2020 Strategic Asset Allocation and Capital Market Assumptions Update Contents At UBS, our Capital Market Assumptions (CMAs) and Strategic Asset Alloca- tions (SAAs) are the fundamental underpinnings that inform our investment 2 Capital Market Assumptions: advice. The Chief Investment Office (CIO) and the Wealth Management USA The Basics (WM USA) Asset Allocation Committee (AAC) regularly review our CMAs and SAAs as asset class expected returns and risks change over time. The lat- 4 Why a CMA update was est review determined that it is appropriate to update the CMAs, which last necessary occurred in February 2019, as well as make changes to the recommended 4 Changes to the major asset SAAs, which last happened in February 2017. class CMAs For the strategic CMAs, meant for one full business cycle, most equity returns 5 How we use CMAs are relatively unchanged, while fixed income returns have declined signifi- 5 SAAs: Allocations are changing cantly in this update. Overall, these CMAs should result in notably lower for CMA and structural reasons returns for more conservative SAAs over the cycle, a lower return for moder- ate SAAs, while aggressive risk portfolio returns are only modestly impacted. 7 Final thoughts: What does this The equilibrium CMA returns, which are applicable for a multi-business cycle mean for investors? horizon, are modestly lower for fixed income and comparably higher for equities. The risk CMAs are the same for both the strategic and equilibrium returns, and the risk estimates are relatively unchanged in this update. Conse- quently, the impact on SAAs’ estimated total risk is modest. The SAA updates reflect changes in the CMAs and include structural modifica- tions designed to enhance portfolio performance and alignment with the UBS Liquidity. Longevity. Legacy. framework. The structural changes include reduc- tions in the cash and hedge fund allocations and breaking the US government bond allocation into three maturity categories. As a result of these changes, the total equity allocation increases between 1% and 5% across all SAAs. There is a small reallocation from US investment grade corporate credit into US high yield corporate credit and emerging markets fixed income in a few SAAs. In addition to changes to the core UBS House View SAAs, updates have been made to all-fixed income, yield-focused, and sustainable investing (SI) SAAs. Liquidity. Longevity. Legacy. disclaimer: Timeframes may vary. Strategies are subject to individual client goals, objectives, and suit- ability. This approach is not a promise or guarantee that wealth, or any financial results, can or will be achieved. 2020 Strategic Asset Allocation and Capital Market Assumptions Update The following sections explain what the CMAs are, the dif- directions. Correlations fall on a range from –1 to 1, where ferences between strategic and equilibrium CMAs, the –1 indicates a perfectly negative correlation, 0 indicates no economic assumptions driving the CMA update, and expla- correlation, and 1 indicates a perfectly positive correlation nations for the major asset class CMAs. Following that is an between the asset classes. explanation of SAAs and their objective, the changes made in this update and their rationale, and a review of the SAAs’ In our last CMA update, we introduced equilibrium return new estimated returns and risks, resulting from both CMA assumptions. The difference between the strategic and equi- and SAA changes. librium return assumptions is their investment time horizons. Specifically: Capital Market Assumptions: The Basics Strategic returns reflect our expectation for the average Prudent asset allocation decision-making starts with reason- annual total return for various asset classes over one full able assumptions about the risk and return prospects for a business cycle. This includes the recovery, expansion, slow- range of asset classes. In total, we estimate CMAs for over down, and recession stages of the cycle. The length of busi- 100 distinct asset classes, including stocks, bonds, hedge ness cycles isn’t fixed, though in advanced economies they fund strategies, private equity, and private real estate. The have been getting longer on average. The starting points for return assumptions represent the average annual estimated estimating returns are typically current values, such as bond return we expect from a particular asset class. The risk yields or equity price-to-earnings ratios, and then we project assumptions include the volatilities, which measure the dis- a future path for this return driver to a long-term sustainable persion of returns, and correlations, which estimate how level. asset classes move together over time. Our process of devel- oping CMAs is not a “hard science” using repeated, con- Equilibrium returns are expectations for average annual trolled experiments, but our methodology combines both returns over multiple business cycles. We assume that asset objective and subjective elements to use the most depend- class return drivers eventually converge to their long-run able aspects of a data-driven approach and enable us to level, which is the economy’s sustainable steady state. Fur- overlay expert judgment in the process where necessary. thermore, we assume that it takes at least 10 but not more than 20 years for the economy to reach this steady state. The Return assumptions represent the average annual estimated expected return is calculated assuming return drivers and return we expect from a particular asset class. They do not returns themselves are in equilibrium. By definition, equi- represent the return of any particular security or investment. librium returns are not influenced by cyclical developments To derive strategic expected returns, we explicitly model the or current circumstances. Instead, they reflect structural expected returns individually for the coming years. assumptions about the economy, including the long-term potential growth rate and the neutral rate of interest. Our volatility estimates seek to quantify the dispersion we expect in the asset class returns on a year-by-year basis. The purpose of having two sets of CMA return assumptions Of course, expected volatility for any particular asset class is to distinguish the average annual return estimates for comes with a range of caveats. For example, an investment investors when their investment horizon changes from one manager or a fund structure might be more or less volatile to multiple business cycles. The strategic returns are applica- than the asset class as a whole depending on factors like the ble for short- to long-term investment horizons over a cycle, investment objective, manager style, or constraints. Addi- such as when constructing strategic asset allocations. In tionally, we’re also making an assumption that asset classes contrast, equilibrium returns are appropriate for multi-cycle themselves will be diversified; one particular position within investment horizons. an asset class can certainly exhibit different risk and return characteristics than the asset class as a whole. We have only one set of volatility and correlation assumptions. Unlike expected returns that can vary quite a bit depending Finally, correlation measures the interaction of two asset on starting economic conditions and valuations, risk proper- classes over time. Positively correlated asset classes tend ties tend to be stable and mean-reverting over time, and are to increase or decrease in value at the same time, whereas unlikely to change from one economic cycle to the next. negatively correlated asset classes tend to move in opposite 2 of 19 April 2020 2020 Strategic Asset Allocation and Capital Market Assumptions Update For the primary asset classes, Fig. 1 lists the updated strategic attractive, but the prospect for earnings growth in the US and equilibrium CMA return assumptions and the updated and internationally has deteriorated. risks, which represent the annualized standard deviation of returns. The full set of CMAs can be found in Appendix 1. The equilibrium returns are generally higher than the stra tegic The updated strategic return CMAs are significantly lower returns, with a few exceptions. For fixed income, the higher for fixed income; moderately lower for hedge funds, private equilibrium returns reflect the fact that US Treasury bond equity, and private real estate; but are the same or higher for yields are significantly below our estimate for their steady equities. Since our last CMA update, US interest rates across state values. Once yields have converged to the steady state, most of the yield curve have fallen to historically low levels, the annual returns for US government fixed income will be greatly reducing fixed income returns. In contrast, equity higher than our strategic return CMA estimates. To estimate valuations relative to bonds have become considerably more the equilibrium returns for all other asset classes, we add an appropriate risk premium to these risk-free equilibrium returns, which results in higher equilibrium returns than stra- tegic returns. Fig 1 Both the strategic and equilibrium returns are point esti- Primary asset class risk and return assumptions mates, but the actual annual returns can deviate quite a bit In % from these returns. An investment that increases in value by 20% one year might easily decline by 10% the follow- Strategic Equilibrium ing year. Combining the return and volatility estimates for Asset Class Ann’l total Ann’l total Ann’l return return risk an asset class provides a more realistic outlook for “normal” US Cash 1.6% 2.6% 0.2% returns. For example, US large-cap equity returns could fall between –9.8% (i.e., 5.7% – 15.5%) and 21.2% (i.e., 5.7% US Government Fixed Income 0.4% 3.4% 3.8% + 15.5%) over 12 months. This range of estimated “normal” US Municipal Fixed Income 0.4% 2.8% 2.8% equity returns is illustrated in Fig. 2 with a number of simu- US Corporate Investment Grade Fixed Income 1.3% 4.6% 5.3% lated paths for US large-cap equities over one year.