sepa-rate
large and small cap manager is: what amount of their existing portfolio is already
invested in mid cap companies? By answering this question, one can determine
whether a separate mid cap manager is necessary. One way to answer this question is
to determine the percent of the total market cap within large cap and small cap indices
represented by mid cap companies. The logic follows that any large or small cap man-ager
attempting to minimize tracking error caused by market capitalization bets will
maintain mid cap allocations close to that of their relevant index.
The following example may shed some light on how much of a typical
large cap/small cap portfolio includes mid cap exposure. In our example, the Russell
2000 represents the small cap index and the Standard & Poor’s 500 will act as the large
cap manager’s bogey. As of July 2001, the Russell 2000 has 73 companies with mar-ket
capitalizations greater than $1.3 billion. Combined,these 73 companies represent
10.0% of the Russell 2000’s total market capitalization. The implication is that any
small cap manager that has allocated a significantly higher or lower percentage of the
client’s portfolio to companies with market caps greater than $1.3 billion is either
over or underweight in mid cap stocks, and thus primed for market capitalization-based
tracking error. For the large cap manager, the Standard & Poor’s 500 has 280
companies with market capitalizations less than $10.3 billion. While large in terms
of sheer numbers, these 280 companies add up to only 13.1% of the index total mar-ket
cap. Again, any large cap manager that has a greater allocation to this range of
companies will be faced with issues of market capitalization-based tracking error
against the index. The chart below illustrates what we term the mid cap gap.
Table 6: Mid Cap Gap between Russell 2000 and Standard & Poor’s 500, July 2001
Source: University of Chicago, Center for Research in Security Prices;and Ibbotson Associates, SBBI 2001.
In order to understand the significance of the mid cap gap, it is necessary
to create hypothetical portfolios consisting of only large and small cap managers with
different weightings. As can be seen from the table below, these portfolios drastical-ly
underweight mid cap stocks relative to the historical average.
Table 7: Theoretical Mid Cap Gap
The theoretical mid cap exposure is determined by the equation (AxC)+(BxD)=E.The mid cap gap is calculated by subtracting E from the mid cap fifty –year average of the total US market capitalization.
In the two-asset class world of only large and small cap managers, the max-imum
total mid cap exposure will typically amount to 12.8%, which is significantly
below the 27% that the asset class has averaged over the last fifty years. As the large
cap exposure declines in favor of small caps, the resulting mid cap exposure contin-ues
to drop. The conclusion that can be drawn from this example is that if the large
and small cap managers are attempting to minimize tracking error, a separate mid cap
manager becomes a requirement. Without a separate mid cap manager, the likely mid
cap exposure ranges from only 10.3%-12.8%, creating a mid cap gap of 14.2%-16.7%,
and thereby threatening the efficiency of the total portfolio.
Attractive Entry Point for Mid Cap Stocks
Thus far, the issues discussed have covered the attractive return/risk attributes of mid
cap stocks and the appropriate allocations to an equity portfolio. A separate, but an
equally important issue revolves around the valuation metrics of the asset class.
Table 8: P/E relative Multiple, Russell Midcap Index versus Standard & Poor’s 500 Index, 1979–2001
Source:The Frank Russell Company.
The above chart plots the Russell Midcap Index p/e multiple relative to
the Standard & Poor’s 500 p/e multiple dating back to December of 1979. The
Standard & Poor’s 500 Index is a good proxy for large cap equities in the u.s. The
chart illustrates that over the last twenty-one years, the mid cap index has held a pre-mium
relative to the large cap index. This premium status makes intuitive sense, as
mid cap stocks produce similar profitability characteristics to large cap stocks, but are
able to grow at higher rates due to their smaller size. These factors should lead to
higher multiples for the category. This graph suggests that relative to large cap stocks,
mid cap stocks appear to offer investors more attractive valuations.
Core, Growth, or Value
A second derivative question of whether or not to invest in mid cap stocks is whether
to invest in a core, growth or value strategy, or in some combination. Consultants sug-gest
investing in large, mid, and small cap stocks for the benefit of diversification.
Core, growth, and value allocations are generally found in portfolios for diversifica-tion
purposes as well. The benefit of diversification from differing styles occurs when
the allocations perform in a contrarian fashion, thus improving the risk-adjusted
returns of the portfolio.
The pertinent question relates to the optimal balance between core, growth
and value strategies within the mid cap asset class. To address this issue, the return data
over the last 15. 25 years from the Russell Midcap, Midcap Value and Midcap Growth
indices are analyzed. In this case, the Russell Midcap Index is considered to be a core
approach. Analyzing the return data for the three indices, it is clear that growth and
value go through boom and bust cycles. More specifically, when one style performs
poorly, the other tends to perform well.
By annualizing the data, it becomes evident that on an absolute and risk-adjusted
basis, neither growth nor value outperformed the core index. As theory
would suggest, growth carries the highest volatility while value produces the lowest.
The core proxy is more heavily weighted towards value with regards to volatility.
Table 9: Return and Risk, 1986–2001: Mid Cap Core, Growth, and Value Indices
Source:The Frank Russell Company.
The important question for asset allocation amongst value and growth will
relate to the correlation between the styles. The matrix below indicates that mid cap
growth and value carry diversification benefits due to their low correlation of 69.3%.
An interesting note on correlation is that the mid cap core proxy carries a 92% cor-relation
to both growth and value, suggesting that much of the diversification bene-fits
of growth and value are combined in the core approach.
Table 10: Mid Cap Correlation Matrix: Core, Growth, and Value Indices
The data suggests that an allocation to the core approach provides the best
absolute returns while producing only slightly greater volatility than the value ap-proach.
As chart 9 and table 10 illustrate, a 50%/50% blended portfolio of value and
growth managers succeeds only in lowering returns and increasing volatility (relative
to the core approach), and increasing correlation to large caps.
Conclusion
The time has come for mid cap stocks. Our study has shown that over the last fifty
years, mid cap stocks have provided investors with a return, risk and correlation to
other equity categories that improves the overall efficiency of portfolios. The asset
class has averaged twenty seven percent of the total u.s. market capitalization over
the last fifty years. This average weight provides a good starting point for determin-ing
the appropriate allocation to mid cap stocks within client portfolios. On an
absolute and risk-adjusted basis, a core approach has outperformed the value and
growth styles over the last fifteen years. Finally, despite the rally in mid cap stocks over
the last year, p/e multiples relative to large cap stocks are still at a discount, offering
upside potential.
____________________________________________________________________________________________
For more information on Segall Bryant & Hamill or questions about our
investment management services, please contact Phil Hildebrandt by phone
at 312.474.4117, via e-mail at phildebr@sbhic.com or visit our website at
www.sbhic.com.
|
|
Sponsored by James J. Eccleston. This Web site contains material of general interest. It is neither intended to, nor constitutes, either legal advice or investment advice.
Always consult an attorney and/or investment adviser when building and protecting your wealth.
All content Copyright © 2010 Advocate Compliance Partners, Inc. except where noted. All rights reserved.
One North Franklin Street, Suite 2620, Chicago, IL 60606
Telephone 312-332-0000 | Fax 312-332-0003
|
|
|
|
