Strategic asset allocation – it’s still a low return world
This edition of Investment Insights, provided by AMP Capital Investors, examines setting an investment portfolio's strategic asset allocation to each asset class and looks at the implications for strategic asset allocation and the subsequent implications for investors.
Thursday, February 19th 2004, 11:14AM
Setting an investment portfolio’s strategic asset allocation (SAA) to each asset class (global shares, bonds, etc) within a diversified investment portfolio is the most important decision most investors make. We have previously argued that this should not be a set and forget decision but should be reviewed annually with respect to a range of factors including changes in markets (or valuations).
It might be tempting to think that the double digit rebound in share markets over the past year indicates that we can look forward to a return to the sustained high returns of the 1990s. However, our analysis indicates that while the bear market is probably behind us we are still in a low return world and that SAA should be set accordingly.
Economic and financial backdrop
The economic and financial backdrop continues to point to lower returns over the next ten years than was the case in the 1980s and 1990s.
- Over the 1980s and 1990s returns from equities and bonds were boosted way above long term (and sustainable) averages by the adjustment from high inflation to low inflation which allowed a sharp fall in interest rates. This produced big capital gains in bonds and equities as bond yields and earnings yields also fell. The adjustment meant that share prices rose a lot faster than growth in underlying corporate earnings (reflected in the rise in price earnings multiples). This once-off adjustment has now run its course with global inflation around two per cent. Any further fall in inflation (ie deflation) might be good for bonds but would not be good for equities, whereas a rise in inflation would be bad for both asset classes as bond and earnings yields would have to rise (PEs fall);
- In the early 1980s price earnings multiples on stocks were low (dividend yields were high) and bond yields were high thus setting the scene for the high average returns of the 1980s and 1990s. Today this is not the case;
- Starting in the 1980s, the ascendancy of free market economics as represented by tax cuts, deregulation, privatisation, free trade and smaller government was positive for private enterprise and hence equities. However, many of these policies have now run their course or are subject to some public backlash;
- The collapse of Communism in the late 1980s ushered in the so-called “peace dividend” in the form of reduced government spending on defence thus freeing resources for use by the private sector. With the “war on terror” this is now subject to some reversal with defence spending now rising again and the heightened terrorist threat adding to investor uncertainty; and
- The sharp fall in share prices between March 2000 and March last year weakened investor confidence in shares and the experience of the 1930s and 1970s suggests that it will take a while to fully recover.
Indicative return expectations
For most assets current prevailing investment yields provide the best guide to returns for the next five to ten years.
- For equities, a simple model of current dividend yields plus trend nominal GDP growth does a good job of predicting medium-term returns. This approach allows for current valuations (which are picked up via the yield) but avoids getting overly complicated. The next chart shows this approach applied to US equities. It can be seen that it broadly tracks the big swings in equity returns (and points to lower returns ahead);
Source: Datastream, Global Financial Data and AMPCI
- For property, a similar approach works well using current rental yields and likely trend inflation as a proxy for rental (and hence capital) growth; and
- For bonds the best predictor of future medium-term returns is the current bond yield.
Using this framework and assuming trend economic growth and inflation (2.5 per cent pa) for each country our projected (pre tax and fees) returns over the next decade for the major asset classes is as follows.
Projected medium term returns, %pa
|
|
Dividend
yield +
|
Growth
# =
|
Return
|
|
US
|
1.5
|
5.25
|
7.0
|
|
UK
|
3.7
|
5.0
|
8.5
|
|
Europe
|
2.4
|
4.5
|
7.0
|
|
Japan
|
1.0
|
3.0
|
4.0
|
|
Asia ex J
|
2.5
|
8.0
|
10.5
|
|
World, local currency
|
1.9
|
4.9
|
7.0
|
|
World, $A *
|
1.9
|
5.9
|
8.0
|
|
Australia
|
3.7
|
6.0
|
9.5
|
|
Direct Property
|
7.0
|
2.5
|
9.5
|
|
Aust Listed Property
|
7.0
|
2.5
|
9.5
|
|
Global Listed Property ^
|
7.0
|
2.5
|
9.5
|
|
Aust Bonds
|
5.5
|
0.0
|
5.5
|
|
Aust Cash
|
5.25
|
0.0
|
5.25
|
# Assumed to equal projected nominal GDP growth.
* Assumes the $A falls 1% pa.
^ Assumes forward points averaging 1% pa (ie, a positive return to hedging back to $A).
Source: AMPCI and Datastream
Key points to note:
The return projections for equities are broadly consistent with the economic and financial backdrop considerations noted in the previous section.
- The low return projections for both equities and bonds generally reflect their low starting point yields. Current low yields for both bonds and stocks contrast dramatically with the situation in the early 1980s, ie at the start of the last secular bull run in both stocks and bonds, when Australian and US bond yields were around 14 per cent or more and dividend yields were around six per cent or more (and PEs were nine times or less);
- For equity markets to do better than projected above:
- earnings growth must be much faster than assumed, which seems unlikely – eg US profits are already high relative to US GDP and if anything the historical record suggests that earnings growth lags GDP growth by two per cent pa or so; or
- share prices must rise faster than profits, ie price earnings multiples must rise – which again seems unlikely given that PE multiples are still quite high; or
- in the case of global shares the $A needs to fall by more than we have allowed for.
- Countries with relatively high dividend yields (like Australia and the UK) and/or strong growth prospects (like non-Japan Asia) should do relatively well. By contrast after outperforming over the last decade the US may well be a relative under performer. Note that the projection for Australia excludes franking credits which adds one per cent or more to post tax returns;
- Non residential property – whether listed or unlisted – is likely to do relatively well because it comes with relatively high yields compared to equities and bonds (around seven per cent plus) and thus only requires modest capital growth to provide a decent return. This does not apply to residential property where values have surged over the last few years, yields have collapsed and imbalances indicate the risk of a bust.
Obviously there are alternative scenarios. But the key point is that the medium term outlook remains for relative modest returns from both equities and bonds.
Implications for investors
The projected returns above imply a five per cent or so real (ie after inflation) return for a typical balanced growth portfolio with a 70 / 30 per cent mix of growth/defensive assets. While this is well below the average of the 1980s and 1990s it is more in line with long term average return that such a combination of assets would have delivered over the last century as a whole. In other words it was the huge returns of the 1980s and 1990s that were the aberration.
The lowered nominal return environment will mean more negative return years (as a result of the average return level shifting down), an increase in market volatility (as investors shift between asset classes in search of higher returns) and a greater deficiency in retirement savings for Australians compared to what otherwise might have been the case had the high returns of the 1990s continued.
Implications for strategic asset allocation
Our return projections are little different from when we last looked at this over a year ago. The obvious thing to do is to look for ways to boost returns to investors portfolios (without taking on more risk) and to make sure that they are roadworthy for the changed environment. Areas to look at remain those that offer high yields or were out of favour through the 1990s. These include:
- Maintain a bias towards Australian equities for now. On a long term basis numerous considerations suggest that Australians should have a greater exposure to international equities than is currently the case. However, likely returns continue to favour Australian over global stocks (particularly after tax) on a medium term basis (notwithstanding the cyclical rebound in markets at present where Australia lags as it is a defensive market);
- Increase exposure to non-US markets within global equities. For example, non-Japan Asian equities come with a higher dividend yield similar and a higher growth potential than the US and are still only into the early stages of a secular recovery relative to global stocks;
- Maintain a decent exposure to non-residential property. Unlisted property fell out of favour in the early 1990s following the crash in office values and the realisation that it is illiquid. However, non-residential property yields remain high relative to those available on bonds and equities and it only requires modest capital growth to produce returns in excess of those available from bonds and equities. Listed property is also attractive for the same reason, and it is worth considering global listed property investments as they also offer high yields but with diversification benefits and unlike Australian listed property trusts trade at a discount to net asset value; and
- Reduce traditional fixed interest exposure in favour of property/corporate debt and cash. Low government bond yields suggest low returns from bonds going forward. One way to offset this is to take on more yield via property investments or corporate debt, but at the same time offset the extra risk involved by taking on more cash. Note that through the 1990s fixed interest provided a decent (three per cent or so) return premium over cash. With bond yields now much lower and having had adjusted to low inflation – this is unlikely to be the case in the coming decade. As a result the incentive to hold traditional fixed interest over cash is diminished.
Additional considerations
- Firstly, the issue of risk also needs to be considered. As returns on traditional asset classes fall there is an inclination to move out along the risk frontier towards sexier products. Often these involve products, which performed well during the bear market, but down side risk is forgotten. An example is hedge funds.
- Secondly, the issue of liquidity also needs to be allowed for. The illiquidity (and related high transactions costs) associated with direct property suggests that investors need to be sensible in increasing their unlisted property exposures despite the attractive relative returns shown above.
- Thirdly, these comments relate to our strategic view – they do not allow for tactical considerations, which suggest that global equity markets should continue to move higher over this year and outperform relatively defensive asset classes like Australian shares. This does not change our strategic view.
- Finally, the $A cannot be ignored. On a strategic basis, having gone from undervalued to now overvalued the $A bull run is now getting long in the tooth.
Conclusion
The cyclical rebound in equity markets of the last year or so does not alter the low medium term return outlook for shares. It does not signal a return to the sustained high double-digit returns from shares that occurred in the 1982 – 2000 global bull market. Absolute share valuations are too rich and the economic backdrop less favourable to indicate that. The medium term outlook remains for share returns to average around high single digit levels over the next five to ten years. Likewise low bond yields presage low medium term returns from bonds.
Within this context an ongoing strategic bias towards high yield investments is desirable. Short of taking on a lot of extra risk it will be very difficult to achieve the sort of double-digit returns achieved through the 1980s and 1990s over the next ten years. However, investors do need to look at their strategic asset allocation to equities, property, bonds and cash to make sure that it will achieve the maximum return for their given risk tolerance.
This article was written by Dr Shane Oliver, Chief Economist and Head of Investment Strategy, AMP Capital Investors
Commenting is closed
