Alternative Thinking I 1Q19

Capital Market
Assumptions for
Major Asset Classes

Executive Summary
This article updates our estimates of year expected returns are somewhat
medium-term (5- to 10-year) expected higher for equities, U.S. Treasuries
returns for major asset classes. It and credit. However, from a historical
also includes a section on estimating perspective, nearly all long-only
expected returns for private equity investments still have low expected
and real estate. Selected estimates are real returns. The expected real return
summarized in Exhibit 1. The year of the traditional U.S. 60/40 portfolio
2018 saw cheapening across many is 2.9%, compared to a long-term
asset classes, and compared to last average of 5% (since 19001).

Exhibit 1
Medium-Term Expected Real Returns for Liquid Asset Classes
6
5.4
5.1
5-10Y Expected Real Return %

5 4.5
4.7
4.3
4.0
4
3.4
3.1 3.0
3
2.1
2 1.6
1.1
1 0.8
0.4
0
-0.2
-0.3
-1
U.S. Non-U.S. Emerging U.S. Non-U.S. U.S. IG U.S. HY Commodities
Equities Developed Market 10Y 10Y Govt Credit Credit
Equities Equities Treasuries Bonds

January 2018 January 2019

Source: AQR; see Exhibits 3-6 for details. “Non-U.S. Developed Equities” is cap-weighted average of Euro-5, Japan,
U.K., Australia, Canada. “Non-U.S. 10Y Govt. Bonds” is GDP-weighted average of Germany, Japan, U.K., Australia,
Canada. Estimates are for illustrative purposes only, are not a guarantee of performance, and are subject to change.
Not representative of any portfolio that AQR currently manages.

1  Based on historical real yields for U.S. large-cap equities and 10-year Treasuries; methodology and sources
described in Appendix.
Contents
Table of Contents 2

Introduction and Framework 1

Equity Markets 2

Government Bonds 3

Credit Indices 5

Commodities 6

Alternative Risk Premia 6

Cash 7

Extending Our Framework to Include Illiquids 8

Concluding Thoughts 12

References 13

Appendix 14

Disclosures 17

Table of Contents
 Capital Market Assumptions for Major Asset Classes | 1Q19 1

Introduction and Framework

For the past five years, the first quarter’s yields) may matter more in judging expected
Alternative Thinking has presented our capital returns. For shorter horizons, returns are
market assumptions for major asset classes, largely unpredictable, and any predictability
with a focus on medium-term expected returns has tended to mainly reflect momentum and
(see 2014, 2015, 2016, 2017 and 2018). We the macro environment.
update these estimates annually, and each
year we provide additional analysis in the form Our estimates are intended to assist investors
of new asset classes or other new material. with their strategic allocation and planning
This year, we update our estimates using decisions and, in particular, with setting
the same methodology as last year and then appropriate medium-term expectations. They
discuss similar yield-based frameworks for are highly uncertain and not intended for
estimating returns for illiquid asset classes. market timing. The frameworks for making
such estimates may be more useful and
As usual, we present local real (inflation- informative than the numbers themselves. As
adjusted) annual compound rates of return 2
one cautionary example, the blue interquartile
for a horizon of 5 to 10 years. Over such error range shown in Exhibit 2, taken from
intermediate horizons, initial market yields last year’s article, suggests that there is a 50%
and valuations tend to be the most important chance that realized equity market returns
inputs. For multi-decade forecast horizons, the over the next 10 years will under- or overshoot
impact of starting yields is diluted, so theory our estimates by more than 3% per annum.
and long-term historical average returns (or

Exhibit 2
Yield-Based Estimates Are the Best We Have but Still Highly Uncertain
10-Year Forecast Errors for Simple Expected Real Return Candidates for U.S. Equities 1900-2017

20%
Upside
surprise
Annualized 10Y Forecast Error

15%

10%

5%

0%

-5%

-10%
Downside
surprise
-15%
Equity Real Yield Historical Real Return Constant Sharpe Ratio
(out-of-sample method) (expanding window) (historical volatility)

Source: AQR. Based on quarterly overlapping 10-year periods. “Error” is realized return minus forecast return. The black diamond indicates
the mean error—if it lies on the x-axis at zero, the predictor has been unbiased on average. The shaded box shows the interquartile range of
errors (containing half the observations), and the whiskers indicate the largest upside (top) and downside (bottom) errors. See Appendix for
construction of each predictor. For illustrative purposes only. Hypothetical data has inherent limitations, some of which are disclosed herein.

2  For a discussion of expected arithmetic (or simple) vs. geometric (or logarithmic, or compound) rates of return, see the 2018 edition.
2 Capital Market Assumptions for Major Asset Classes | 1Q19

Equity Markets
For equities our starting point is the classic 2. Payout-based: Our estimate of net total
dividend discount model (DDM), under which payout yield (NTY) is the sum of current
expected real return is approximately the sum dividend yield and smoothed net buyback
of dividend yield (DY), expected trend growth yield for each country. To this we add
(g) in real dividends or earnings per share an estimate of long-term real growth
(EPS), and expected change in valuation (Δv), of aggregate payouts that includes net
that is: E(r) ≈ DY+g+Δv. We take the average issuance. This country-specific growth
of two approaches, described below. We
3
estimate, gTPagg, is an average of smoothed
assume no change in valuations, i.e., no mean historical geometric aggregate earnings
reversion from today’s (mostly high) valuations growth and forecast GDP growth. So our
toward historical averages. 4
payout-based expected return is:
E(r) ≈ NTY + gTPagg, where NTY = DY +
1. Earnings-based: We start from the inverse net buyback yield (NBY).
of the CAPE ratio (cyclically adjusted P/E),
which is the 10-year average of earnings, All estimates are higher than last year, due
inflation-adjusted to today’s price levels, to cheapening from falling equity prices in
divided by today’s price. We multiply by 0.5 2018 (see Exhibit 3). In a reverse of last year’s
(roughly the US long-run dividend payout changes, lower valuations are partly offset by
ratio) and add a real earnings growth slightly lower growth estimates. Estimates
rate of 1.5% (roughly the U.S. long-run remain low by historical standards.
geometric average). Our earnings-based
expected return 5 is therefore:
E(r) ≈ 0.5* Adjusted Shiller E/P + gEPS

3  See Alternative Thinking Q1 2017 and its online appendix for details and discussion of the methodology.
4  See Alternative Thinking Q1 2015 for a discussion of mean reversion in stock and bond valuations and our decision to exclude it.
5 For our earnings-based estimate, we apply a 50% payout ratio to all countries, and use g = 1.5% for all countries except for emerging
markets, where we use a higher growth rate of 2%. Adjusted Shiller EP is Shiller EP * 1.075 where the scalar accounts for average
earnings growth during the 10-year earnings window of the Shiller EP.
 Capital Market Assumptions for Major Asset Classes | 1Q19 3

Exhibit 3
Expected Local Real Returns for Equities, January 2019
1. Earnings-Based 2. Payout-Based Combined

Adjusted 0.5 * EP + Dividend DY+NBY + 2019 1yr

  ShillerEP gEPS gEPS Yield NBY gTPagg gTPagg Est. Change

U.S. 4.2% 1.5% 3.6% 2.2% 0.2% 2.6% 5.0% 4.3% (+0.3%)

Euro-5 6.0% 1.5% 4.5% 3.7% -0.5% 2.4% 5.7% 5.1% (+0.5%)

Japan 4.8% 1.5% 3.9% 2.6% 0.1% 2.1% 4.8% 4.3% (+0.6%)

U.K. 6.3% 1.5% 4.6% 5.0% -0.3% 2.4% 7.1% 5.9% (+0.7%)

Australia 5.6% 1.5% 4.3% 4.9% -0.9% 2.7% 6.7% 5.5% (+0.3%)

Canada 5.5% 1.5% 4.2% 3.5% -1.4% 2.6% 4.7% 4.5% (+0.5%)

Global Developed 4.7% 1.5% 3.8% 2.7% 0.0% 2.5% 5.2% 4.5% (+0.3%)

Global Dev. ex US 5.6% 1.5% 4.3% 3.7% -0.3% 2.4% 5.9% 5.1% (+0.6%)

Emerging Markets 7.9% 2.0% 5.9% 2.9% -- -- 4.9% 5.4% (+0.6%)

Source:AQR, Consensus Economics and Bloomberg. Return assumptions and methodology subject to change and based on data as of
December 31, 2018. See main text for methodology. For earnings yield, US is based on the S&P 500; UK on the FTSE 100 Index; “Euro-5”
is a cap-weighted average of large-cap indices in Germany, France, Italy, Netherlands and Spain; Japan on the Topix Index; and Emerging
Markets on the MSCI Emerging Markets Index. The period for net buyback yield (NBY) is 1988 to 2018. For payout-based estimates, all
countries are based on corresponding MSCI indices. “Global Developed” is a cap-weighted average. For emerging markets, the payout-
based estimate is dividend yield + forecast GDP per capita growth. Hypothetical performance results have certain inherent limitations,
some of which are disclosed in the back. Estimates are for illustrative purposes only, are not a guarantee of performance and are subject to
change. Not representative of any portfolio that AQR currently manages.

Government Bonds
Government bonds’ prospective medium-term rolling yields for six countries, converted to
nominal total returns are strongly anchored local real return estimates by subtracting a
by their yields. For bond portfolios with stable survey-based forecast of long-term inflation.
duration, so-called rolling yield is a better
measure (assuming an unchanged yield Since last year, our estimate for U.S.
curve). For example, a strategy of holding
6
Treasuries has increased due to higher yields,
constant-maturity 10-year Treasuries has an whereas German Bunds see a decrease
expected annual (nominal) return of 2.9%, due to lower yields and a flatter curve. The
given the starting yield of 2.7% and expected estimate for Japan remains negative. Low
capital gains of 0.2% from rolldown as the bond yields should be considered in the
bonds age. Exhibit 4 shows current local context of exceptionally low cash rates. Any

6  If we assumed a more realistic random-walk (rather than unchanged) yield curve, our estimate would theoretically need to include
convexity and variance drag components. However, since these terms are small and mostly offsetting for single bonds, we ignore
them here.
4 Capital Market Assumptions for Major Asset Classes | 1Q19

adjustment to these expected returns boils to falling/rising yields dominate returns over
down to expected future changes in the yield short horizons but are highly uncertain and
curve level or shape. Capital gains/losses due matter less over longer horizons.

Exhibit 4
Expected Local Real Returns for Government Bonds, January 2019
Y RR I Y + RR - I

10-Year Nominal Rolldown 10-Year Forecast Expected 1yr

  Bond Yield Return Inflation Real Return Change

U.S. 2.7% 0.2% 2.2% 0.8% (+0.4%)

Japan 0.0% 0.6% 1.2% -0.6% (+0.1%)

Germany 0.2% 1.1% 1.8% -0.5% (-0.5%)

U.K. 1.3% 0.8% 2.1% -0.1% (+0.2%)

Australia 2.3% 0.2% 2.4% 0.1% (-0.4%)

Canada 2.0% 0.3% 2.0% 0.3% (-0.1%)

Global Developed 1.9% 0.4% 2.0% 0.3% (+0.2%)

Global Developed ex US 0.7% 0.7% 1.7% -0.3% (-0.1%)

Source: Bloomberg, Consensus Economics and AQR. Estimates as of December 31, 2018. “Global Developed” and “Global Developed
ex US” are GDP-weighted averages of the country estimates. Rolldown return is estimated from fitted yield curves and based on annual
rebalance. Estimates are for illustrative purposes only, are not a guarantee of performance, and are subject to change. Not representative
of any portfolio that AQR currently manages.

Inflation, Currency and Cash Considerations in the Current Environment

The local real returns that we report don’t tell the whole story. To convert local real to nominal total
returns, we must add expected inflation. To convert local real to excess-of-cash returns, we must
subtract an estimate of expected real cash return. To convert local returns to those seen by a foreign
investor, we must correct for the expected risk-free rate differential plus possible cross-currency basis
(if hedged) or the expected exchange rate return from spot rate changes (if unhedged).

The significance of some of these conversions has increased in recent years. We therefore include in
the Appendix our estimates of hedged excess-of-cash returns for equities and government bonds.

For European investors, the widening differential between U.S. dollar and euro short rates makes
hedging appear increasingly expensive, although this differential is not exactly a cost. Rather, it
ensures you can only earn your own risk-free rate unless you accept currency risk (and hedged
investors will only “pay” the differential, compared to unhedged investors, if the exchange rate remains
constant—a big “if”). Expected inflation differentials are one possible basis for medium-term expected
exchange rate return estimates, and after adjusting for higher expected inflation in the U.S., the dollar
versus euro differential is somewhat narrower.
 Capital Market Assumptions for Major Asset Classes | 1Q19 5

Credit Indices
To estimate expected real returns for credit rolldown and the additional spread curve
indices, we first apply a haircut of 50% to both rolldown as bonds age and roll down the OAS
IG and HY spreads to represent the combined curve. Exhibit 5 shows our updated estimates
effects of expected default losses, downgrading for US credit indices,8 as well as hard-currency
bias and bad selling practices. We assume 7
emerging market sovereign debt, which we
no change in the spread curve, say, through include for the first time. Over the past year,
mean-reversion. We add the expected real return estimates have risen due to both wider
yield of a duration-matched Treasury. Finally, credit spreads and higher Treasury yields.
we add rolldown return — both Treasury

Exhibit 5
Expected Real Returns for US Credit Indices, January 2019
A. Spread Return B. Treasury Real Yield C. Rolldown Return

  S S * 0.5 Y I Y-I RT RC RT+RC A + B + C 

Exptd Dur- Dur-M’d Expected

Option-Adj. Excess M’tched Forecast Real Tsy OAS Total Real 1yr
  Spread Return Tsy Yld Inflation Tsy Yld Rolldown Rolldown Rolldown Return Change

US IG 1.5% 0.8% 2.6% 2.2% 0.5% 0.2% 0.3% 0.5% 1.6% (+0.5%)

US HY 5.3% 2.6% 2.6% 2.2% 0.4% 0.1% 0.0% 0.1% 3.1% (+1.0%)

EM Debt 3.9% 1.9% 2.7% 2.2% 0.5% 0.1% -- 0.1% 2.6%  (+1.0%)

Source: Barclays, Bloomberg, AQR. Estimates as of December 31, 2018. OAS and duration data is for Barclays US. Corporate Investment
Grade (IG) Index, Barclays US Corporate High Yield (HY) Index and Barclays Emerging USD Sovereign (EM Debt) Index. Index durations are
7.1 years, 4.0 years and 7.0 years respectively. Estimates are for illustrative purposes only, are not a guarantee of performance and are
subject to change. Not representative of any portfolio that AQR currently manages.

7  Consistent with Giesecke et al. (2011), who find that over the long term, the average credit risk premium is roughly half the average
spread. “Bad selling” refers to the practice of selling bonds that no longer meet the rating or maturity criteria of the index.
8  Exhibit 5 shows spreads for cash bonds in the popular Barclays indices. Actively traded synthetic indices (Markit North America CDX)
have tended to have slightly tighter spreads. The difference was small for most of 2017 and 2018 but increased in Q4 2018. For
EM debt we have insufficient data to calculate an OAS rolldown term. This year for US corporate indices, our final estimates include
corrections for Treasury convexity and variance drag — these terms are small and partly offsetting but not as closely offsetting for
indices as they are for single bonds.
6 Capital Market Assumptions for Major Asset Classes | 1Q19

Commodities
Commodities do not have obvious yield equal-dollar-weighted portfolio of commodity
measures, and we find no statistically futures, in early decades holding only five or six
significant predictability in medium-term assets but the universe growing to 13 by 1970
returns (Alternative Thinking Q1 2016). Our and 22 by 1990. This portfolio has earned about
estimate of 5- to 10-year expected return is 3% geometric average excess return over cash.
therefore simply the long-run average return. To this we add a small U.S. real cash return to
Exhibit 6 shows the performance of an give our expected real return of 3.4%.

Exhibit 6
Historical Performance of Equal-Dollar-Weighted Portfolio of Commodity Futures

1877–2018 1951–2018

Excess-of-Cash Return (AM) 4.5% 4.0%

Excess-of-Cash Return (GM) 3.0% 3.1%

Annualized Volatility 17.6% 13.4%

Sharpe Ratio (AM) 0.26 0.30

Source: AQR, Bloomberg, Chicago Board of Trade, Commodity Systems Inc. Portfolio consists of 5 to 25 of the most actively traded
commodity futures, weighted equally and rebalanced monthly, with the universe generally increasing over time as new data becomes
available. AM = arithmetic mean. GM = geometric mean. Risk-free rate used to calculate Sharpe ratio is 3M T-bill. Hypothetical data has
inherent limitations, some of which are disclosed herein. Data presented are based on hypothetical portfolios and are not representative of
any AQR product or investment.

Alternative Risk Premia

Style-Tilted Long-Only Portfolios achieve a higher expected net active
return of around 1.0% at a similar tracking
We believe a hypothetical value-tilted, error. Finally, we think a defensive or low-
diversified long-only equity portfolio that is risk equity portfolio may be assumed to
carefully implemented and reasonably priced have an expected return similar to that
may be assumed to have an expected real of the relevant cap-weighted index but
return 0.5% higher than the cap-weighted may achieve this with lower volatility.9
index, after fees. A multi-style strategy
— which we assume to include balanced Long/Short Style Premia
allocations to three highly complementary,
“tried and true” strategy styles, notably Alternative risk premia strategies apply similar
value, momentum and defensive — may tilts as long-only smart beta strategies, but in

9  Style-tilted strategies exhibit many design variations. Our estimates are purely illustrative and do not represent any AQR product or
strategy.
 Capital Market Assumptions for Major Asset Classes | 1Q19 7

a market-neutral fashion and often in multiple and shorting costs.11 Strategies that are less
asset classes. Because long/short strategies well-designed or poorly implemented may have
can be invested at any volatility level, it makes much lower expected returns.
sense to focus on expected Sharpe ratios. The
degree of diversification is essential. A single What about current style valuations?
long/short style applied in a single asset class Aggregate valuations across multiple styles
might have an expected Sharpe ratio of only remain somewhat higher than long-term
0.2 to 0.3. For a diversified composite, we averages. Some styles are rich (notably, the
believe an expected Sharpe ratio of 0.7 to 0.8, defensive style in equities) but not off-the-
net of trading costs and fees, can be feasible chart. Others are on the cheap side — equity
when multiple styles are applied in multiple value, for example, suffered losses in 2018
asset classes. At a target volatility of 10%, and is cheap by some measures. Our research
such a hypothetical portfolio would have an suggests there is only a weak link between the
expected return of 7 to 8% over cash. We 10
value spreads of style factors and their future
stress that this requires careful craftsmanship returns, making it difficult to use tactical
in portfolio construction as well as great timing based on valuations to outperform a
efficiency in controlling trading, financing strategic multi-style portfolio.12

Cash
The prospects for nominal cash returns Central Bank has finally ended QE, and the
depend on the expected path of inflation and Bank of Japan may soon follow suit, but with
of real cash rates. Long-term U.S. inflation low inflation expected to continue, their real
expectations have remained well-anchored policy rates may remain negative or near zero
just above 2%. The Federal Reserve remains in over our forecast horizon. Some tentative
policy-tightening mode, and the U.S. real cash estimates for expected real cash returns are
rate is now in positive territory. The European shown in the Appendix, Exhibit A1.

10 Consistent with historical data, we assume low correlations between the styles to produce our Sharpe ratio range for a diversified
composite of long/short styles. As transaction costs depend on implementation and both transaction costs and fees vary with target
volatility, our estimates are based on a transaction-cost-optimized strategy targeting 10% volatility with fees of 1 to 1.5%. Refer
to Alternative Thinking Q1 2015 for details of our style premia assumptions, which we believe are plausible and conservative. All
assumptions are purely illustrative and do not represent any AQR product or strategy.
11 See Israel, Jiang and Ross (2017), “Craftsmanship Alpha: An Application to Style Investing.”
12 See Asness, Chandra, Ilmanen and Israel (2017), “Contrarian Factor Timing Is Deceptively Difficult.”
8 Capital Market Assumptions for Major Asset Classes | 1Q19

Extending Our Framework

to Include Illiquids
Investors allocating to illiquid private asset quality and transparent data. Nevertheless,
classes alongside public markets may find here we extend our discounted-cashflow-based
themselves asking whether the former’s approach for equity and fixed income to the
substantially higher fees are justified by higher realms of private equity and private real estate.
expected returns. This comparison of public A common framework helps highlight how the
and private assets is not straightforward. expected returns (ER) of private assets have
Illiquid assets are inherently harder to model, similar drivers — say, yield and growth — as
and this is exacerbated by a lack of good public assets.

Private Equity

Our estimate is for U.S. buyout funds. We Then we estimate the levered return to equity
present net-of-fee expected returns, as fees by applying leverage and the cost of debt, and
are a substantial component of returns for finally we add expected multiple expansion
illiquid assets. Admittedly, each of our inputs and subtract fees to arrive at a net ER. The
is debatable as PE data limitations necessitate building blocks are as follows:
many simplifying assumptions. Still, the
broad framework explains the mechanism of • Yield: We proxy PE’s unlevered payout
how PE firms have the potential to generate yield by a quarter of its EBITDA/EV at
higher returns than public equity. PE firms deal inception.
can employ multiple levers to boost returns:
namely, higher starting yields through deal • Growth rate: We assume an unlevered real
selection; higher earnings growth rates growth rate of 3%.
through operational improvements; multiple
expansion through opportunistic timing of • Leverage: We use estimates of debt-to-
entries/exits; and financial leverage. However, equity at deal inception.
we should expect yields and growth rates for
PE to be at least loosely anchored to those for • Cost of debt: We estimate PE’s cost of debt
public equities. as real cash rate plus a third of the HY
index OAS.
Exhibit 7 illustrates our framework. First, we 13

estimate unlevered ER using the DDM: E(r) • Multiple expansion: We assume partial
≈ yU + g U, where yU = unlevered payout yield richening of PE multiple toward public
and g U = real earnings-per-share growth rate. market multiple.

13 See Ilmanen, Chandra and McQuinn (2019a) for a detailed discussion of this framework, our input choices, and the sources. Strictly
speaking, the framework applies to the current vintage rather than the entire PE market. This paper also discusses the theoretical
rationales and historical average returns to assess expected PE returns.
 Capital Market Assumptions for Major Asset Classes | 1Q19 9

• Fees: We assume all-in PE fees of 5.7%, Our current estimate of PE expected return
as per average experience of large and expected outperformance over public
institutions. 14
equity are clearly low compared to historical
averages. A downtrend in expected returns
Putting it all together, our yield-based estimate from the 1990s to the 2010s was driven by
is about 4% net-of-fee real return, half the richening PE multiples (resulting in both lower
long-run average. This compares to 3.5% for our yields and lower multiple expansion) and a
earnings-based estimate for U.S. public equity, gradual decline in PE leverage.
net of a 0.1% fee for passive investing. We thus
find PE to have a roughly 0.5% higher net-of-fee On the next page we outline an alternative
ER. We interpret this expected outperformance approach for generating an expected return
as a warranted risk premium given PE’s higher for PE, which gives a higher estimate of
equity risk, assuming that any warranted 5.6%. Taking a simple average of the two
illiquidity premium is offset by investors paying approaches gives a final estimate of 4.7%,
for the smoothness in private asset returns. compared to our combined U.S. large cap
equity estimate of 4.3%.

Exhibit 7
Building Blocks for U.S. Private Equity Expected Real Returns
Financial
Unlevered Leverage Levered

r| = ru +
ru = yu (D/E) * rg = r|
yu gu + gu D/E kd (ru - kd) m +m
Real Multiple Net Exp.
Earnings Growth Real Debt to Real Cost Real Expansion Gross Real
Yield Rate Return Equity of Debt Return (Ann.) Real ER Fees Return

Current 2.1% + 3.0% = 5.1% 109% 1.2% 9.3% + 0.3% = 9.6% - 5.7% = 3.9%

Historical
Average
3.1% + 3.0% = 6.1% 181% 2.3% 12.9% + 1.0% = 13.8% - 5.7% = 8.1%
(1993–
2018)

Source: AQR, Pitchbook, Bain & Company, Bloomberg, CEM Benchmarking, Consensus Economics. Current estimate as of September 30,
2018, and subject to change. Historical averages cover period January 1, 1993 to September 30, 2018. Strictly speaking, our inputs are log
returns and should be converted to simple returns before leverage is applied, then converted back to log returns. This “round-trip” has only a
small impact, so we omit it here. Please see the Appendix for further detail. For illustrative purposes only and not representative of any portfolio
or strategy that AQR currently manages. There is no guarantee, express or implied, that long-term return targets will be achieved. Realized
returns may come in higher or lower than expected.

14 Assumptions as of September 30, 2018, and subject to change. EBITDA/EV (earnings before interest, tax, depreciation and
amortization/enterprise value) data from Pitchbook. Our yield proxy reflects limited available PE data and historical experience for
public equities, where dividend payout has averaged roughly half of earnings, and unlevered earnings (EBIT) has averaged roughly half
of EBITDA. The 3% growth rate is more than double what we assume for public equities (due to potential operational improvements
and sector composition), and it is further amplified through financial leverage. Leverage estimates are from Pitchbook. The cost of debt
includes a large haircut to the HY spread, as PE debt is mainly secured bank loans financed at a lower cost of debt, and further, the
entire credit spread may overstate the cost of debt actually borne by firms on average. The fee estimate is from a CEM Benchmarking
analysis quoted in Doskeland-Stromberg (2018) and McKinsey (2017) and includes 167 funds. For more details on assumptions, see
Ilmanen, Chandra and McQuinn (2019a). All assumptions are purely illustrative and do not represent any AQR product or strategy.
10 Capital Market Assumptions for Major Asset Classes | 1Q19

A Simpler Approach: Start from a Public Equity

Estimate and Adjust for Size and Leverage
Rather than separately modeling all the different components listed above, we could simply assert
that on average, at the industry level, any PE alpha is retained by managers in the form of fees. Under
this assumption, PE end-investors receive, on average, levered small-cap exposure with some useful
return-smoothing but zero net alpha (i.e., the same net Sharpe ratio as public equity). Here is an
example of how this method could be employed (see Appendix for details):15

Step 1:
Start from expected real return of large cap public equity; subtract real cash return and add variance
drag to give excess-of-cash return; use volatility estimate from historical data to calculate expected
Sharpe ratio:
Expected Real Excess-of-Cash Expected Expected
Return (GM) Exp. Return (AM) Volatility Sharpe Ratio

U.S. LC Public Equity 4.3% 5.0% 14.6% 0.34

Step 2:
Assume same Sharpe ratio for small caps; estimate small cap return based on higher volatility:
Expected Excess-of-Cash
Sharpe Ratio Expected Volatility Exp. Return (AM)

U.S. SC Public Equity 0.34 19.2% 6.5%

Step 3:
Estimate private equity expected net-of-fee excess-of-cash return by applying a leverage adjustment,
then add real cash return and subtract variance drag to get estimated PE real return:
PE Expected Premium Over
SC Public Private Real Return LC Public
Equity Leverage Equity (GM) Equity

Excess-of-Cash Return 6.5% x 1.2 = 7.8% 5.6% 1.3%

Volatility 19.2% x 1.2 = 23.0%

Sharpe Ratio 0.34 0.34

The advantages of this top-down public proxy approach are its simplicity (fewer design choices
to get right but still economically intuitive and consistent with the empirical evidence) and higher
quality input data.

The disadvantages are that it may be too simple: it tells us nothing about any time-varying
premium of private over public equities, for example, due to changing relative valuations,
and it gives fewer insights into the fundamental drivers of PE’s expected return.

15 Source: AQR, Bloomberg. Assumptions of December 31, 2018, and subject to change. Hypothetical data has inherent
limitations, some of which are disclosed herein. For illustrative purposes only and not representative of any AQR product or
strategy. Please see Appendix for further detail on proxies and methodology.
 Capital Market Assumptions for Major Asset Classes | 1Q19 11

Private Real Estate

We estimate expected returns for unlevered market cap), since one-third tends to be
U.S. direct real estate (RE) as represented by offset by capital expenditure. As of Q3
the NCREIF indices. We caveat that returns 2018, the NCREIF NOI yield was roughly
for individual RE funds can vary vastly from 4.4%, leading to an estimated cashflow
the industry average, due to a wide dispersion yield of 2.9%.
of returns by geographic region, sector, and
• Growth rate: We expect that on average,
manager (this is also true of PE).
the long-term growth rate in real estate
cashflows should equal inflation, i.e., the
Real estate returns can be characterized as the
real growth rate in earnings is zero.
sum of a relatively steady income component
and more volatile price appreciation. Our • Multiple expansion: As with public
yield-based approach for RE is similar to our equities and bonds, we assume no change
DDM-based approach for equities, where we in multiples.
sum payout yield and expected long-term
growth rates.16 Our assumptions for RE’s Putting this together in Exhibit 8 gives us a
building blocks are as follows: 17
gross expected real return of roughly 3% for
unlevered RE (to make it comparable to the
• Yield: We follow Pagliari (2017) in unlevered returns reported by NCREIF). This
approximating RE cashflow yield by two- is similar to our real return estimate of 2.9%
thirds of net operating income yield (NOI/ for a 60/40 portfolio.

Exhibit 8
Real Expected Returns for Private Real Estate
NOI C ≈ NOI / 3 CF ≈ NOI - C g ER = CF + g

NOI Capital Cashflow Real Unlevered

  Yield Expenditure Yield Growth Real Return

U.S. Real Estate 4.4% 1.5% 2.9% 0.0% 2.9%

Source: AQR, NCREIF Webinar Q3 2018. Estimates as of September 30, 2018, and subject to change. Hypothetical data has inherent
limitations, some of which are disclosed herein. For illustrative purposes only and not representative of any AQR product or strategy.

16 As with private equity, an alternative approach would be to use a public proxy (i.e., REITs) and adjust for leverage as required. We omit
the details here for the sake of brevity.
17 See Ilmanen, Chandra and McQuinn (2019b) and Pagliari (2017) for details. Our growth assumption is a compromise given mixed
evidence of mildly positive or mildly negative long-run real growth.
12 Capital Market Assumptions for Major Asset Classes | 1Q19

Concluding Thoughts
Yield-based expected returns for equities framework helps to illustrate the “moving
and bonds may be our best estimates of parts” underlying expected returns for private
medium-term prospects for these asset classes. equity and real estate. This framework is a
As of January 2019, these estimates are work in progress, but we believe it is a step
mostly higher than they were a year ago, but toward a more intuitive and transparent
compared to historical norms, they remain comparison between public and private assets.
soberingly low. They continue to suggest that
over the next decade, many investors may We again emphasize that our return estimates
struggle to meet return objectives anchored to for all asset classes are highly uncertain. The
a rosier past. estimates in this report do not in themselves
warrant aggressive tactical allocation
Growing investor interest in illiquid assets responses — but they may warrant other
encouraged us to add them to our report kinds of responses. For example, investment
this year. We believe some investors may objectives may still need to be reassessed,
have inflated return expectations for such even if this necessitates higher contribution
assets. This may be due to inherent modeling rates and lower expected payouts. And the
difficulties, as well as lack of transparency case for diversifying away from traditional
on fees and performance. While some of equity and term premia remains strong despite
our assumptions are debatable, we hope our recent modest rises in expected returns.

And Finally… Historical Context for 2018,

the Year with Nowhere to Hide

The year 2018 was exceptional not in the liabilities). The year 2018 also stands out
depth of market losses, which were nothing because of what preceded it—in 2017 all 15
like the losses of a decade earlier, but investments outperformed cash, and in 2016
in the breadth of pain for investors. We it was all but one.
illustrate this in Exhibit 9 below, which
shows the annual performance of 15 major One simple take-way from this observation:
investments compared to cash. Not since investors whose portfolios suffered broad pain
1981 (not shown), when cash rates were in the in the past year—and who may be questioning
double digits, have we seen 2018’s breadth their strategic asset allocation—should not
of underperformance. Equity losses were overreact. Any one-year period can exhibit
much worse in 2008, but that year at least extreme outcomes, and 2018, the year when
Treasuries were up (though this was a double- diversification failed for almost everyone, was
whammy for pension funds with unhedged a prime example of this.
 Capital Market Assumptions for Major Asset Classes | 1Q19 13

Exhibit 9
Annual Performance of Selected Investments 2000–2018

Source: AQR, Bloomberg. See Appendix for proxies. Past performance is not a guarantee of future performance. Please read important
disclosures in the Appendix.

References
Asness, C., S. Chandra, A. Ilmanen and R. Israel, 2017, “Contrarian Factor Timing Is
Deceptively Difficult,” Journal of Portfolio Management Special Issue.
Døskeland, T. M. and Strömberg, P., 2018, “Evaluating Investments in Unlisted Equity for the
Norwegian Government Pension Fund Global,” Norwegian Ministry of Finance report.
Giesecke, K., F. Longstaff, S. Schaefer and I. Strebulaev, 2011, “Corporate Bond Default Risk: A
150-Year Perspective,” Journal of Financial Economics, 102, 233-250.
Ilmanen, A., S. Chandra and N. McQuinn, 2019a, “Demystifying Illiquid Assets: Expected
Returns for Private Equity,” AQR white paper.
Ilmanen, A., S. Chandra and N. McQuinn, 2019b, “Demystifying Illiquid Assets: Expected
Returns for Real Estate,” AQR white paper.
Israel, R., S. Jiang and A. Ross, 2017, “Craftsmanship Alpha: An Application to Style Investing,”
Journal of Portfolio Management Multi-Asset Strategies Special Issue.
McKinsey, 2017, “Equity investments in unlisted companies,” Norwegian Ministry of Finance
report.
Pagliari, J. L., 2017, “Some Thoughts on Real Estate Pricing,” Journal of Portfolio Management 34
(6), 44–61.
14 Capital Market Assumptions for Major Asset Classes | 1Q19

Appendix
Translating Local Real Returns to Hedged Excess-of Cash Returns

In Exhibit A1 we translate our local real return estimates to nominal excess-of-cash returns by
subtracting intermediate-horizon estimates of real cash return for each market. These tentative
real cash return estimates are a weighted combination of current nominal cash rates and 10-year
bond yields, minus expected inflation.

These excess-of-cash returns are the expected excess returns accessed by hedged investors
irrespective of their base currency. While real returns are often relevant for assessing
expectations versus investment objectives, excess-of-cash returns are the appropriate unit for
making international allocation decisions. When viewed in excess of local cash, non-U.S. assets
look relatively more attractive, and U.S. assets relatively less attractive.

Exhibit A1
Translating Local Real Returns to Hedged Excess-of Cash Returns
Hedged Hedged
Local Excess- 10-Year Local Excess-
Local Real Real Cash of-Cash Government Local Real Real Cash of-Cash
Equities Return Return Return Bonds Return Return Return

U.S. 4.3% 0.4% 3.9% U.S. 0.8% 0.4% 0.4%

Euro-5 5.1% -1.5% 6.6% Japan  -0.6%  -1.2%  0.6%

Japan 4.3% -1.2% 5.6% Germany -0.5% -1.5% 1.0%

U.K. 5.9% -1.0% 6.9% U.K.  -0.1%  -1.0%  0.9%

Australia 5.5% -0.2% 5.7% Australia 0.1% -0.2% 0.4%

Canada 4.5% -0.1% 4.5% Canada  0.3%  -0.1%  0.4%

Global Developed 4.5% -0.2% 4.7% Global Developed 0.3% -0.2% 0.5%

Developed ex US 5.1% -1.1% 6.2% Developed ex US  -0.3%  -1.0%  0.7%

Emerging Mkts 5.4% 1.4% 4.0%

Source: AQR, Consensus Economics and Bloomberg. Return assumptions and methodology subject to change and based on data as of
December 31, 2018. See main text and Exhibits 3 and 4 for methodology. Global Developed and Developed ex US are cap-weighted for
equities but GDP-weighted for bonds. Hypothetical performance results have certain inherent limitations, some of which are disclosed in
the back. Estimates are for illustrative purposes only, are not a guarantee of performance and are subject to change. Not representative of
any portfolio that AQR currently manages.
 Capital Market Assumptions for Major Asset Classes | 1Q19 15

Sources for Long-Term Historical Expected Returns

Sources for historical equity and bond expected returns are AQR, Robert Shiller’s data library,
Kozicki-Tinsley (2006), Federal Reserve Bank of Philadelphia, Blue Chip Economic Indicators,
Consensus Economics and Morningstar. Prior to 1926, stocks are represented by a reconstruction
of the S&P 500 available on Robert Shiller’s website, which uses dividends and earnings data
from Cowles and associates, interpolated from annual data. After that, stocks are the S&P 500.
Bonds are represented by long-dated Treasuries. The equity yield is a 50/50 mix of two measures:
50% Shiller E/P * 1.075 and 50% Dividend/Price + 1.5%. Scalars are used to account for long-term
real Earnings Per Share (EPS) Growth. Bond yield is 10-year real Treasury yield minus 10-year
inflation forecast as in Expected Returns (Ilmanen, 2011), with no rolldown added.

Methodology for Forecast Error Analysis (Exhibit 2)

We first produce a long historical time series for our yield-based estimate for U.S. equities (the
analysis starts in 1900, but we use data from the 1870s onward). We then calculate two intuitive
non-yield-based candidates: one is simply the historical average real return up to the date of
the forecast while the other assumes a constant Sharpe ratio of 0.2 and converts this into a real
return estimate using historical volatility and real cash return. These candidates are described
below. We test their predictive power using quarterly overlapping 10-year periods since 1900. See
Alternative Thinking Q1 2018 for more details.

Candidate Construction Hindsighted?

Equity Real Yield Only in use of DDM model (~1956) &

Avg(adj. Shiller EP * o-o-s payout, DP) + o-o-s G
(out-of-sample) CAPE (1988)

Historical Real Return Annualized geom. real return

No
(1881-) (1881 - forecast date)

Constant SR=0.2, (0.2 * annualized vol) + real cash ret.

Yes: SR = 0.2
Historical Vol (1881 - forecast date)

Source: AQR, Shiller data library, Blue Chip Economic Indicators, Consensus Economics and the Federal Reserve Bank of Philadelphia. All
inputs based on historical data use expanding windows starting in 1871. Out of sample (“o-o-s”) estimates are calculated from available data,
using an expanding window. EP and DP are earnings-to-price and dividend-to-price ratios respectively. SR is Sharpe ratio. For illustrative
purposes only.
16 Capital Market Assumptions for Major Asset Classes | 1Q19

Assumptions for Simple Proxy-Based PE Estimate

We start from our 4.3% expected real return for US large-cap public equity as stated in Exhibit
3. We add 1.1% variance drag, then subtract 0.4% real cash return (see Exhibit A1) to give an
arithmetic excess-of-cash return of 5.0%. Our volatility estimate of 14.6% is based on the S&P
500 since 1990 (monthly data), and implies an expected Sharpe ratio of 0.34. Our small-cap
volatility estimate of 19.2% is based on the Russell 2000 since 1990; with the same Sharpe ratio,
this implies an excess return of 6.5%. We multiply this by 1.2 to account for PE’s higher leverage,
then add the real cash return and subtract variance drain of 2.7%, based on estimated volatility
of 23% (small-cap volatility x 1.2). This gives our final proxy-based PE estimate of 5.6% real
return, as of December 31, 2018.

Proxy Indices for Annual Returns Chart (Exhibit 9)

Asset Class Proxy Index

U.S. Large-Cap Equities Russell 1000

U.S. Small-Cap Equities Russell 2000

European Equities MSCI Europe USD Unhedged

Japanese Equities MSCI Japan USD Unhedged

Emerging Markets Equities MSCI Emerging Markets USD Unhedged

U.S. 10Y Treasuries U.S. 10-Year Treasuries

U.S. IG Credit Excess Barclays U.S. Corporate IG (excess of Treasuries)

U.S. HY Credit Excess Barclays U.S. HY (excess of Treasuries)

Global Aggregate Bonds Barclays Global Aggregate USD Hedged

EM Debt JP Morgan EMBI

Inflation-linked Bonds Barclays Global Inflation-Linked USD Hedged

Commodities S&P GSCI

Gold S&P GSCI Gold Index

Real Estate FTSE NAREIT All REITs

Hedge Funds Credit Suisse Hedge Fund Index

Source: AQR, Bloomberg.

 Capital Market Assumptions for Major Asset Classes | 1Q19 17

Disclosures
This document has been provided to you solely for information purposes and does not constitute an offer or solicitation of an offer or any advice or
recommendation to purchase any securities or other financial instruments and may not be construed as such. The factual information set forth herein
has been obtained or derived from sources believed by the author and AQR Capital Management, LLC (“AQR”) to be reliable, but it is not necessarily all-
inclusive and is not guaranteed as to its accuracy and is not to be regarded as a representation or warranty, express or implied, as to the information’s
accuracy or completeness, nor should the attached information serve as the basis of any investment decision. This document is intended exclusively
for the use of the person to whom it has been delivered by AQR, and it is not to be reproduced or redistributed to any other person. The information set
forth herein has been provided to you as secondary information and should not be the primary source for any investment or allocation decision.
Past performance is not a guarantee of future performance.
This presentation is not research and should not be treated as research. This presentation does not represent valuation judgments with respect to
any financial instrument, issuer, security or sector that may be described or referenced herein and does not represent a formal or official view of AQR.
The views expressed reflect the current views as of the date hereof, and neither the author nor AQR undertakes to advise you of any changes in the
views expressed herein. It should not be assumed that the author or AQR will make investment recommendations in the future that are consistent
with the views expressed herein or use any or all of the techniques or methods of analysis described herein in managing client accounts. AQR and its
affiliates may have positions (long or short) or engage in securities transactions that are not consistent with the information and views expressed in
this presentation.
This analysis is for illustrative purposes only. This material is intended for informational purposes only and should not be construed as legal or tax
advice, nor is it intended to replace the advice of a qualified attorney or tax adviser. The recipient should conduct his or her own analysis and consult
with professional advisers prior to making any investment decisions.
The information contained herein is only as current as of the date indicated and may be superseded by subsequent market events or for other reasons.
Charts and graphs provided herein are for illustrative purposes only. The information in this presentation has been developed internally and/or obtained
from sources believed to be reliable; however, neither AQR nor the author guarantees the accuracy, adequacy or completeness of such information.
Nothing contained herein constitutes investment, legal, tax or other advice nor is it to be relied on in making an investment or other decision.
There can be no assurance that an investment strategy will be successful. Historic market trends are not reliable indicators of actual future market
behavior or future performance of any particular investment, which may differ materially, and should not be relied upon as such. Target allocations
contained herein are subject to change. There is no assurance that the target allocations will be achieved, and actual allocations may be significantly
different from that shown here. This presentation should not be viewed as a current or past recommendation or a solicitation of an offer to buy or sell
any securities or to adopt any investment strategy.
The information in this presentation may contain projections or other forward-looking statements regarding future events, targets, forecasts or
expectations regarding the strategies described herein and is only current as of the date indicated. There is no assurance that such events or targets
will be achieved and may be significantly different from that shown here. The information in this presentation, including statements concerning financial
market trends, is based on current market conditions, which will fluctuate and may be superseded by subsequent market events or for other reasons.
Performance of all cited indices is calculated on a total return basis with dividends reinvested.
Diversification does not eliminate the risk of experiencing investment losses. Broad-based securities indices are unmanaged and are not subject to
fees and expenses typically associated with managed accounts or investment funds. Investments cannot be made directly in an index.
The investment strategy and themes discussed herein may be unsuitable for investors depending on their specific investment objectives and financial
situation. Please note that changes in the rate of exchange of a currency may affect the value, price or income of an investment adversely.
Neither AQR nor the author assumes any duty to, nor undertakes to update forward-looking statements. No representation or warranty, express or
implied, is made or given by or on behalf of AQR, the author or any other person as to the accuracy and completeness or fairness of the information
contained in this presentation, and no responsibility or liability is accepted for any such information. By accepting this presentation in its entirety, the
recipient acknowledges its understanding and acceptance of the foregoing statement.
The data and analysis contained herein are based on theoretical and model portfolios and are not representative of the performance of funds or
portfolios that AQR currently manages. Volatility-targeted investing described herein will not always be successful at controlling a portfolio’s risk or
limiting portfolio losses. This process may be subject to revision over time.
HYPOTHETICAL PERFORMANCE RESULTS HAVE MANY INHERENT LIMITATIONS, SOME OF WHICH, BUT NOT ALL, ARE DESCRIBED HEREIN.
NO REPRESENTATION IS BEING MADE THAT ANY FUND OR ACCOUNT WILL OR IS LIKELY TO ACHIEVE PROFITS OR LOSSES SIMILAR TO THOSE
SHOWN HEREIN. IN FACT, THERE ARE FREQUENTLY SHARP DIFFERENCES BETWEEN HYPOTHETICAL PERFORMANCE RESULTS AND THE
ACTUAL RESULTS SUBSEQUENTLY REALIZED BY ANY PARTICULAR TRADING PROGRAM. ONE OF THE LIMITATIONS OF HYPOTHETICAL
PERFORMANCE RESULTS IS THAT THEY ARE GENERALLY PREPARED WITH THE BENEFIT OF HINDSIGHT. IN ADDITION, HYPOTHETICAL
TRADING DOES NOT INVOLVE FINANCIAL RISK, AND NO HYPOTHETICAL TRADING RECORD CAN COMPLETELY ACCOUNT FOR THE IMPACT
OF FINANCIAL RISK IN ACTUAL TRADING. FOR EXAMPLE, THE ABILITY TO WITHSTAND LOSSES OR TO ADHERE TO A PARTICULAR TRADING
PROGRAM IN SPITE OF TRADING LOSSES ARE MATERIAL POINTS THAT CAN ADVERSELY AFFECT ACTUAL TRADING RESULTS. THERE
ARE NUMEROUS OTHER FACTORS RELATED TO THE MARKETS IN GENERAL OR TO THE IMPLEMENTATION OF ANY SPECIFIC TRADING
PROGRAM, WHICH CANNOT BE FULLY ACCOUNTED FOR IN THE PREPARATION OF HYPOTHETICAL PERFORMANCE RESULTS, ALL OF
WHICH CAN ADVERSELY AFFECT ACTUAL TRADING RESULTS.
The hypothetical performance results contained herein represent the application of the quantitative models as currently in effect on the date first
written above, and there can be no assurance that the models will remain the same in the future or that an application of the current models in the future
will produce similar results because the relevant market and economic conditions that prevailed during the hypothetical performance period will not
necessarily recur. Discounting factors may be applied to reduce suspected anomalies. This backtest’s return, for this period, may vary depending on
the date it is run. Hypothetical performance results are presented for illustrative purposes only. In addition, our transaction cost assumptions utilized in
backtests, where noted, are based on AQR Capital Management LLC’s, (“AQR’s”) historical realized transaction costs and market data. Certain of the
assumptions have been made for modeling purposes and are unlikely to be realized. No representation or warranty is made as to the reasonableness
of the assumptions made or that all assumptions used in achieving the returns have been stated or fully considered. Changes in the assumptions may
have a material impact on the hypothetical returns presented. Actual advisory fees for products offering this strategy may vary.
Gross performance results do not reflect the deduction of investment advisory fees, which would reduce an investor’s actual return. For example,
assume that $1 million is invested in an account with the Firm, and this account achieves a 10 percent compounded annualized return, gross of fees, for
five years. At the end of five years that account would grow to $1,610,510 before the deduction of management fees. Assuming management fees of
1 percent per year are deducted monthly from the account, the value of the account at the end of five years would be $1,532,886, and the annualized
rate of return would be 8.92 percent. For a 10-year period, the ending dollar values before and after fees would be $2,593,742 and $2,349,739,
respectively. AQR’s asset-based fees may go up to 2.85 percent of assets under management and are generally billed monthly or quarterly at the
commencement of the calendar month or quarter during which AQR will perform the services to which the fees relate. Where applicable, performance
fees are generally equal to 20 percent of net realized and unrealized profits each year, after restoration of any losses carried forward from prior years.
In addition, AQR funds incur expenses (including start-up, legal, accounting, audit, administrative and regulatory expenses) and may have redemption
or withdrawal charges up to 2 percent based on gross redemption or withdrawal proceeds. Please refer to AQR’s ADV Part 2A for more information
on fees. Consultants supplied with gross results are to use this data in accordance with SEC, CFTC, NFA, or the applicable jurisdiction’s guidelines.
There is a risk of substantial loss associated with trading commodities, futures, options, derivatives and other financial instruments. Before trading,
investors should carefully consider their financial position and risk tolerance to determine whether the proposed trading style is appropriate. Investors
should realize that when trading futures, commodities, options, derivatives and other financial instruments, one could lose the full balance of their
account. It is also possible to lose more than the initial deposit when trading derivatives or using leverage. All funds committed to such a trading
strategy should be purely risk capital.
 www.aqr.com

AQR Capital Management, LLC Two Greenwich Plaza, Greenwich, CT 06830 P +1.203.742.3600 F +1.203.742.3100