Bonds: Too Much For Now, Or Too Little?

Government bonds had a tough first quarter, recording losses, and so far the second quarter is seeing a bit more of the same as market rates edge higher on signs the US economy is doing better than expected and the eurozone crisis has eased.

Shares are looking good, although the big sell off around the world on Wednesday is a reminder that the current rebound in equities isn't solidly built.

Gold, oil and other commodities were also sold down and the Aussie dollar fell to around $US1.0250, a three month low.

If it continues, investors will again start looking at bonds for safety.

AMP Capital Investors' Chief Economist and Strategist, Dr Shane Oliver looks at whether holding too many bonds is a good or bad thing for investors.

There has recently been much debate about whether Australian super funds have too much in shares and not enough in bonds.

The basic argument is that compared to other major countries Australian pension funds have a higher share allocation and a lower bond allocation and that this leaves members exposed should shares plunge.

And after shares have underperformed bonds in recent years such concerns resonate.

There are several aspects to this debate.

Why is the share allocation relatively high?

Is now a good time to be thinking about switching into bonds?

Is the real issue the underdeveloped corporate bond market?

Share allocations - super versus total wealth

Roughly speaking the Australian superannuation system has about 50% invested in equities and 18% in bonds.

The share allocation is higher than that of other major countries and the bond allocation is lower as can be seen in the following table.

While the share allocation in super may be relatively high, this does not appear to be the case if Australian households' total wealth is looked at.

Thanks to a huge allocation to housing, Australians tend to have a much higher allocation to non-financial assets compared to other countries and a relatively low exposure to shares outside of superannuation.

As such the total allocation of Australian households to shares as a proportion of total household wealth is not out of line with households in other countries, being above the UK, Japan and Germany but below the US and Canada.

Shares in super

There are several reasons why Australian superannuation funds may have a higher allocation to shares.

The first thing to note is over long periods shares provide higher returns than bonds and cash.

This can be seen in the next chart which shows that since 1900 Australian shares have returned 11.9% pa compared to 6% pa from bonds.

Similarly the next chart shows that over most rolling ten year periods shares provide higher returns than bonds, and when every so often they don't the share return usually rebounds.

Second, Australian shares have tended to do better than most over the long term, so the higher allocation to shares which has been biased to Australian shares may have served Australians' well (notwithstanding poor recent years).

Third, dividend imputation provides a tax boost to the return on Australian shares equal to around 1.5% pa.

In a world of relatively low returns where a diversified mix of assets is expected to return around 8% pa, this is quite significant.

Fourth, the government bond market is relatively small thanks to years of budget surpluses and asset sales which have led to a relatively low level of public debt.

On average, countries in the OECD have a stock of government bonds equal to around 100% of GDP to invest in.

By contrast, in Australia gross public debt is just 24% of GDP which means there is not a lot of public debt to invest in.

Fifth, Australia's corporate bond market is relatively small as corporates have traditionally relied more on either the share market or banks for funding.

And in recent years they have found it easier to issue debt overseas.

In fact, this is a key issue because if there were a more developed corporate bond market it would provide a higher return alternative to government bonds with less risk than shares.

Sixth, reflecting its high immigration level and fertility rate, Australia has a relatively younger population compared to Japan and many European countries so as a result a greater proportion of pension fund members can afford to take the longer term horizon necessary for investing in shares.

As such it stands to reason that the share allocation might be higher in a country like Australia, than in Germany or Japan.

Finally, and perhaps partly reflecting this, the "default" superannuation system has tended to focus on funds with a 70% allocation to growth assets (including shares & property) and a 30% allocation to defensive assets (mainly bonds and cash).

Some superannuation funds adopt a "life stages" approach where the proportion of shares declines with a member's age.

However, many members may not be in a system that offers this so unless they receive advice they may not adjust their growth allocation as they get older.

What is appropriate?

There is no easy answer as to what is the appropriate allocation to shares and bonds.

It is essentially a function of age (the younger the member the higher the share allocation), wealth (the wealthier the investor the easier it is for an investor to accept short term volatility so a higher share allocation may be appropriate) and risk tolerance.

There is a case for superannuation funds to invest more in alternatives such as property, infrastructure, private equity and distressed assets to provide some diversification away from a reliance on shares, though it should be recognised that such assets are usually less liquid and more expensive to manage limiting any asset allocation to them.

There is also a strong argument that a better way to manage superannuation funds, particularly for older members with less wealth, may be with a particular return (or outcome) objective over time rather than to manage to a particular benchmark allocation to shares, bonds and other assets.

For these reasons, along with My Super reforms, the ageing population and hopefully the growth of the local corporate bond market it is likely that over time the share allocation in the Australian superannuation system may decline a bit.

Not the best timing

The trouble is that now is not a good time to undertake a structural reweighting from shares towards bonds.

Shares have already had several tough years resulting in very poor returns.

And at the same time bond yields have plunged to record lows in the US and generational lows in Australia and elsewhere, making it very hard for the strong bond returns of recent years to be repeated. The next chart shows the gap between the grossed up for franking credits dividend yield on Australian shares (i.e. the annual cash flow investors receive from shares) relative to the 10 year bond yield.

The dividend yield has risen relative to the bond yield, with the latter pushing down to levels not seen since the early 1950s.

Putting the period through the GFC aside, the gap between the two has blown out to levels not seen since the 1950s when the post war share boom was getting underway.

With shares offering a 2% or so yield pick up over bonds, they only require modest capital growth to deliver a total return sufficient to cover the extra risk of owning them.

At the same time the generational lows in bond yields are at risk of reversing resulting in capital losses for investors in bonds.

Corporate bonds offer a higher yield than government bonds, but their yield advantage over shares has receded.

For example, since 1997 A rated Australian corporate debt has yielded an average 1.4% pa more than shares, but yields are now in line.

See the next chart.

While average share market returns are likely to remain constrained in the years ahead, relatively low bond yields are likely to mean that shares will provide a relatively better return on a 5 to ten year horizon.

So now may not be the best time to undertake a structural shift away from shares to bonds in superannuation funds.