- Consistent with our recent focus on objective-driven investing, we believe that investors should rethink their approach to asset allocation and look across asset classes to identify the performance characteristics that they want—income, gains and alpha—when constructing their asset allocations.
The investment world is dominated by the notion that equities will provide the highest return over the long term and that bonds should be included in a portfolio to offset some of the risk that comes with equities while providing incremental income. The remainder of the portfolio—typically 10% or less in most asset allocation models—includes either cash as a safe haven or alternatives that can potentially provide uncorrelated return streams.
We would like to challenge these ideas in the current environment. We believe that income should drive (and has driven) portfolio returns over the long term and that income can come from any sector of the market. Additional allocations should be made seeking gains and alpha to supplement the return from income. In effect, we believe investors should consider replacing the traditional “stocks, bonds and alternatives” with “income, gains and alpha.”
As can be seen in Exhibit 1, dividends have been a key contributor to equity returns over the long term. Looking back even further, from 1871 through 1982 the cumulative return from dividends was roughly equal to the compounded return on price. The same held true for the period from 1958 through the start of the bull market for equities in August of 1982. It was primarily during the bull market of 1982 through 2000 when price appreciation significantly eclipsed income as a source of returns for equities.
Since the end of the bull market, equity price returns have been less than 3% per annum and have been spectacularly wild, despite recent low levels of volatility. This has occurred despite the fact that earnings have reached record levels and have nearly doubled since the end of the bull market. This growth has been offset by a dramatic decline in price/earnings (P/E) ratios to levels that are closer to long-term averages.
As earnings have grown and P/E ratios have reverted toward the mean, the idea that equities are now fairly priced has been reinforced and investors may believe that equities should be the best performing asset class over the long term. The data that support this notion are heavily biased by the price performance of the 1980s and 1990s. During this period, equity prices soared as deregulation, leveraged financing, globalization, technological revolution and positive demographics propelled them. These data are also buoyed by substantially higher dividend levels than are available today.
Traditional bonds—the constituents of the Barclays U.S. Aggregate Bond Index—are also offering less coupon income today than they have historically. In recent years the return from bonds has been consistent with long-term expectations but has been less dependent on income and more dependent on price appreciation, as extraordinary monetary policy has reduced yields to historic lows.
The benefits are obvious: higher bond returns and cheap financing rates for households and businesses. The risks are all in the future: potential inflation, low coupons and negative price returns. This puts investors in a precarious position. If bond coupons are lower than they have been historically and price returns become negative or zero (assuming rates either rise or remain at current levels, respectively), an investor has to assume that traditional bond returns will be below expectations over the short to intermediate term. At the same time, if inflation increases, then real returns on traditional bonds could be decidedly negative.
Putting Income at the Core
By starting with income as the primary objective, an investor can more easily estimate long-term returns. If an investor pursues this path, then income generating assets should represent a significant percentage of the investor’s overall portfolio and should be characterized by the objective rather than by the sector in which the asset resides. While this approach may reduce the expected level of total return over the long term, it can significantly reduce volatility over both the long and short runs and can reduce the tail risk associated with traditional equity beta-dependent portfolios.
Corporate pension plans have been moving in this direction, but for different reasons. They view a pension plan as an asset/liability hedge. Industry practices and regulations dictate how they must value their liabilities. As a result, they are shifting assets into security types that behave more like their liabilities. This approach typically focuses on investing in long-term Treasuries and long-term corporate bonds as the core assets. While these institutions are not typically thinking of liability-driven investing as an income-based approach, technically it is. By shifting assets out of more volatile equity markets and into bonds, they are increasing portfolio income and reducing portfolio volatility. They are not only reducing absolute levels of volatility, but also are seeking to reduce tracking error versus their liability.
The key difference between this approach and what we are suggesting in this paper is the duration of the assets. In a corporate pension plan, the liabilities tend to be long-dated. This leads plan managers to buy long-dated assets to match the performance characteristics of the assets.
For investors that are more total-return focused, the duration of these assets should shift tactically as the relative attractiveness of the prevailing interest-rate environment changes. With rates today at or near historic lows and default risk being relatively low, we would suggest low-duration, high-income (more credit sensitive) assets as core portfolio holdings.
Note that the objective of generating income results in a different type of allocation than either traditional asset allocation or the major indices would suggest. In our suggested approach, allocations would be to higher yielding sectors of both the equity and bond markets, rather than the core constituents of either the S&P 500 or the Barclays Aggregate. The net impact should be that overall portfolio risk either remains constant or can be reduced while portfolio income rises. It also seeks to balance the potential sources of return by increasing the dependence on income and reducing the dependence on price appreciation.
By focusing on gains as a secondary objective, allocations would also differ substantially. With dividends low and equity prices high, we might consider gains from commodities and currencies as well as from stocks. The stocks that we choose under this model would likely be higher beta names that are globally diversified, and the percentage allocation to these types of opportunities would likely be significantly lower than the equity-heavy, broad market focused allocations we typically see under the traditional model.
A key concern with broad market equities is that given the dramatic rise in after-tax corporate profits, one would at least hope that dividends would be near their highs rather than their lows. They are not. This is a complex situation that has more to do with tax policy than economics. Corporations today can manufacture anywhere and sell everywhere. This means that a US company may manufacture something in Asia and sell it in South America. This generates profits for the company, but these profits must be repatriated and taxes must be paid before companies can distribute them to shareholders in the US as dividends. Many choose not to, based on the sound economic reasoning that roughly 50% of those profits will be owed as taxes, despite favorable treatment of dividends. This substantially reduces income as a source of return and largely explains why companies today are more likely to return money to shareholders through stock buybacks instead of dividend distributions.
Based on the data above, hedge funds have not completely succeeded in fulfilling their promise. As a result, their reliability as a source of alpha should be questioned. So where will the uncorrelated alpha come from? Alpha can come from multiple sources, including excess returns from active management of bonds, equities and alternatives as well as uncorrelated returns from commodities, direct investments in real assets (i.e., timber and real estate) and selected hedge fund strategies. But unlike bonds and stocks, there is likely to be much more dispersion around the market benchmark in alternatives based on the available data.
Establish Reasonable Objectives
If maximizing total return is an objective that provides excessive dependence on equity beta, then what would a reasonable long-term return objective be? In the foundation and endowment world, the prevailing objective is CPI plus 5%. This is a perfectly pragmatic objective as they are typically building a pool of capital over time and are required by law to pay out 5% of their assets each year to maintain their tax-exempt status. So if they grow assets by CPI, they preserve the purchasing power of their dollars, and if they pay out only earnings above CPI their capital will last indefinitely. Given the dearth of positive yielding, inflation-sensitive assets, these investors largely assume a total return target. They make significant allocations to private equity, venture capital and other alternative strategies. A complementary strategy would be to assume a long-term inflation forecast (say 3%) and establish a target return equal to this assumption plus their 5% excess return target (3% + 5% = 8%).
This is essentially what public pension plans do. They generally have a fixed-return target that is based on their assumed long-term rate of return. Today this target is generally around 7.75%. Achieving this level of return enables them to maintain their current funded status. Exceeding this objective allows them to increase funded status and meet their long-term liabilities.
In both of these cases, shifting from sector-based asset allocations to income, gains and alpha would increase the probability of achieving their return targets. In periods where income is insufficient to meet these objectives, additional diversification into gain-seeking assets and uncorrelated alpha would be justified. In periods of higher yields, it would be rewarding to have a majority of the portfolio in income producing assets.
Insurance companies tend to hedge their liabilities rather than seek to maximize total return. They have always focused more on income as an objective. But they too must re-think their asset allocation as coupon income opportunities have diminished and must now consider alternative sources of income. Unfortunately, risk-based capital rules limit their options and will likely force them to stay in traditional bonds.
Individual investors typically have a less clearly defined objective and are heavily influenced by the notion that over the long-term, equities have the highest expected returns. Target date funds and most investor portfolios typically have more than 50% of assets allocated to equities, resulting in significant equity beta dependence.
We would suggest that these investors, collectively and individually, should seek to reduce their dependence on equity beta and increase their dependence on income for long-term returns. Income provides the benefits of compounding, self-funded dollar cost averaging and lower volatility. We would also point out that income is especially beneficial in tax deferred investment programs as it is generally taxed at an investor’s top marginal rate.
Sector-based asset allocation has been and likely will be challenged to deliver on its promise in the coming years. We believe this approach has led to an overdependence on equity beta, significant volatility and disappointing returns since the end of the bull market in August of 2000. We suggest that pension plans and individual investors re-assess their investment objectives and define an investment plan that has a higher probability of achieving more modest and concrete objectives. Income-producing equities, higher-yielding bonds and income producing alternatives should be combined to form the core of the portfolio, and this core should generate returns primarily through income.
Price gains should represent a secondary source of return as they can be extremely volatile and, at times, unreliable and can potentially be generated from sectors other than equities. Investors should also seek alpha to provide uncorrelated return streams that can complement both income and price gains.
We believe income, gains and alpha should be the building blocks of a portfolio regardless of source. Taxable investors will have to make adjustments to take into account the favorable tax treatment of capital gains and dividends. Regardless of tax status, investors can create a potentially higher yielding, less volatile portfolio. Those who focus on maximizing total return as their primary objective will have to weather the volatility of the equity markets and, while they may have a higher expected return, will have to accept a lower probability of achieving the desired result.