ETF STRATEGIES

A multi-asset-class portfolio can achieve the ideal balance between risk and return

How to build a perfect ETF portfolio

There are various ways that exchange traded funds (ETFs) can be used as part of the investment strategies for both long-term and shorter-term traders. Here is how you can achieve diversification using a multi-asset class portfolio of ETFs.

By investing in one asset class, even on a diversified basis, investors are basically stuck with the risk and return characteristics of that asset class.

As seen in the chart at right, equities, for example, typically produce the highest long-run returns but also have the greatest return volatility (as most recently seen during the GFC, when Australian equity returns slumped 38% in 2008). Cash and bonds tend to produce lower returns but with less volatility or “risk”, as broadly defined. At least over the past decade or so, listed property also had volatile returns, largely due to significant swings before and after the GFC.

Wouldn’t it be great to be able to develop a portfolio that offers the correct blend of risk and return that you are most comfortable with? By investing across asset classes, that is precisely what multi-asset-class portfolios aim to do.

What’s more, another advantage of a “multi” over “single” asset class portfolio is even greater diversification, allowing investors to reduce risk without necessarily sacrificing a lot in returns.

The key reason is that, as with different securities within an asset class, asset classes themselves tend to perform differently over time. For example, equities tend to do best in periods of strong economic growth with low inflation and interest rates. Bonds do best in high (but falling) interest rate environments, while commodities perform best in high-inflation environments.

Indeed, in the 12 years from 2004 to 2015, the correlation in annual returns for Australian bond and equities was negative 0.75. Equities tended to do well when bonds did not, and vice versa.

Blending different assets within a portfolio, therefore, can allow you to better withstand changing market environments. Indeed, research shows that blending different asset classes with low correlation in their returns into a single portfolio can often produce a better long-run return performance – for a given level of year-to-year portfolio return volatility – compared with only owning a single asset class, such as only bonds or equities.

The exact asset blend that is best depends on each investor’s relative risk and return preferences.

The table below shows the average annual returns – and the volatility (or standard deviation) in those returns – investors would have experienced had they invested across a range of asset classes over recent years, with weights attached to each asset class varying within their portfolio. The weights used broadly conform to typical weightings used within the financial industry for long-run “model portfolios” for different risk exposure.

“Higher-risk” portfolios with a greater weighting toward equities would have produced the highest average annual returns over this period but also with the greatest return volatility.

For investors, however, the main point is that there are now several relatively cheap and transparent ETFs that can provide broad exposure to each asset class. Depending on the particular ETFs chosen, average management fees for developing a similar multi-asset class portfolio could be about 0.5%pa, or even less.

For example, relatively young investors with several decades until retirement might opt for a higher-risk core portfolio, comprised mainly of equities and relatively fewer lower-return cash and bonds. However older investors, either in or close to retirement, might prefer a less volatile portfolio that provides steadier annual income returns, as might best be achieved with a portfolio heavily weighted toward cash or bonds.

Further diversification is also possible through including international equities, foreign currencies and even commodities, such as gold.

ETFs offer a way for investors (especially those running a self-managed superannuation fund) to develop a highly diversified multi-asset portfolio that meets their desired risk-return profile (as decided either by themselves or in consultation with their financial planners) at an often much cheaper cost than using actively managed funds.

STRATEGIC ASSET ALLOCATION

One issue to consider when developing a multi-asset portfolio is how actively managed you would like it to be over time. Are you a “set-and-forget” investor, or one who would like to try to actively tilt your asset allocation based on market conditions.

Strategic portfolios are those that blend exposures to different asset classes based on the long-run expected risks and returns of each asset class. As such, these portfolios aim to provide the best long-run returns for a buy-and-hold investor, subject to their maximum risk tolerance. Clearly, investors who either can’t stomach a lot of return volatility or can’t risk a major loss (as they’re either in or near retirement) will need to settle for more investments in less volatile (or lower-return) assets, such as cash or bonds. Those who can afford to take on more risk can invest in more growth assets, such as equities.

Although expert views differ on the margin, the broad asset allocations referred to in the table are likely to be close to what many industry professionals would regard as reasonably optimal strategic asset allocations for different risk levels. As noted, investors might seek even greater diversification by swapping some Australian equity exposure (usually around a third to a half) for international exposure, or allocate 5% to 10% of their portfolio to commodities and/or international currencies.

For long-term strategic investors, all that is then required is to “rebalance” their portfolio at regular intervals. Opinions differ on the appropriate timing, but a reasonable compromise might be once a year.

For example, if equities have done particularly well in a given year (for example, their weight in your portfolio has increased from, say, 50% to 60%), rebalancing would involve selling down some equities at year end so that their overall weight fell back to 50%. Instead, you would buy exposure to the asset classes that had not done so well, so that their weight in your portfolio was restored to the long-run desired levels.

Annual rebalancing allows investors to take advantage of any short-run momentum within asset classes within a given year (which can be a feature of markets), while also eventually reducing exposure to any potentially “overheated” asset classes so that the portfolio remains in line with the investor’s risk profile. Annual rebalancing also has the advantage of ensuring any realised capital gains are eligible for a capital gains tax discount.

TACTICAL ASSET ALLOCATION

For strategic investors, the only changes to their multiasset portfolio over time would be for rebalancing purposes. For tactical investors, however, changes in the asset allocation might also be pursued based on either a fundamental and/or technical analysis of respective asset classes.

For example, your strategic asset allocation might suggest that 60% of your portfolio should be in equities and 40% in bonds. The tactical inter-asset class decision, however, is whether you should invest more or less than 60% in equities over the short term due to expected short-term asset class returns.

Indeed, although equities might be expected to return 10%pa on average over the long term, year-to-year asset class returns are rarely smooth. Over any given period (or say two to five years), it is likely that some asset classes (such as equities or bonds) will produce returns above their long-run expected average, while other asset classes will produce below-average returns. In theory, if we knew in advance which asset classes would produce the best returns over the short run, we could tilt our portfolio accordingly and enjoy even higher long-run returns.

Of course, determining which asset classes will perform best over the short run is not always as easy in practice.

• Fundamental or valuation-based strategies

One way to predict likely short-run returns between asset classes is based on valuations.

Most valuation-based tactical asset allocation strategies tend to exploit regression to the mean. When, for example, asset returns have had a strong run, their valuations tend to reach expensive levels. Regression to the mean dynamics suggest that a period of belowaverage returns for that asset class is then likely for a few years. Similarly, after a major market downturn that forces valuations to cheap levels, regression to the mean analysis suggests a period of above-average returns is then likely.

• Momentum- and trend-based strategies

Another more mechanical option is to make tactical switches between asset classes relying on relative price momentum and/or trending behaviour.

Indeed, financial research suggests that momentum effects – or the tendency of asset classes (or asset class sub-segments) that have performed strongly (poorly) over the recent past to keep performing well (poorly) for a while longer – are evident in many markets. This happens because markets can be slow to recognise and adjust to new fundamentals, and also because often “herd-like” behaviour can develop among investors.

In essence, asset prices seem to be driven by momentum in the short term but by valuations over the long term. These two forces are reconciled by corrective “regression to the mean” periods, which push prices back to fair value once momentum has driven them too far (either to the upside or downside).

In fact, if momentum effects become too strong, it can create “bubbles” where valuations reach extreme levels. This often results in large, jarring corrections at some stage. America’s dot-com bubble earlier last decade is one such example.

Related to momentum investing is trend investing, such as only investing in an asset class that is trending up based on, say, whether current prices are above or below a moving average of past prices. Trend signals could also be used in place of relative performance in choosing between investments – that is, invest in one investment choice when the ratio of its price relative to that of an alternative investment choice is trending up.

Of course, the trick in trend or momentum investing is to jump on an identified wave as early as possible – and also try to get off before it comes crashing down! In choosing a time period, there’s a trade-off between being sure that an apparent shift in relative performance is not simply noise versus having a signal early enough to profitably act on it.

Longer-term time periods (say one or two years) suffer from lags. This means that one can miss out on an early trend and hold on too long once a trend changes. Investing in something that has performed well for numerous years also exposes oneself to the risk of a corrective “regression to the mean” period. Indeed, it is often noted that past performance is not necessarily an indicator of future performance. In fact, it may be a potentially better indicator of a poor performance ahead!

However, very short-term trends (say a few days to a few weeks) may be too erratic, meaning one may overreact and jump on trends just as they are about to change once again. This problem is known as “whipsawing”.

The trick in momentum investing is to jump on a wave as early as possible – and get off before it crashes

Most research on momentum investing tends to find that a past performance period of between six months to one year tends to perform best, in terms of being most likely to predict the relative strength of returns over the next month or so.

CORE/SATELLITE PORTFOLIOS

Some investors may decide that they want to try to add an “alpha strategy” overlay that attempts to beat the benchmark performance of one or all core indexed asset class investments in your portfolio. In this case, ETFs also then lend themselves to another complementary strategy: core/satellite investing.

The idea behind core/satellite investing is to first use low-cost index funds (such as ETFs) to achieve, at a minimum, the “beta” returns from each asset you seek to be exposed to. This is the “core” portfolio element. The second, or “satellite”, element of the portfolio is to add some portfolio “tilts” within one or more asset classes to try to generate extra “alpha” returns within each asset class.

Core/satellite investing has several benefits. Most obviously the approach allows for more granulated risk control. What you’re doing is effectively unbundling the “beta” (or market) and “alpha” (market-beating) returns from a certain desired asset class. By investing in an index fund, you know exactly how much of your portfolio will be “beta” and track a certain asset class – and so how much exposure you may then wish to allocate to “alpha” investments that may be more volatile but offer potentially higher longer-run returns within that asset class.

With active funds, you never can be too sure of your beta versus alpha portfolio exposure, especially if the fund’s senior managers and/or investment strategies change.

By retaining core beta exposure, moreover, you can also probably afford – at the margin – to consider more aggressive (though volatile) active investment opportunities that may have an even lower correlation with your beta investment holdings, and be more likely to outperform over time.

• Passive v active managed funds

One source of alpha investment exposure you could seek is via an actively managed fund – assuming that you can find one you are happy with and are confident might offer market-beating returns.

For example, if you decide you want 60% of your portfolio to be in Australian equities, the traditional approach would be to allocate, say, 20% to three separate actively managed funds. Each fund would be expected to at least provide returns that matched the market, plus a little extra “alpha” return. You’d then watch each one carefully and swap funds when and if you thought that one fund was not doing its job, or you had a specific new investment theme (such as resources or global markets) that you wanted to pursue.

A core/satellite approach, however, might instead invest, say, 45% of the money earmarked for equities into a single low-cost “core” equity ETF, and the remaining 15% equity allocation among the preferred actively managed funds, which are known in this case as equity “satellites”. The overall equity allocation remains 60% in each case, though there’s clearer separation between alpha and beta components with the core/satellite strategy.

This approach has numerous advantages. For starters, it can help reduce management fees. By buying an actively managed fund, you might be paying a high management fee to a manager, even though a good chunk of their return is only beta. You might instead get this beta return with a lower-cost ETF. For example, if all three active funds charged a 1.5% management fee and an equity ETF charged only 0.25%, then allocating three-quarters to the ETF and one quarter among the active funds would reduce overall equity management fees to 0.56% – more than halving the cost.

Another advantage is that core/satellite investing also saves on transaction costs and taxation, if you decide to swap active managers. For example, if your portfolio contained 20% of an active fund that consistently lagged the market, then selling it for something else would involve turning over one-fifth of your portfolio. But if, due to a larger index fund holding, the active fund accounted for just 5% of the portfolio, then making the change only involves turning over a much smaller portfolio share with lower transaction costs and lower potential capital gains tax to pay.

Of course, while you might leave decisions over how to beat asset class benchmark returns to an active manager(s), you might also try to do it yourself, through various intra-asset class tilts. In this case, ETFs can provide both the core and satellite parts of your portfolio.