Last month, I wrote that I was happy with my 18% international exposure in my SMSF as a stepping-stone to a more substantial holding. Since then, I have built models to help me work out what that optimal exposure should be for me. After discussing the results, I will then comment on how I plan to get there.
In classic tradition, the optimal allocation across asset classes is achieved by combining forecasts of expected returns, volatilities and correlations in an optimiser. The outcome is the best risk-adjusted return portfolio for a given cash rate.
I start by considering only two assets: domestic equities (represented by ASX 200) and US equities (represented by S&P 500) but valued in Australian dollars. In that sense, the international exposure is unhedged. I personally choose to use the iShares ETF, IVV, to make my international investments. Unhedged means that the S&P 500 return – expressed in $US – will be increased when the $A depreciates against the $US and vice versa.
I used the same volatility forecasting method I use in my domestic equity sectoral allocations and share portfolios for both indexes. I show my rolling volatility forecasts for 12-months ahead for both of the ASX 200 and the S&P 500 in Chart 1 to the end of 2014. The essence of the model is that volatility is assumed to go through periods of stability – but this ‘mean’ occasionally shifts to a new regime – and sometimes sharply. Recently, the ASX 200 has been stable and a little less volatile than the US index – but sometimes in the past, the deviations have been quite large. The Woodhall methodology paper under the Market Updates tab on www.woodhall.com.au [1] gives more details for both returns and volatilities forecasts.
Chart 1: 12-month-ahead forecasts of volatility
[2]Source: Thomson Reuters and Woodhall Investment Research
My January 1 estimates for yield from Thomson Reuters data were 2.2% for the S&P 500 and 4.8% for the ASX 200. When franking credits are included for an Aussie SMSF in pension mode, my total returns forecasts, including dividends, franking credits and capital gains, for 2015 were 9.6% for Wall Street and 11.1% for the ASX 200. Since my broker-based forecasts for both are cast in their local currencies, I am implicitly assuming that the $A will not change over the course of the year ahead.
Since the forecast correlation between the two indexes is only 0.21, there is ample scope to gain from diversifying across both indexes. The ASX 200 gets a 70% share and the S&P 500 gets 30% in the best risk-adjusted return portfolio.
The expected portfolio return is a simple weighted average of the two expected returns and is equal to 10.7% in this case. But there are some nice gains to be had in lowering expected volatility. The portfolio is expected to have a volatility of 8.2% against 9.6% for the ASX 200 and 11.8% for the unhedged S&P 500. Note that 8% was the level of volatility over the ‘quiet days’ of 2003 – 2005 on the ASX 200.
If a depreciation of 10% in the $A were to be expected over 2015 – and many analysts expect that or more – the optimal allocation changes to 37% ASX 200 and 63% S&P 500. Currency forecasts are notoriously difficult to make and a no-change forecast is often thought to be about as good as one can do on average. However, from a risk perspective, I assume that a depreciation of the $A against the greenback is more likely than an appreciation. Therefore, I will err towards a higher exposure to international than my no-change calculations of 30%. But I will not go anywhere near the full way to 63%. If, at some point, I feel that the $A might start to appreciate, I would change my exposure to a hedged ETF, such as iShares’ IHVV, as I wouldn’t want the appreciation to erode any gains on the S&P 500. More likely is to keep the aggregate exposure of IVV and IHVV constant but change the balance of the two international ETFs.
In related work, I found that I could not justify a broader international exposure than just to the S&P 500. I found that the Rest of the World attracted a zero weight using my assumptions and reasonable variations from them. So my new target IVV exposure is 30% ‘or more’. At a current 18% IVV exposure, I have a long way to go even to get to 30%! However, data from January 1st 2015 shows me that I didn’t lose much by my tardiness. The S&P 500 has been relatively flat so far for 2015 and the $A has not fallen a lot. The ASX 200 is well ahead over this period and my domestic portfolio is well ahead of the ASX 200. Even had I been losing from not moving quicker, I am in favour of taking measured steps rather than jumping in with both feet.
Next month I will look at the broad sectors of the ASX 200 – and update you on how my internationalisation is going.
Important: This content has been prepared without taking account of the objectives, financial situation or needs of any particular individual. It does not constitute formal advice. Consider the appropriateness of the information in regards to your circumstances.