Morningstar - Q1 2024 - 35

in this article based on exchange-traded funds.
Even if one does not adopt any of the
models presented here, keeping in mind how
Sharpe-optimized portfolios differ from
what's typical could help practitioners make
better asset-allocation decisions.
Optimizer Results
Our study includes seven asset classes. The
Morningstar US Market Index holds about
1,500 stocks and accounts for 97% of the U.S.
equity market's capitalization. The Morningstar
US Corporate Bond Index, Morningstar US
High-Yield Bond Index, Morningstar US MortgageBacked
Securities Index, and Morningstar
US Treasury Bond Index are all broad measures of
their respective domestic asset classes. The
Morningstar Emerging Markets Composite Bond
Index gauges fixed-rate, investment-grade,
and high-yield sovereign and corporate bonds
issued by emerging-markets entities but only those
denominated in U.S. dollars. Rebranded since
Morningstar acquired Leveraged Commentary &
Data from S&P Global in mid-2022, the Morningstar
LSTA US Leveraged Loan Index measures
U.S.-dollar-denominated bank loans that have
been syndicated in the United States, though
issuers can be located anywhere.
EXHIBIT 1 shows the Sharpe-optimized portfolio
that results from our resampled data without
requiring a set minimum equity allocation as a
starting point.2
Unlike the prototypical 60%
stock/40% bond portfolio, optimizing for excess
return relative to its standard deviation across
the seven asset classes of our study leads
to a 5% stock/95% bond portfolio. Severe equity
market downturns since 2000, including
2000-02, 2007-09, and in early 2020, dimmed
the attractiveness of a stock allocation from the
standpoint of the Sharpe ratio.
The Sharpe-optimized portfolio's bond portion
is also distinctive. EXHIBIT 1 shows how
the portfolio's weightings compare with those
of the high-yield-inclusive Morningstar US
Asset Class
Morningstar US Market Index
Morningstar US Corporate Bond Index
Morningstar US High-Yield Bond Index
Morningstar US Mortgage-Backed Securities Index
Morningstar US Treasury Bond Index
Morningstar Emerging Markets Composite Bond Index
Morningstar LSTA US Leveraged Loan Index
Core Plus Bond Index. Within the Sharpe-optimized
portfolio, the investment-grade corporate
market is not represented, even though it accounts
for one fifth of the index. The optimizer seems
to have taken its credit-risk exposure from stocks,
the modest stakes in high-yield and emergingmarkets
bonds, and most of all from low-volatility
leveraged loans, or bank loans.
Indeed, the Sharpe-optimized portfolio's
leveraged loan weighting is 30% compared with
zero for the Morningstar US Core Plus Bond
Index benchmark (which excludes this more than
$1 trillion asset class as a matter of index
construction). It also helped that bank loans are
much less correlated with Treasuries than
investment-grade corporates.
The optimizer made Treasuries, mortgages, and
bank loans the workhorses of its portfolio. That's
because for the whole period, Treasuries had
a negative correlation with bank loans and a weak
one with agency mortgages. Meanwhile, the
correlations between Treasuries and high-yield and
emerging-markets debt were low.
EXHIBIT 1
The Sharpe-Optimized Portfolio Not only is the allocation of the most optimized
portfolio 5% equity/95% bond, but also the portfolio's bond portion is distinctive.
Sharpe-Optimal
Portfolio Weighting
%
5.0
0.0
3.0
26.0
30.0
6.0
30.0
Morningstar US Core
Plus Bond Index
Weighting %
0.0
21.0
4.8
26.6
40.3
1.1
0.0
+/-
5.0
-21.0
-1.8
-0.6
-10.3
4.9
30.0
Source: Morningstar Direct; Intercontinental Exchange. Data as of Sept. 30, 2023. The Morningstar US Core Plus Bond and Morningstar Emerging Markets
Composite Bond Indexes shared in common 78 investment-grade, USD-denominated emerging markets corporate bonds and 146 quasi and foreign government
bonds, which together accounted for 1.11% of the Core Plus index and 17.55% of the Emerging Markets index, respectively. The 78 corporate bonds
also appeared in the Morningstar US Corporate Bond Index, where they had a 1.39% weighting, versus a 4.29% weighting in the Emerging Markets index.
Bank loans have risks to bear in mind. They aren't
regulated in the same way as stocks and bonds,
and trades can take longer to settle. Holding
loans through a fund mitigates those longer
settlement times, but longer settlement times can
also make the asset class vulnerable to market
shocks. The bank loan space has also evolved and
is less staid than it was during the earlier years
of our dataset.
Although the future will not repeat the past,
our use of resampled historical data to
build Sharpe-optimized portfolios assumes that
key aspects of asset-class relationships will
persist. Bank loans may not remain negatively
correlated with Treasuries, but they are
not likely to be highly correlated either. It is also
unlikely that investment-grade corporates'
positive correlation with Treasuries, mortgages,
and so on will cease. The correlations between
Treasuries and high yield on the one hand
and Treasuries and emerging-markets debt on
the other have been weak or nonexistent
for most of their respective histories, which
could continue.
2 To assess this data through September 2023, we used the statistical technique of resampling, which involved drawing random samples to account for the uncertainty and variability of financial
markets while also limiting sensitivity to any given starting or end point. In our case, from the more than 23-year period in question, we created 10,000 separate samples of 180 months
(15 years), with replacement, and have averaged the optimal allocations based on those samples. We used Python's SciPy package to set up the optimizer, which constrained all optimizations
to be long-only (thus no shorting) and to add up to 100%.
morningstar.com/products/magazine
35
http://www.morningstar.com/products/magazine
Morningstar - Q1 2024

Table of Contents for the Digital Edition of Morningstar - Q1 2024
