For marketing purposes

For professional / qualified /

institutional clients and investors only

Finding opportunities
as growth slows
Five-year capital market expectations

By UBS Asset Management Investment Solutions

Slower growth and market performance
Our five-year (2019-2024) expectations in brief

UBS Asset Management’s Investment Solutions team conducts ongoing macroeconomic

research to develop our Baseline five- and 10-year return expectations. Drawing on the breadth
and depth of expertise provided by more than 100 professionals and an over-36-year asset
management track record, our capital market expectations quantify risk/return expectations for
a broad set of asset classes, incorporate current market and economic conditions and provide a
key component for modeling a portfolio’s strategic asset allocation.

– As global growth slows and interest rates remain low across the world, our five-year returns for the capital markets have been
revised downward slightly.
– Over the next five years, we expect average annual global growth of 2.8% and average global inflation of 2.3%.
– Real growth on average should be slightly lower than in the previous decade.

Equities Alternatives
– US equities are overvalued and we expect mid-single digit – For the next few years, we have lower expectations for
returns. unlevered real estate.
– Most non-US markets are fairly valued or undervalued. We – For US private equity we expect returns in the high single
expect high single digit returns. digits, but not on par with some of the touted returns of
the past. We expect unhedged global private equity to
Fixed income outperform the US.
– With negative nominal yields still predominant in many
developed markets, we expect very low fixed income Special topic: Chinese financial markets
returns. For many markets, investors should prepare for – As China continues its rapid transition from an emerging to
negative nominal returns. a developed market, we believe that Chinese equities and
fixed income will continue to offer investors unique
opportunities.

2
 Equilibrium expectations
(an average over the next 30+ years)
Reflect our views for a broad
set of asset classes in a
long-term or Equilibrium setting

Baseline expectations
(the next 5–10 years)
Take Equilibrium views and
incorporate current market
conditions with an intermediate
time horizon
Tactical views
(the next 3–18 months)
Set tactical asset allocation
views in portfolios. Used to
determine overweights and
underweights around SAA

The UBS approach to capital market assumptions

Capital market expectations (CMEs) are critical inputs in valuations, market conditions and key forward-looking inputs
designing an investment strategy that will help investors meet to generate five- and 10-year expected returns by asset class
specific objectives. A pension plan, for example, has liabilities and region. It is important to note that Baseline forecasts
with certain wage, payout and inflation assumptions; an should be taken as measures of tendency rather than point
endowment may plan for distributions based on university estimates, and can be subject to considerable swings in the
budget growth; or a family office may have income and real five-year period as the economic and market cycle progresses.
growth objectives. Ultimately, the CMEs must have an Our Baseline expectations are used to design and evaluate
economic logic and consistency behind them that tie into the existing strategic asset allocations.
larger setting that investors face.
Valuation is an important component of our Baseline process.
We produce long-term, or ‘Equilibrium,’ assumptions (average Valuation focuses on the deviation of asset prices from their
return for an asset class over the next 30-40 years, for the intrinsic value; well established as a key driver of returns over
purpose of setting general benchmarks), and Baseline multi-year periods.1 When assets are undervalued they tend to
expectations covering five- and 10-year returns, for the be set for a period of strong performance, and when they are
purpose of building strategic asset allocations. We update overvalued they typically lag, though timing may vary. For
Equilibrium and Baseline assumptions quarterly. We rely on example, investing in US large cap growth equities in the late
shorter term catalysts – such as earnings revisions, manager 1990s or long-dated bonds in the 1970s would have been
positioning, market stress and momentum – to set tactical ill-advised. Forward-looking inputs may vary by asset class but
asset allocation views in portfolios. generally include inflation, economic and earnings growth,
cash yields, 10-year government bond yields and default and
We call our long run assumptions Equilibrium because we recovery rates for investment grade and high yield bonds. We
assume that economic and financial variables ultimately offer direct estimates of value on 11 equity markets, 14 fixed
regress to an internally consistent regime around long-term income markets and 33 currency markets. We can use these
equilibriums. Our Baseline process incorporates current to determine some broad measures of global expectations.

1
Intrinsic value deviation as a driver of returns has been broadly validated by several academic studies, most notably in John Y. Campbell and Robert J.
Shiller (2001) “Valuation Ratios and the Long-Run Stock Market Outlook: An Update.” NBER Working Papers: 8221), as well as our team’s experience
spanning over three decades.

3
Equities

4
Exhibit 1 – Global equities 5-year expectations

5-Year 10-Year 5-Year 10-Year

Equilibrium Expected Expected Local Expected Expected Unhedged
Geometric Return Return Standard Return in Return in Standard
Return % Local % Local % Deviation % USD Unh % USD Unh % Deviation %

US Large Cap 7.4 5.1 5.5 16.1 5.1 5.5 16.1

Australia 6.5 8.5 8.7 18.3 8.7 8.8 22.1

Canada 6.6 6.8 6.9 18.3 7.6 7.5 22.1

Eurozone 7.0 8.3 8.6 18.9 9.1 9.8 21.4

Japan 6.9 5.7 5.6 18.9 8.3 8.1 21.4

Switzerland 5.8 7.3 7.6 17.3 7.3 8.4 19.8

United Kingdom 6.9 8.6 9.8 18.3 9.5 10.5 20.9

Developed Mkts 7.5 6.1 6.5 14.9 6.6 7.0 15.4

Developed x US 7.3 7.5 7.9 15.1 8.7 9.1 16.7

Emerging Mkts 8.3 11.0 11.6 21.9 11.2 11.3 26.0

Global Equities 7.7 6.8 7.2 15.3 7.2 7.6 16.0

Global x US 7.7 8.5 8.9 16.1 9.3 9.7 17.8

Source: UBS Asset Management. Data as of 30 June 2019.

For equities, we develop our Baseline, or five- and ten-year, Our valuation models indicate that the US equity market is
equity assumptions based on several inputs: projected overvalued, while most of the rest of the markets are near fair
earnings growth, dividend yields, and a reversion to fair value value or undervalued. We project for the next five years US
as determined by our proprietary valuation models. We base large-cap equity returns in the 5.0% range, global equities
these inputs on current economic conditions and market around 7.2%, developed ex-US around 8.7%, and emerging
pricing, analysis of historical trends and relationships, and our markets equities around 11.2% (note all returns are in USD
judgment about future trends. terms).

5
It is helpful to break expected returns down into their building outside the US we project an expansion of multiples to add
block components. For US equities, we expect valuation around 0.8% in local terms (boosted by a 1.1% appreciation
contraction of 1.3% per year, while in developed markets for undervalued currencies).

Exhibit 2 – Baseline 5-year expected returns: Equity return decomposition (in %)

UBS 5-Year Projections: June 2019

US US Dev
Equilibrium Large Cap Large Mkts Emg
US Large Cap History1 Cap x US Mkts ACWI

Inflation (Local) 2.0 2.9 2.0 1.4 3.0 1.9

Real Growth 2.3 2.2 2.2 1.9 3.9 2.3

Dividends 2.9 3.9 2.2 3.2 2.7 2.6

Valuation 0.0 0.7 -1.3 0.8 0.9 -0.3

Compound 0.2 0.3 0.0 0.2 0.4 0.1

Total Return Local 7.4 10.0 5.1 7.5 10.9 6.8

Currency Effects

Unhedged 0 0 0 1.1 0.2 0.4

Hedged 0 0 0 1.9 -1.9 0.4

Total Return Unhedged 7.4 10.0 5.1 8.7 11.2 7.2

Total Return Hedged 7.4 10.0 5.1 9.4 9.0 7.2

1
1926-2018: UBS estimates from SBBI data and Shiller (Yale University) Data as of 30 June 2019.

Exhibit 3 – Composition of 5-year returns: Local terms

12

10

-2
Large Cap Equilibrium Large Cap History US Large Cap Dev Mkts ex US Emg Mkts ACWI

Inflation (local) Real growth Dividends Valuation Compound

Source: UBS Asset Management. Data as of 30 June 2019.

6
Fixed income

7
To develop our fixed income Baseline return expectations, we Outside of the US, most developed markets are still under a
start with current yields and apply our judgment about the very low interest rate regime with real rates negative in many
future direction of interest rates (returns are today’s yield + markets. This, of course, augurs for low returns. Further
changes in price and income over time). We develop a cyclical undermining returns is that we project yield increases across
view of interest based upon sustainable levels given our most markets.
growth, inflation, and cash projections.

Exhibit 4 – Expected yields for 10-year government bond yields in five years and 10 years

June In 5 In 10 5-Year 10-Year

2019 % Years % Years % Increase % Increase %

US 2.01 3.10 3.20 1.1 1.2

Australia 1.32 2.50 3.50 1.2 2.2

Canada 1.47 2.90 3.20 1.4 1.7

Germany -0.33 0.90 1.40 1.2 1.7

France -0.01 1.15 1.55 1.2 1.6

Italy 2.10 2.15 1.85 0.1 -0.2

Spain 0.39 1.55 1.80 1.2 1.4

Japan -0.16 0.40 0.70 0.6 0.9

Switzerland -0.55 0.85 1.25 1.4 1.8

UK 0.83 2.50 2.75 1.7 1.9

China 3.23 3.50 3.60 0.3 0.4

Source: UBS Asset Management. Data as of 30 June 2019.

Exhibit 5 – Fixed income 5- and 10-year expectations

5-Year 10-Year 5-Year 10-Year
Expected Expected Local Expected Expected Hedged
Return Return Standard Return in Return in Standard
Local % Local % Deviation % USD Hdg % USD Hdg % Deviation %

Government Bonds US 1.0 2.0 4.6 1.0 2.0 4.6

Australia 0.1 0.8 5.8 0.8 1.1 5.8

Canada 0.2 1.4 5.8 0.5 1.6 5.8

Eurozone -0.8 0.0 6.5 1.6 2.1 6.5

Japan -1.0 -0.5 8.3 1.3 1.7 8.3

Switzerland -3.1 -1.4 6.1 -0.3 1.0 6.1

United Kingdom -2.1 0.1 8.1 -0.7 1.0 8.1

Global Government -0.2 0.7 5.6 1.2 1.9 5.6

Other US Corporates 1.4 2.8 6.7 1.4 2.8 6.7

US High Yield 3.7 4.5 11.1 3.7 4.5 11.1

US TIPS 1.5 2.3 5.3 1.5 2.3 5.3

EMD Hard Currency 4.5 4.6 6.0 4.5 4.6 6.0

Cash 3-Mo Libor 2.4 2.5 1.4

Note: We express equity returns in unhedged terms, but fixed income in hedged terms. Indices used are Bloomberg Barclays except for: EMD Hard
Currency which is the JPM EMBI Global Diversified and US TIPS, which is the ICE BoA Merrill Lynch US TIPS index.
Source: UBS Asset Management. Data as of 30 June 2019.

8
Cash

9
Although cash may seem like a ‘residual’ asset class, it is We base our cash projections off of current cash rates and our
actually a key component of the capital markets. First, it expectations of monetary policy. Over the longer run we expect
represents in nominal terms the least risky asset class and forms all central banks to move toward a sustainable rate that is
the basis from which we measure risk premiums and assess the consistent with the level of inflation and growth of a region.
attractiveness of equity and fixed income. Additionally, cash Below is our projection of nominal return for cash, realized
reflects monetary policy and the pricing of derivatives. inflation and subsequent real returns from cash for the next five
years for key markets. (Note: we define cash here as 3-month
Libor rates or their equivalent.)

Exhibit 6 – Expected 5-yr real cash returns for selected markets in local currency

Nominal Realized Real Cash

Region Cash Return Inflation Return

United States 2.5 2.0 0.5

Eurozone 0.4 1.7 -1.3

China 3.2 2.8 0.4

United Kingdom 1.4 2.2 -0.8

Japan 0.3 0.3 0.0

Switzerland 0.1 1.0 -0.9

Korea 2.6 2.0 0.6

Taiwan 2.2 1.9 0.3

Brazil 5.3 4.2 1.1

Mexico 5.5 3.3 2.2

South Africa 6.1 4.5 1.6

Turkey 14.0 11.5 2.5

Source: UBS Asset Management. Data as of 30 June 2019.

Exhibit 7 – Expected 5-yr real cash returns for selected markets

3.0%
2.5%
2.5% 2.2%
2.0%
1.6%
1.5%
1.1%
1.0%
0.5% 0.6%
0.5% 0.3% 0.4%
0.0%
0.0%
-0.5%

-1.0% -0.9% -0.8%

-1.5% -1.3%
-2.0%
Eurozone Switzerland UK Japan Taiwan China US Korea Brazil South Africa Mexico Turkey

Source: UBS Asset Management. Data as of 30 June 2019.

10
Currencies

11
We model currency movement based on a long-run trend to The systematic component is the impact of currency hedging
fair value as determined by our adjusted purchasing power versus the USD. It is simply the difference in nominal cash terms
parity measure (PPP) and the path of relative inflation and between the two currencies and is subtracted against the local
interest rate differentials (i.e., real rates). return to get the hedged USD terms. The high US rates relative
to most developed countries means that non-US investors must
To help investors assess currency impacts, we calculate exchange pay a ‘premium’ to hedge US returns to their local currency.
rate impacts in both hedged and unhedged terms and translate Thus, the high US rates are cut back to near local levels of
to any reference currency as necessary. The valuation return in EUR, JPY and CHF terms.
component of currency returns includes both our adjusted PPP
valuation of currencies as well as the path of inflation between
currencies and degree of convergence to fair value.

Exhibit 8: Expected total currency movements

5-Year 5-Year 10-Year 10-Year

Exchange Rate Effects in USD terms Valuation % Systematic % Valuation % Systematic %

EUR 0.8 -2.4 1.2 -2.1

GBP 1.0 -1.4 0.8 -0.9

JPY 2.6 -2.3 2.5 -2.2

CHF 0.0 -2.8 0.9 -2.4

CAD 0.8 -0.3 0.5 -0.1

AUD 0.3 -0.7 0.1 -0.3

CNY -0.2 0.5 -0.4 0.6

Baskets

EAFE 1.1 -1.9 1.2 -1.6

ACWII 0.4 -0.4 0.4 -0.4

ACWI x US 0.9 -0.9 0.8 -0.7

EME 0.2 1.9 -0.3 1.7

Global Gov 0.9 -1.4 1.0 -1.2

Global Gov ex-US 1.4 -2.2 1.5 -1.9

Global Credit 0.2 -0.3 0.2 -0.2

EMD Local 0.3 2.1 -0.3 1.9

Source: UBS Asset Management. Data as of 30 June 2019.

12
Alternatives

13
Exhibit 9: Net-of-fees projected 5-year returns and risk for alternative asset classes

5-Year 10-Year
Expected Expected
Equilibrium Return Return Standard
Return % USD % USD % Deviation %

US Real Estate (unlevered) 6.2 4.9 5.6 9.0

Hedge Funds (Low Vol) 5.2 3.9 3.9 4.3

Hedge Funds (High Vol) 5.8 4.5 4.6 7.4

US Private Equity 10.0 7.7 8.0 24.0

Global Private Equity (Unh) 10.0 9.1 9.5 24.5

Infrastructure (Unhedged) 6.6 6.2 6.5 17.1

Source: UBS Asset Management. Data as of 30 June 2019.

Note: We assume the following fees that have been deducted from gross-of-fee performance. US unlevered real estate 1.5%; hedge funds 1.8%, private
equity 2.5%; and infrastructure 2.0%. Actually, fees and total expenses will vary, and depend on base fees, performance fees, and administrative expense.
These figures are purely for illustrative purposes.

We model alternatives on a case-by-case basis. Some of them Thus, for unlevered real estate we expect returns in the 5.0%
– private equity, for example – can be viewed as a range and with leverage it is possible to see returns in the
straightforward extension of equities. Others, like hedge 6.5% range. We also see some pressured markets outside the
funds, are very idiosyncratic and are based on cash-plus US, after years of steady returns and above-average
models. By definition, private markets in real estate, equity performance; we temper our expectations and expect mid- to
and credit have less data to work with than public markets. low-single-digit returns.
However, enough history of these investments has
accumulated to allow for some understanding of the For private equity, we assume that the performance of these
relationships to public markets from which we can make some partnerships has an equity beta around 1.8 to 2.0.3 This
crude estimates of return. translates to a risk premium around 2.5% over large cap
stocks, which we apply in modeling private equity returns.
For the private asset sector, we need to distinguish between Thus, for US private equity we expect returns in the high
economic volatility and appraised volatility. Economic volatility single digits, but not on par with some of the touted returns
reflects the full risks of the asset class, including leverage, of the past.
concentration, and illiquidity. Appraised volatility is what
investors will see from their custodial statements. In our work, Theoretically, hedge funds are pure alpha plays with little to
we are projecting economic risk with the standard deviation, no beta exposures. In practice, we see a significant correlation
not appraised volatility. To illustrate this difference, compare to the equity markets (the quarterly data for various hedge
the five- and 10-year expected standard deviation shown in fund indices are 0.7 to 0.8). We believe that well-constructed,
Exhibit 9 above with the reported performance of alternative well-diversified hedge fund portfolios can have volatilities in
assets shown in Exhibit 27 (page 38). the 4.0% to 5.0% range (this is our Low Vol assumption),
while more aggressive portfolios will be around 7.4% (High
The most well understood alternative market is real estate. Vol assumption). We set the returns relative to cash and apply
The long run performance of unlevered property falls a constant risk premium over time.
between stocks and bonds in both return and risk. Although
good data for the US exists going back into the 1980s, it is
still based on appraisals and some careful analysis is needed to
develop proper risk characteristics.2

2
In the extreme, it can be argued that property is a type of equity investment and should have equity-like volatility. We won’t debate this point here, as
certainly the REIT market clearly has equity-like volatility. However, the nature of the property contract is quite different than holding equity shares. Even
the high end estimates of unlevered property to adjust for appraisal valuations don’t quite equal equity volatility. (Levered property does begin to
resemble equity volatility.)
3
The literature on the beta for various types of private equity are all over the place. See Korteweg, Arthur G., Risk Adjustment in Private Equity Returns
(November 16, 2018), available at SSRN: https://ssrn.com/abstract=3293984 or http://dx.doi.org/10.2139/ssrn.3293984 for a good review of the
empirical analysis. The betas from venture cap and buyout funds range from 0.6 to 3.2.

14
China in transition:
Five-year expectations

15
With the broader inclusion of China into global equity and fixed – The exchange rate is tightly managed. Although the bands
income indices, investors need to understand this unique have expanded and currency volatility has increased, this is
market. one of the largest currencies that is not completely free-
floating.
In our view, China is neither a traditional emerging market (i.e.,
commodity-driven) nor a developed one. It has some similarities – China’s equity market has historically been quite volatile. But
to its smaller Asian neighbors in its focus on export-driven volatility is declining, and while it is still out of line compared
growth. The government has gradually eased its control and to other markets, we believe it is trending towards emerging
ownership over many sectors of the economy and we expect market volatility levels. In the past, with an isolated and not
this to continue into the future, opening up significant well-developed institutional market, Chinese equity market
opportunities for investors. volatility was propelled in part by its large retail market. As
access to the Chinese market increases for institutional
Here are some unique characteristics of the Chinese markets: investors (which tend to have a longer-term outlook) and
the weight to China expands in the major indices, we
– Nominal government bond yields have been well below expect this volatility driver to continue to decline. It is worth
nominal growth rates since the early 2000s. Although many noting that this increasing share of institutional investors is
Western countries have also had yields well below the significantly driven by foreign investors as China continues
nominal growth rate since the global financial crisis (GFC), to open its financial markets. China has tightened rules
China had them before the GFC and we expect this trend to around trading suspensions, which has greatly reduced the
continue. number of trading suspensions in recent years. This is a
change from years ago, when Chinese companies were
– The working age population in China has started to decline. more likely to suspend trading in times of crisis.4
This should help keep yields low and we would expect the
productivity growth rate to drop with this. The heady days
of 10% real growth are of the past and we expect real (not
nominal) growth to be below 6% going forward.

Exhibit 10: Chinese nominal growth and yields Exhibit 11: China’s equity market volatility is gradually falling
towards the level of emerging markets

25 50

20 40
Last 12-Month Growth Yield in %

15 30

10 20

5 10

0 0
2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2005 2007 2009 2011 2013 2015 2017 2019
Nom Growth 10-Yr Gov Yld 3-Mo Gov Yld China EME S&P

Source: Macrobond. Data as of 31 March 2019. Source: MSCI China A, S&P, Morningstar Direct, based on monthly data. Data
as of 30 June 2019.

4
F or example, this Financial Times article cites more than 1,476 suspensions in the 2015 downturn:
https://www.ft.com/content/1bf693dc-24f9-11e5-9c4e-a775d2b173ca

16
Exhibit 12: 10-year standard deviation by market (local terms) through June 2019

30

25

20

15

10

0
Switzerland Global UK EAFE Australia EM S&P500 Taiwan S Korea Euro India Canada Japan HK Brazil China China
(ACWI) H Lcl

Note: Global market designated by MSCI ACWI Index, local terms.

Source: MSCI China A, Standard & Poor’s, Morningstar Direct. Data as of 30 June 2019.

– China’s historical correlation to the global market is low.

We would expect this to rise slightly in the future.

Exhibit 13: 10-year correlation to global equities by market (local terms) through June 2019

1.0

0.8

0.6

0.4

0.2

0.0
China Brazil India China Japan Taiwan S Korea Australia HK Switzerland UK Canada EM Euro EAFE S&P 500 Global
H Lcl (ACWI)

Note: Global market designated by MSCI ACWI Index, local terms.

Source: MSCI, Morningstar Direct. Data as of 30 June 2019.

17
– Unlike other emerging markets, China exhibits a negative stock-bond correlation in times of stress. Because of its size and
enormous investor base, which will demand a safe haven in times of turmoil, we believe that China will maintain this negative
stock-bond correlation during stressful periods. However, as indicated in the following chart, the relationship is not stable.

Exhibit 14: Rolling 12-month correlation: MSCI China A (in CNY) with China treasuries
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019

Source: MSCI, ChinaBond, Morningstar Direct. Data as of 30 June 2019.

Within the Chinese fixed income market, we follow three sectors: treasuries, policy banks (finance bonds), and corporates. Here are
the summary characteristics of these sectors according the Bloomberg Barclays Indices.

Exhibit 15: China’s bond market – Bloomberg Barclays Indices, June 2019
Yield to Size in Percent of
Issues worst Duration Maturity bn USD Aggregate Index

Treasuries 156 3.15 6.6 9.7 1,570 28

Policy Banks 189 3.51 4.4 5.3 1,789 32

Corporate 1,128 4.07 2.8 4.1 620 11

Aggregate 2,495 3.45 4.9 6.5 5,655 100

Source: Bloomberg LLC. Data as of 30 June 2019.

In summary, we project the following for Chinese assets in CNY terms. Although we believe that the domestic markets and
off-shore markets will converge over time, our equity assumption specifically refers to the on-shore market.

Exhibit 16: Summary 5-year return assumptions for Chinese asset classes
5-Year Baseline 10-Year Baseline

GRR % ARR % Std Dev % GRR % ARR % Std Dev %

Cash 2.4 2.4 1.3 3.4 3.4 1.3

China Comm Paper 3.4 3.4 1.4 3.7 3.7 1.4

China Government Bonds 2.6 2.7 4.6 2.8 3.0 4.6

China IG Credit 3.9 4.1 4.7 4.2 4.3 4.7

China Finance Bond 3.2 3.3 4.2 3.5 3.6 4.2

Domestic Equities 8.2 10.4 22.7 8.2 10.2 22.7

Source: UBS Asset Management. Data as of 30 June 2019.

In CNY terms, we expect Chinese equities to earn just over 8.0% over the next five to ten years. This represents a healthy premium
to both cash investments (around 2.5% to 3.0%) and to Chinese Government Bonds (around 3.0%).

18
Economic background:
Growth and inflation

19
Exhibit 17: Key market summary: Cyclical inflation and growth assumptions, in percent

Growth Inflation 3-Mo Libor

United States 2.0 2.0 2.5

Eurozone 1.5 1.8 1.0

United Kingdom 1.6 2.0 2.5

Switzerland 1.7 1.5 0.9

Japan 0.5 0.3 0.9

China 5.8 2.8 3.3

Global Markets 2.8 2.3

Developed Markets 1.7 1.8

Emerging Markets 4.6 3.2

Source: UBS Asset Management. Data as of 30 June 2019.

Over the next five years (2019-2024), we expect average Real growth on average should be slightly lower than in the
annual global growth of 2.8%, slightly lower than the 2.9% previous decade. This should translate to real earnings growth
average annual growth from 2014-2018.5 We expect inflation in the 2.8% range globally, with some markets having higher
to average 2.3% across the globe. We believe this will allow growth (emerging markets) and others having lower growth
the gradual normalization of monetary policy. We do not (Japan).
expect real rates to rise to their peaks of the 1980s or early
1990s, but we do expect several of them to finally become In general, the Anglosphere countries – United States, UK,
positive or only very slightly negative. Canada and Australia – should have inflation at or around
2.0%. In continental Europe and Japan, we expect lower
inflation: around 1.8% for the eurozone, 1.6% for
Switzerland, and 0.3% for Japan.

5
P ast growth is based on the IMF’s data in the World Economic Outlook, April 2019 edition. This is also based on our universe of countries that we cover
(approximately 36 countries), not the full universe in the WEO data.

20
Risk and correlation

21
The covariance matrix We expect equity market volatility generally to increase from the
Risk and correlation play an important role in portfolio low levels seen prior to the fourth quarter of 2018. After spiking
construction. We use a factor approach to build our with the GFC, equity volatility declined substantially into 2017.
covariance matrix. This approach produces an internally As seen in exhibit 18, volatility reached mid-single digit ranges
consistent covariance matrix with fewer assumptions than and has started to rapidly increase. We would not be surprised
needed for directly estimating volatilities and pair-wise to see further increases in the short run as trade issues and
correlations.6 geopolitical risks remain elevated. We also note the realized
volatility of the 10-year Treasury return is below its average and
we expect this to become more elevated in the future.

Exhibit 18: Historic volatility: 1997-2019 - S&P 500 and 10-yr Treasury
30

25
Rolling 12-month volatility

20

15

10

0
1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017 2019
Eq Vol 10-Year Vol

Source: S&P, Bloomberg Barclays, Morningstar Direct. Data as of 30 June 2019.

The stock-bond correlation has been persistently negative for next few years. However, investors must be aware that fits of
most of the last 20 years (after being positive for the prior 40 positive correlation will occur with surprises in inflation or if
years). We project slightly negative to zero correlation for the fiscal crises occur.

Exhibit 19: Historic stock-bond correlation: 1997-2019

0.8

0.6

0.4
Rolling 12-month correlation

0.2

0.0

-0.2

-0.4

-0.6

-0.8

-1.0
1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017 2019

Source: S&P, Bloomberg Barclays, Morningstar Direct. Data as of 30 June 2019.

6
 nother advantage is that it allows us to easily add new asset classes via the factor model, as well as model liabilities and other variables needed in
A
different analyses (wage growth, health care costs, etc.).

22
Part 2
Our methodology

23
Our methodology for setting market expectations
and valuation

This paper presents our approach to intermediate-term capital market expectations,

which we call our Baseline. We produce them by starting from very long-term capital
market expectations, which we call Equilibrium assumptions, and taking into account
current market and economic conditions as well as asset class valuation.

Equilibrium Capital Market Expectations (CME) are a vector of Asset prices tend to fluctuate around intrinsic value.
expected returns (cash rate plus risk premiums) and a matrix Discrepancies between market price and intrinsic value
of covariance for the asset classes in which we either invest or arise from market behavior and market structure, providing
may invest client assets. They are used to compute strategic opportunities to outperform. Common behavioral biases of
asset allocations and long-term simulations. In addition, the market participants include over-reaction to short-term noise
risk premiums included in the expected returns are used as and under-reaction to structural change. Our price/intrinsic
discount factors in our valuation models. These expectations value approach supports our ability to maintain discipline in
are the projection of average asset class behavior over the the face of short term noise and is forward-looking so as to
next 30-40 years or longer. We use the term Equilibrium incorporate structural change and key trends.
because we assume that economic and financial variables
ultimately regress to an internally consistent regime around a ValMod (Valuation Model) is a tool developed by Investment
long-term steady state. Solutions (IS) within UBS Asset Management, and used for the
fair valuation of various asset classes within local and global
These Equilibrium expectations are useful in indicating a markets. The model comprises a discounted cash flow (DCF)
“ballpark” benchmark for the portfolio, helping portfolio analysis which means a theoretical price for the considered
manager and client express in a common language what the asset, such as equity or bond, is derived.
suitable risk and return characteristics would be for a portfolio
over very long time periods. However, when establishing a The DCF technique is based on the estimation of the two
strategic asset allocation, we find it more useful to utilize main financial attributes of an asset: (i) all future cash flows
Baseline expectations, which include more information about to the investor stemming from the given asset need to be
the current market conditions. forecast, and (ii) proper assumptions on the applicable
discount rates have to be made in order to bring the derived
We believe that the intrinsic value of an investment is cash flows back in time via present value calculation. Finally,
determined by the fundamentals that drive its future cash the summation of all those discounted cash flows results in
flow. This is true for all investments in the capital market the model-based fair value of the underlying asset.
be they stocks, bonds or alternative assets. Our estimate
of intrinsic value is forward-looking and discounts back to Through comparing the corresponding current market
the present the future cash flows available to current and price to the fair value obtained from the model, portfolio
successor shareholders. It does not use shortcut ratios such as managers are able to draw conclusions on how to position
price/earnings, price/book value, price/sales or dividend/price themselves and decide whether to overweight or underweight
(yield) as proxies for intrinsic value. their exposures to certain asset classes or markets. Using
these valuation metrics we formulate our intermediate-term
expectations for a range of asset classes. We now present our
methods and estimates by broad asset class.

24
Integrating demand side and supply side estimates

There are two general approaches to setting capital market Our analysis, which includes balancing both a demand side and
assumptions: the supply side and the demand side. When supply side, approach indicates that equities should have a 3.7%
done properly, these two approaches tie in together and are risk premium over Libor. So, with a risk premium of 3.7%, 2.0%
consistent with each other. inflation and 1.7% real cash rate, we obtain a return of 7.4%.

Demand side Supply side

The demand side approach stipulates that investors demand The supply side approach looks at how the market can
a risk premium for taking on risk. It starts from cash as the deliver on the demands of investors. For equities, returns
riskless rate and adds on a premium that varies with the risk consist of dividend income and price appreciation. This price
taken. A simple approach is to start with current cash yields and appreciation can be further decomposed into earnings growth
add on historic risk premiums. So, for example, in the US from (which can be separated into inflation and real growth) and
1926 through 2018, US large cap equities have earned 10.0% changes in valuation. For fixed income, this approach begins
while cash (1-month Treasury bills) has earned 3.3%, providing with starting yields and adjusts for future movements in yields
a premium of 6.7% on a geometric basis. Thus, with 1-month and defaults (as well as prepayments or inflation depending
t-bills around 2.4%, a simple estimate of expected return is on the type of bonds). This approach follows the research by
9.1%. This approach is consistent with the “beta” approach Grinold, Kroner and Siegel.7
of the traditional Capital Market Pricing Model, where riskier
assets have a higher required return (risk premium). From the supply side perspective, we also start with inflation,
but instead look at payouts (which is dividend income and
For example, let’s consider the Equilibrium return for US Large buybacks). With inflation at 2.0%, we add on real earnings
Cap equities. From the demand side, we base the nominal growth (2.3%) and total payouts of 2.9%, which we divide
return off of a premium over 3-month Libor rates, which can into a dividend yield of 2.0% and buybacks of 0.9%. When
be further divided into expected inflation and the real cash we link these together geometrically, there is another 0.2%
rate. Currently, we set inflation at 2.0% and real cash at 1.7%. compounding effect to include.

Demand side Supply side

Expected Inflation 2.0% Expected Inflation 2.0%
+ Expected real cash rate 1.7% + Expected real earnings growth 2.3%
+ Risk premium 3.7% + Dividend yield 2.0%
+ Change in valuation 0.0% + Buyback yield 0.9%
Total 7.4% + Change in valuation 0.0%
+ Compound effect 0.2%
Total 7.4%
Source: UBS Asset Management. Data as of 30 June 2019.

Exhibit 20: Demand and supply side components: Equilibrium US equity returns
8%
Compound effect 0.2%
Buyback yield 0.9%
6%
Risk premium 3.7% Dividend yield 2.0%
4%
Expected real cash rate 1.7% Expected real earnings growth 2.3%
2%
Expected Inflation 2.0% Expected Inflation 2.0%
0%
Demand side Supply side

Source: UBS Asset Management. Data as of 30 June 2019.

7
 ichard Grinold, Kenneth Kroner, and Laurence Siegel, "A Supply Model of the Equity Premium," in B. Hammond, M. Leibowitz, and L. Siegel, eds.,
R
Rethinking the Equity Risk Premium, Charlottesville, VA: Research Foundation of CFA Institute, 2011

25
At a glance: The global investment universe

As a part of our annual review of our capital market believe that markets can overvalue or undervalue assets in the
expectations, we estimate cyclically-adjusted weights for all short term – this is, after all, the very basis of the valuation
asset classes we cover, listed and unlisted. Cyclically adjusted approach to investing. As a consequence, for the purpose of
refers to the median weight for each asset class over the estimating Equilibrium (or steady state) risk premiums, we will
10 years ending on 31 December 2018. From the point of need to estimate Equilibrium relative weights of each asset
view of finance, this is “the market,” against whose excess class. The cyclically adjusted weight for established public
performance we compute betas for the Capital Asset Pricing asset classes can be obtained by looking at the values of the
Model. The cyclically adjusted market value, in our estimate, asset’s capitalization over a period of time that covers at least
was about USD 101 trillion as of 31 December 2018. We also one full market cycle – for example, ten years.
report our estimate of current value for the market, which is the
sum of all asset class values as of end 2018. The value is USD Asset classes that are not listed on public markets have less
114 trillion, down from USD 121 trillion at the end of 2017. For frequent and complete data and we use a variety of sources,
“cash” we include the worldwide total of money market fund internal and external, to establish estimates.
assets provided by the Investment Company Institute.
The purpose of having capitalization weights for all the asset
Capitalization weights for listed asset classes (i.e., asset classes for which we provide capital market expectations is
classes that have liquid markets with at least daily pricing) twofold: the calculation of risk premiums using the Black-
are generally easy to find because indexes are capitalization- Litterman approach and the aggregation of asset classes to
weighted and the capitalizations of individual indices are used less granular aggregates (e.g., “rolling up” country- or region-
to compute world aggregates. The only problem is that, while specific equity markets to obtain a world equity aggregate).
we believe that markets are efficient on average, we also

Exhibit 21: Cyclically adjusted market capitalizations (as of December 2018)

Money Market Funds 5.4%

Infrastructure, Timber, Farm 0.3%
REITs 1.1%
Non-US Direct Real Estate 4.4%
US Direct Real Estate 2.3%
US Equity 19.7%
Private Equity 2.3%
High-yield Bonds 1.7%
EM Bonds (EMBI, CEMBI, GBI-EM) 3.0%

Non-US Industrialized Bonds 20.0% Non-US Industrialized Equity 15.0%

EM Equity 5.4%

US Bonds 19.4%

Source: UBS Asset Management. Data as of December 2018.

26
The function of tactical (short-term) expectations

Valuation measures, such as Professor Shiller’s Cyclically framework) provide useful information about market returns
Adjusted P/E Ratio or our proprietary valuation model ValMod over the medium term.
(which covers both equities and fixed income in a consistent

Exhibit 22: Cyclically adjusted S&P 500 Index P/E ratio vs. 10-year annualized total returns
25% 0

20%
10
Annualized TR over Next 10 Years

Shiller P/E (Inverted Axis)

15%
20
10%
30
5%

40
0%

-5% 50
1960 1963 1966 1969 1972 1975 1978 1981 1984 1987 1990 1993 1996 1999 2002 2005 2008 2011 2014 2017
S&P 500 Annualized Total Return over the Next 10 Years S&P 500 Index, Cyclically Adjusted P/E Ratio (Shiller CAPE)

Source: Standard & Poor’s, Macrobond; Professor R.J. Shiller; UBS Asset Management. Indexes are not directly investable. Data as of 30 June 2019.

For example, the picture above shows that the Shiller P/E ratio investor to sell at the bottom of the market. Therefore, our
mirrors 10-year annualized total returns for US stocks objective is to provide our clients with a smoother path –
(represented by the Standard & Poor’s 500 total return index). while taking advantage of the information from valuation
One clearly notices that when the (tan) P/E line is high, the models, we also consider shorter-tem market trends.
(dark brown) line of future total returns is high, and vice versa.
This is the difference between the functions of intermediate-
However, by showing the average return over the next ten and short-term expectations in our process. Investment
years we hide any volatility that actually occurred over those Solutions can provide returns analysis for a range of time
years. It is our experience that an investment strategy based frames upon request.
purely on valuation will experience massive volatility and, as
behavioral economics research has shown, may lead the

27
Strategic asset allocation

One of the most important decisions investors make is to Based on these criteria, it would not make sense to invest in a
determine their asset allocation. This sets the portfolio market that provides economic value but has no viable legal
composition in terms of asset classes, such as US stocks, system to protect ownership of this value. Brinson, Diermeier
European stocks, US bonds, etc., for reasons of achieving the and Schlarbaum considered over 80 asset classes and
best risk-return ratio for the investor’s needs. The strategic subclasses for inclusion in their Multiple Markets Index (MMI).
asset allocation (SAA) represents the average or so-called They describe the investable capital market as “primary
“normal” allocation. wealth-generating assets where sufficient markets have
developed and legal hurdles do not prohibit meaningful
Investors need to establish portfolio risk and return guidelines investment by tax-exempt investors.”
for a variety of reasons, including:
Per se, hedge funds are not an asset class – they are a legal
– Suitability of the portfolio (risk capacity and risk tolerance) form consisting of a multi-faceted universe of unconstrained
– Benchmarking and performance attribution managers investing in stocks, bonds and derivative
– Choice of instruments, manager styles and asset classes instruments. Since we already account for stocks and bonds,
– Liability-driven investing hedge funds are just like derivatives, that is, a strategy or
instrument based on other asset classes. We provide
An SAA is a clear and intuitive way to express guidelines and expectations for hedge funds for convenience, even if we do
quantify the “ballpark” for managers – since this approach has not believe them to be an actual asset class (violating the
been widely taught for decades in business schools, it is requirement of non-redundancy above).
understood by practitioners of different experience levels and
therefore helps convey risk and return expectations to clients. Its Arithmetic or geometric?
function is to examine long-term trade-offs between risk and We provide both arithmetic and geometric estimates of
return and help quantify expected gains from diversification. expected returns. Which should you use? Geometric returns
are the most intuitive as we use them any time we look at
Moreover, an SAA can be used in simulations to study something that grows with compounding. For example, if I
whether a client’s financial plan is sustainable (e.g., if a invest $100 in a stock and after 3 years the portfolio is worth
pension plan has a high likelihood of meeting future financial $130, my return rate will be (130/100)1/3 – 1 = 9.14% and not
obligations). 10% as one would obtain by computing the average
arithmetic return. The geometric return is what an investor
What is an asset class? would look at in this situation.
First of all, to understand SAA we need to define the relevant
investment universe. Brinson, Diermeier and Schlarbaum8 However, arithmetic returns are useful. In simple mean-
suggest “admission” criteria to the universe of investable variance optimization, we are facing a one-period problem
assets. The criteria are: and look for the asset allocation that gives the highest
expected return per unit of risk. There is no compounding in
– Analytical: this problem, and therefore arithmetic returns (or, more
– Adequate Control and Regulation precisely, excess returns) should be used to follow Modern
– Marketability and Liquidity Portfolio Theory.
– Meaningful Impact
– Non-redundant Interestingly, also in simulations of future values of a portfolio,
– Manageable Estimation Risk one should use arithmetic expectations. If, as is done
commonly, one assumes that returns are randomly drawn
– Rule of Law: from the lognormal distribution to randomly draw and create
– Contract Enforcement a large number of wealth paths (trials), the mean of the
– Shareholder Rights simulated wealth distribution will match the arithmetic return
while the median of the wealth distribution will asymptotically
– Talent Availability match the geometric return. Since, in a multi-year simulation,

8
 ary P. Brinson, Jeffrey J. Diermeier, and Gary G. Schlarbaum, “A Composite Portfolio Benchmark for Pension Plans,” Financial Analysts Journal,
G
March-April, 1986.

28
we will be interested in the geometric return, we will notice Over the long run, there is a strong relationship across asset
that in the special case of lognormal random numbers, classes:
geometric mean of the simulated wealth growth and median
of the distribution will be extremely close. – Cash has the lowest return and risk. Generally, the risk is in
the 0.5% to 1.5% range. One issue is that the more
What about discounting models? In general, since geometric granular the data, the greater the serial correlation.
growth rates of earnings and other variables are used in – Broad investment grade fixed income has volatility in the
discounted cash flow or Gordon models, it is logical to use 4.0% to 5.0% range, but is has been quite low in the last
geometric risk premia to discount such cash flows. couple of decades. Long fixed income has volatility in the
high single digits to the low double digits (8.0% to 12.0%).
Readers interested in these topics may find relevant – Non-investment grade fixed income shows volatility in the
information in the Ibbotson SBBI Yearbook, published 10.0% range, but this can vary quite dramatically, being
annually by Duff & Phelps. very stable in growth periods, but spike in volatility
suddenly.
Historical comparisons – Unlevered real estate generally falls in risk in the mid to low
The characteristics of asset classes can be very dependent on single-digit range when looking at quarterly data. However,
the time period chosen. It is not hard to find 10-year time as discussed in the alternatives section, the high serial
periods for which equities have underperformed fixed income correlation masks some of the actual annual volatility. Using
and there are even 20-year periods for which equities have some different measures, we get unlevered real estate
underperformed segments of the fixed income market. volatility to be around 9.0%-10.0%.
– Equities have volatilities that range as low as 12.5% for
For example, the 20 years up to June 2019 shows equities diversified portfolios to the 25.0% range for emerging
underperforming. markets countries. Generally, broadly diversified equities
have volatility set around 14.5% to 16.0%.

Exhibit 23 – Long-run asset class returns

Global Global
Global Fixed Short Global
Equity Income Fixed Global High Global Hedge
Unhedged Hedged Income Credit Yield ILBs Funds Property Commod Cash Inflation

Since 2001

Geometric Return 5.2 4.6 2.6 5.1 8.4 5.5 7.9 8.0 -0.5 1.4 2.1

Arithmetic Return 6.6 4.5 2.6 5.1 8.6 5.6 7.8 8.0 1.2 1.4 2.1

Standard Deviation 17.3 2.9 1.3 3.9 10.3 6.8 5.4 6.7 17.8 0.8 2.0

Serial correlation 0.1 0.0 0.3 0.1 0.3 0.2 0.3 0.9 0.2 1.0 -0.1

Correlation to Eq 1.0 -0.3 -0.5 0.3 0.8 0.4 0.3 0.1 0.5 -0.1 0.1

Since 2Q 2009

Geometric Return 12.0 4.1 1.4 5.9 11.5 4.1 4.8 8.7 -2.6 0.4 1.8

Arithmetic Return 12.6 4.6 2.6 5.2 9.1 6.0 8.2 8.0 1.3 1.3 2.1

Standard Deviation 15.3 3.0 1.3 4.0 10.6 6.9 5.5 7.0 18.3 0.8 2.0

Serial correlation 0.0 -0.1 0.3 0.1 0.3 0.2 0.3 0.9 0.2 1.0 -0.1

Correlation to Eq 1.0 0.0 -0.1 0.5 0.8 0.5 0.3 -0.5 0.6 -0.1 0.3

Global Equities MSCI All Country World NR USD Hedge Funds Credit Suisse Global Macro HF USD
Global Fixed Income Bloomberg Barclays Global Aggregate TR Hdg USD Property NCREIF Fund ODCE
Global Short FI Bloomberg Barclays Global Treasury 1-3Y Hdg TR USD Commodities Bloomberg Commodity TR USD
Global Credit Bloomberg Barclays Global Agg Credit TR Hdg USD Cash Ibbotson Associates SBBI US 30 Day TBill TR USD
Global High Yield Bloomberg Barclays Global High Yield TR Hdg USD Inflation Ibbotson Associates SBBI US Inflation
Global ILBs Bloomberg Barclays Global Infl Linked TR USD

Source: UBS Asset Management. Data as of 30 June 2019.

29
Exhibit 24: Historic performance: 1Q2001-1Q2019
10

Hedge Funds Property High Yield

8
Annual Arithmetic Return %

Global Equities
6
ILBs
Credit
Global Fixed Income
4

Global Short Fixed Income

2
Cash Commodities

0
0 5 10 15 20
Standard Deviation

Source: UBS Asset Management. Data as of 30 June 2019.

Exhibit 25: Historic Performance: 2Q2009-1Q2019

14
Global Equities
12
Annual Arithmetic Return %

10
Hedge Funds Property High Yield
8

6
Credit ILBs
Global Fixed Income
4
Global Short Fixed Income
2
Cash Commodities
0
0 5 10 15 20
Standard Deviation

Source: UBS Asset Management. Data as of 30 June 2019.

30
Equities In understanding the expected returns, it is helpful to break
Very long run estimates of equity returns generally show them down into their building block components. For US
about a 5.0% real return. For example, the Dimson-Marsh- equities, we expect valuation contraction of 1.8% per year,
Stanton research (Credit Suisse Yearbook 2014) has data going while in developed markets outside the US we project an
back to 1900 showing real returns ranging from over 6.5% expansion of multiples to add around 3.0% in local terms
(Australia, the US) to under 2.0% (Austria, Italy). (boosted by a 1.2% appreciation for undervalued currencies).

Our valuation measures indicate a large opportunity in Fixed income

emerging markets over the coming five years. We believe For fixed income, current yields are a good first estimate of
these strong valuation measures are supported by quite low subsequent returns. As the chart below shows, the five-year
P/E levels in these markets. index returns have closely matched the yields at the beginning
of the five-year period.

Exhibit 26: Bloomberg Barclays US Aggregate bond yield and subsequent 5-yr returns

20

15
Yield/Return (in %)

10

0
1976 1981 1986 1991 1996 2001 2006 2011 2016
Yld 5-Yr Ret

Source: Bloomberg Barclays, Morningstar Direct. Data as of 30 June 2019.

However, some notable gaps exist when yields move up or default rate is quite low around 2.0% and we project rates to
down substantially. In turn, we can improve return estimates rise from this level to their historic averages. Finally, we use a
by projecting future interest rate movements. We also adjust lower recovery rate for high yield bonds (around 40%).
for defaults in the corporate sector, inflation for inflation-
linked bonds and prepayment risks for securitized bonds. We also project that corporate spreads will widen from their
current levels. Along with the rise in government bond yields
For corporates and high yield bonds, we also project default projected above, we expect an additional spread widening of
rates and recovery rates. Historically, the cumulative default 25 bps for investment grade credit and 37 bps for high yield
rate for investment grade bonds over five years is small, on the bonds. This would put them in line with their historic averages.
order of 0.5% or so with a recovery rate around 60% upon We also present our Equilibrium returns, which is a different
default. High yield bonds have had much higher cumulative process and establishes a long run time horizon based on
default rates. Typically, they have averaged between 3.5% and sustainable growth levels over 20+ year time periods.
4.5% and will spike considerably in recessions. Recently the

31
Alternatives
By definition, private markets in real estate, equity and credit have less data to work with than public markets. However, enough
history of these partnerships has accumulated to allow for some understanding of the relationships to public markets from
which we can make some crude estimates of return.

Exhibit 27: Performance of alternative asset classes

Quarterly data: March 1994-March 2019

Real Real
US Estate Estate Private Fund of Equal Leveraged
Equity Unlevered Core Equity1 Funds Weighted Broad Commod Gold Loans

Geometric Return 9.5 9.3 9.0 14.3 5.3 8.9 7.4 2.3 4.8 5.4

Arithmetic Return 10.5 9.1 8.9 14.1 5.4 8.9 7.4 3.6 5.7 5.5

Standard Deviation 15.8 4.1 5.8 9.9 5.7 8.5 7.6 16.4 13.4 7.2

Serial Correlation 0.0 0.8 0.9 0.4 0.2 0.2 0.2 0.2 0.0 0.2

Skew -0.6 -3.0 -3.5 -0.6 -0.4 0.0 -0.1 -0.8 -0.3 -2.1

Kurt 0.7 11.8 14.6 2.5 2.9 1.6 3.4 2.0 1.4 25.4

Correlation to US Eq 1.0 0.2 0.2 0.8 0.7 0.8 0.7 0.2 -0.1 0.6

US Equity S&P 500 TR USD(1936) HF-Eq Wght CISDM EW Hedge Fund USD
Unlevered RE NCREIF Property HF-Broad Credit Suisse Hedge Fund USD
Core Real Estate NCREIF Fund ODCE Commod Bloomberg Commodity TR USD
Private Equity Cambridge Associates US Private Equity Gold LBMA Gold Price AM USD
HF-FoFs CISDM Fund of Funds Diverfd USD Leveraged Loans Credit Suisse Leveraged Loan USD
1
Through December 2018
Source: Morningstar Direct, UBS Asset Management. Data as of 31 December 2018.

The most well understood alternative market is real estate. going forward. Thus, for unlevered real estate we expect
The long run performance of unlevered property falls returns in the 5.0% range and with leverage it is possible to
between stocks and bonds in both return and risk. Although see returns in the 6.5% range. We also see pressured markets
good data for the US exists going back into the 1980s, it is outside the US. After years of steady returns and above-
still based on appraisals and some careful analysis is needed to average performance, we temper our expectations and expect
develop proper risk characteristics.9 mid- to low-single digit returns.

The historical record for real estate in the US has been quite The quarterly data indicate a high degree of serial correlation,
good. Returns approach equity levels but with notably lower which distorts the volatility. When we apply some adjustments
volatility. For the next few years, we have lower expectations and look at different annual periods, the serial correlation
for unlevered real estate. Like the US equity market, analysts remains (around 0.3), but the volatilities increase to around
have been surprised by the resiliency of the market, but 7.4%. The volatility for the Core index in this case is 10.0%.
expect softening with the slightly lower growth expectations

9
In the extreme, it can be argued that property is a type of equity investment and should have equity-like volatility. We won't debate this point here,
as certainly the REIT market clearly has equity-like volatility. However, the nature of the property contract is quite different than holding equity shares.
Even the high end estimates of unlevered property to adjust for appraisal valuations don't quite equal equity volatility. (Levered property does begin to
resemble equity volatility.)

32
For private equity, we assume that the performance of assets listed above there are no index funds with low fees.
partnerships has an equity beta around 1.8 to 2.0.10 This Instead, fees are significant and can vary widely. It is not
translates to a risk premium around 2.5% over large cap unusual for a well performing private equity partnership to
stocks, which we apply in modeling private equity returns. have total fees in the 5% to 8% range. To simplify, we assume
Thus, for US private equity we expect returns in the high that these are net-of-fees performance.
single digits, but not on par with some of the touted returns
of the past. The historic performance from the Cambridge As the historical data indicate, commodities have been a
series suffers from serial correlation (0.4) and we find disappointment for investors. The volatility has been larger
adjustments give estimates closer to 14.0% for this appraised than equities and the returns have been quite low. This would
volatility. However we believe that the true economic volatility be acceptable if the correlation to equities was negative (as
for a diversified private asset portfolio (buy-outs, venture cap, was argued 25 years ago when this asset class was introduced
mezzanine and special situations) is around 24.0%. to institutional investors), except that we see a slight positive
correlation with equities. With commodities becoming a
Hedge funds are one of the hardest asset classes to project. smaller part of inflation over time, their ability to hedge
Theoretically, they are pure alpha plays with little to no beta broader inflation risk has declined. Moreover, they are now
exposures. In practice, we see a significant correlation to the seen as correlated with the economic cycle. In the case of
equity markets (the quarterly data for various hedge fund huge commodity shocks (1970s, 2007-2008), commodities
indices are 0.7 to 0.8). We believe that well-constructed, may provide some hedging characteristics, but the case has
well-diversified hedge fund portfolios can have volatilities been far weaker than alleged.
in the 4.0% to 5.0% (this is our Low Vol assumption), while
more aggressive portfolios will be around 7.4% (High Vol Currencies
assumption). We set the returns relative to cash and apply a Historically, the currencies with the lowest volatility against
constant risk premium over time. the US dollar are Asian currencies. This is partly a result of
explicitly managed exchange rates (CNY, HKD) or partially
Fees are a thorny issue for alternative investments. For public managed (IRD, TWD, MYR). Developed countries tend to have
assets, we simply assume index-like performance, so the volatilities around 8.0% to 11%. Finally, the emerging markets
impact of fees can be very small. However, for the alternative exhibit the most volatility in the 10% to 15% range.

Exhibit 28: Return-risk graph: Currency vs USD Monthly July 1999- June 2019
5
CHF CZK
CNY NZD
HKD TWD ILS CAD DKK AUD
KRW
0 PLN
Annualized Cumulative Return %

SAR MYR
GBP EUR JPY NOK,SEK HUF
BRL
CLP COP ZAR
INR MXN IDR RUB
-5

EGP
-10
TRY
-15
ARS

-20
0 5 10 15 20
Standard Deviation

Source: UBS Asset Management. Data as of 30 June 2019.

10
T he literature on the beta for various types of private equity are all over the place. See Korteweg, Arthur G., Risk Adjustment in Private Equity Returns
(November 16, 2018), available at SSRN: https://ssrn.com/abstract=3293984 or http://dx.doi.org/10.2139/ssrn.3293984 for a good review of the empiri-
cal analysis. The betas from venture cap and buyout funds range from 0.6 to 3.2.

33
Also, notice that except for the emerging markets, the returns Exhibit 29: Longer-term US dollar moves
do center around zero. In fact, the trade-weighted index for
advanced countries has a 0.4% return over the last 20 years. Number of Cumulative Annualize
Months Change % Change %
However, there does appear to be longer trends in the US
dollar. The late 1990s was the ‘Rubin Strong Dollar’ period Oct 1994-Feb 2002 89 47.3 5.4
which saw the dollar appreciate 42% in a steady trend in an Post-Tech decline 75 -26.0 -4.7
eight-year period. This over-valuation was gradually corrected
GFC Flight to Quality 11 21.0 23.1
from 2002 to March 2008, when global market began to
sense the impending global financial crisis and a huge flight- Post-GFC decline 26 -18.4 -9.0

to-quality occurred with the dollar appreciating 21% in an Since 2011 98 33.9 3.6
11-month period. Starting in March 2009, the markets began
to rally and the dollar started to decline again in value over
the next two years. Since 2011, the dollar has generally been Last 20-years 240 8.4 0.4

rising in value, though recently it has started to show signs of Since 1994 302 41.1 1.4
hovering around current levels. Source: Macrobond, Federal Reserve: Broad US Trade-weighted FX Index.
Data as of 30 June 2019.

Exhibit 30: US trade-weighted exchange rate: Broad index

120

110
Late 1990s: GFC flight-to-quality
strong dollar period
100
2014–2016:
Strengthening
Post-tech decline
90

80

70
1994 1997 2000 2003 2006 2009 2012 2015 2018

Source: Macrobond. Data as of 30 June 2019.

34
As to serial correlation, most currencies exhibit no serial correlation, though there are a few outliers. On a short-term basis,
currencies are hard to predict, but over the longer run we do see longer periods of moves toward fair value.

Exhibit 31: Serial correlation of currencies (monthly data, April 1999-June 2019)

0.4

0.3

0.2

0.1

0.0

-0.1
CHF BRL KRW CAD EBP ZAR NZD HKD IDR HUF GBP NOK EUR MYR DKK CZK ILS SEK CLP MXN JPY AUD COP PLN INR ARS TRY TWD RUB CNY

Source: Macrobond. Data as of 30 June 2019.

Overall, from first principles we would expect currency returns the performance of foreign assets is sometimes negatively
to be zero on a geometric basis, have volatility around 6% correlated with foreign currency movements. Australia,
to 12% (depending on individual securities or baskets) and being a large commodity producer, often sees the AUD
have a zero correlation to local and foreign capital markets. strengthen as the global economy improves (higher demand
Several large currencies exhibit these features (USD, euro). for commodities). Thus, as foreign equities increase, their
However, several currencies exhibit strong movement with currencies depreciate, creating a negative correlation. In
local and foreign markets. For example, from an emerging this instance, currency hedging doesn’t lower volatility, but
market perspective as well as a few developed markets, increases it.

35
Sources
Gary P. Brinson, Jeffrey J. Diermeier, and Gary G. Schlarbaum (1986) “A Composite Portfolio Benchmark for Pension Plans.”
Financial Analysts Journal, 42(2): 15-24.
John Y. Campbell and Robert J. Shiller (2001) “Valuation Ratios and the Long-Run Stock Market Outlook: An Update.”
NBER Working Papers: 8221).
Credit Suisse “Global Investment Returns Yearbook 2014.”
And Ibbotson Associates (2015) Stocks, Bonds, Bills and Inflation (SBBI) Yearbook. Chicago, IL: Morningstar.
Richard Grinold, Kenneth Kroner, and Laurence Siegel, “A Supply Model of the Equity Premium,” in B. Hammond, M. Leibowitz,
and L. Siegel, eds., Rethinking the Equity Risk Premium, Charlottesville, VA: Research Foundation of CFA Institute, 2011).

36
For marketing and information purposes by UBS. For professional clients / qualified / institutional investors only.
This document does not replace portfolio and fund-specific materials. Commentary is at a macro or strategy level and is not with reference to any
registered or other mutual fund.

Americas our business, a number of risks, uncertainties and other important factors
The views expressed are a general guide to the views of UBS Asset could cause actual developments and results to differ materially from our
Management as of July 2019. The information contained herein should expectations.
not be considered a recommendation to purchase or sell securities or any UK
particular strategy or fund. Commentary is at a macro level and is not Issued in the UK by UBS Asset Management (UK) Ltd. Authorised and
with reference to any investment strategy, product or fund offered by UBS regulated by the Financial Conduct Authority.
Asset Management. The information contained herein does not constitute APAC
investment research, has not been prepared in line with the requirements This document and its contents have not been reviewed by, delivered to
of any jurisdiction designed to promote the independence of investment or registered with any regulatory or other relevant authority in APAC. This
research and is not subject to any prohibition on dealing ahead of the document is for informational purposes and should not be construed as an
dissemination of investment research. The information and opinions offer or invitation to the public, direct or indirect, to buy or sell securities.
contained in this document have been compiled or arrived at based upon This document is intended for limited distribution and only to the extent
information obtained from sources believed to be reliable and in good permitted under applicable laws in your jurisdiction. No representations are
faith. All such information and opinions are subject to change without made with respect to the eligibility of any recipients of this document to
notice. Care has been taken to ensure its accuracy but no responsibility is acquire interests in securities under the laws of your jurisdiction.
accepted for any errors or omissions herein. A number of the comments Using, copying, redistributing or republishing any part of this document
in this document are based on current expectations and are considered without prior written permission from UBS Asset Management is prohibited.
“forward-looking statements”. Actual future results, however, may prove Any statements made regarding investment performance objectives, risk and/
to be different from expectations. The opinions expressed are a reflection or return targets shall not constitute a representation or warranty that
of UBS Asset Management’s best judgment at the time this document was such objectives or expectations will be achieved or risks are fully disclosed.
compiled, and any obligation to update or alter forward-looking statements The information and opinions contained in this document is based upon
as a result of new information, future events or otherwise is disclaimed. information obtained from sources believed to be reliable and in good
Furthermore, these views are not intended to predict or guarantee the faith but no responsibility is accepted for any misrepresentation, errors or
future performance of any individual security, asset class or market omissions. All such information and opinions are subject to change without
generally, nor are they intended to predict the future performance of any notice. A number of comments in this document are based on current
UBS Asset Management account, portfolio or fund. expectations and are considered “forward-looking statements”. Actual future
EMEA results may prove to be different from expectations and any unforeseen risk
The information and opinions contained in this document have been or event may arise in the future. The opinions expressed are a reflection of
compiled or arrived at based upon information obtained from sources UBS Asset Management’s judgment at the time this document is compiled
believed to be reliable and in good faith, but is not guaranteed as being and any obligation to update or alter forward-looking statements as a
accurate, nor is it a complete statement or summary of the securities, result of new information, future events, or otherwise is disclaimed.
markets or developments referred to in the document. UBS AG and / or You are advised to exercise caution in relation to this document. The
other members of the UBS Group may have a position in and may make a information in this document does not constitute advice and does not
purchase and / or sale of any of the securities or other financial instruments take into consideration your investment objectives, legal, financial or tax
mentioned in this document. situation or particular needs in any other respect. Investors should be aware
Before investing in a product please read the latest prospectus carefully and that past performance of investment is not necessarily indicative of future
thoroughly. Units of UBS funds mentioned herein may not be eligible for performance. Potential for profit is accompanied by possibility of loss. If
sale in all jurisdictions or to certain categories of investors and may not be you are in any doubt about any of the contents of this document, you
offered, sold or delivered in the United States. The information mentioned should obtain independent professional advice.
herein is not intended to be construed as a solicitation or an offer to buy Australia
or sell any securities or related financial instruments. Past performance is This document is provided by UBS Asset Management (Australia) Ltd, ABN
not a reliable indicator of future results. The performance shown does not 31 003 146 290 and AFS License No. 222605.
take account of any commissions and costs charged when subscribing to Source for all data and charts (if not indicated otherwise): UBS Asset
and redeeming units. Commissions and costs have a negative impact on Management
performance. If the currency of a financial product or financial service is The key symbol and UBS are among the registered and unregistered
different from your reference currency, the return can increase or decrease trademarks of UBS.
as a result of currency fluctuations. This information pays no regard to
the specific or future investment objectives, financial or tax situation or
particular needs of any specific recipient.
The details and opinions contained in this document are provided by UBS
without any guarantee or warranty and are for the recipient’s personal use
and information purposes only. This document may not be reproduced,
redistributed or republished for any purpose without the written permission
of UBS AG.
This document contains statements that constitute “forward-looking
statements”, including, but not limited to, statements relating to our future
business development. While these forward-looking statements represent
our judgments and future expectations concerning the development of

37
 www.ubs.com/am-linkedin

© UBS 2019. All rights reserved.

AMMA-2469 7/19
www.ubs.com/am

For professional / qualified / institutional clients and investors only.