Capital Market Assumptions

2020 Long Term Capital Market Assumptions

Executive summary


We use our building block methodology to estimate long-term CMAs for equities. This approach reflects a total return approach to equities — accounting for both income and capital appreciation (i.e., the change in price over time). The building blocks therefore consist of estimates for yield (as a driver of income) and earnings growth and valuation change (as drivers of capital appreciation).

A closer look at how emerging-market equity CMAs differ from developed ones

Emerging Market (EM) economies are structurally different than Developed Markets (DM) leading to differences in the way their CMA building blocks are estimated. Even amongst the broad EM category are economies different enough to justify individual marginal adjustments. Some EM economies like Korea and Taiwan, are more mature in their economic development, are export oriented, and have similar characteristics to DM economies. China, however, is a high growth economy, driven by credit and money growth, that is moving to more long-term sustainable growth levels and is much more oriented to domestic growth.

Emerging Market Building Blocks (including Hong Kong as more than half of market capitalization of Hong Kong listed firms are Chinese companies):

  • Yield. For Korea and Taiwan, our yield approach is the same as DM (ex-US). Hong Kong’s market focuses on recurring trailing dividend yield. With many Hong Kong listed firms being family or state owned, non-recurring special dividends can occur, so we exclude special dividends from our analysis to prevent outliers within our yield estimates.
  • Earnings growth. For Taiwan and Korea, we follow the same process as for DM. China used to be a high growth market and is now slowing down. Because of this, we adjust the historical average growth by calculating the decline of other Asian economies that have been in similar economic positions and deduct that figure from China’s 10-year average real GDP growth.
  • Valuation change. Similar to DM, we assume valuation such as the price-to-book ratio will return to the long-term mean after adjusting for macroeconomic variables. For Taiwan and Korea, exports make up more than 60% and 40% of GDP, respectively. As export driven economies, currency (FX) has a bigger impact than inflation on valuation change. For Hong Kong, growth is influenced by inflation in China. As the HK$ is pegged to the USD, we do not look at FX but look at the HK “risk free” rate as liquidity conditions are influenced by the HK$ peg and therefore the US Fed’s monetary policy.

Fixed income

Within fixed income, we also utilize the building block methodology, seeking to isolate and identify the individual drivers of the specific asset class risk premium. As with equities, the drivers of return for fixed income are income (yield) and appreciation (roll return, valuation change, and credit loss).


A significant, and increasing, number of institutional portfolios contain private or alternative assets1. This trend is likely due to shrinking expected returns and yields in traditional public assets. Private asset market capitalization has grown to $5.8T globally in 20192, Private Equity, a large subset, has experienced 7.5x growth since 2002 compared to 3x in public markets2. Composed of a broad array of heterogeneous investments, private assets are anything but standardized. As the space is evolving to include new assets and creates unique challenges to investors, we attempt to assess the economics of some common types of investments.

Private equity

In the 2020 long-term CMA methodology we will present our views on; Private Equity, specifically Leveraged Buyouts (LBO), Private Direct Real Estate (DRE), and both Private Infrastructure Equity and Debt. To properly introduce private assets into a portfolio, we suggest taking a building blocks based approach to understand and forecast return, which we will address in detail in the following sections.

While an investor in public assets can simply buy an index of an asset class, own a portion of the universe, and experience average results, an investor in private assets cannot. To align our private CMAs with our public CMAs, but still provide the custom nature private asset classes require, we built enough flexibility into our private asset models to analyze the whole market, an individual fund, or a single deal.

To emphasize the underlying reason for a customizable private model, there is simply no investable benchmark for private assets. These assets are unlisted on any tradeable market, provide at-best quarterly reporting or tender dates, and lack transparency of the underlying investments required to create a proper benchmark.

Hedge Funds and Listed Real Assets

Estimating returns for such investments is more complex than evaluating equities and fixed income, as the range of alternatives (”alts”) available runs the entire spectrum of risk. Long/short strategies, for example, behave differently than commodities, and both behave differently than global macro. And for any alternative category, it can be a challenge to know how much of the return is true, uncorrelated alpha, and how much can be attributed to broad market exposures (e.g., S&P 500 Index). In fact, academic research (Hasanhodzic and Lo, 2007; and Fung and Hsieh, 2004) suggests that a significant portion of hedge fund returns is attributable to conventional asset class and factor risks. Leaning into this research, we construct linear models using available market indexes from our traditional asset class CMAs, and measure the proportion of the estimated returns and volatility that are attributable them.

Our expectations

About our methodology

We employ a fundamentally based “building block” approach to estimating asset class returns. Building blocks represent a “bottom-up” approach in which the underlying drivers of asset class returns are used to form estimates (Figure 13).

First, these sources of return are identified by deconstructing returns into income and capital gain components. Next, estimates for each driver are formed using fundamental data such as yield, earnings growth and valuation, and combined to establish estimated returns.

By incorporating fundamental data, our approach allows for the relative attractiveness of asset classes to vary over time. Other approaches based on historical relative returns can provide relatively static risk-premiums through time in which certain asset classes contain constant return advantages. The following sections will detail and present the estimates across various equity, fixed income and alternative asset classes.