ANALYSIS | How much you should invest overseas right now? | Business

If you believe that South Africa will denigrate into another Zimbabwe, then the choice is simple – move as much capital offshore as you possibly can.

However, if we believe that our great country will, as it has in the past, withstand the many economic and political challenges that we’ve faced, with the determination we’ve historically portrayed and with the strength of our financial institutions and legal system, then the pressure to go to the maximum offshore limit (in equities) is less obvious.

To find out what the most optimal allocation would’ve been in the past, we conducted a detailed analysis of data over a 50-year period. We used a combination of the FTSE/JSE All Share Index (Alsi) and the MSCI World Index – limiting the choice between these two indices as potential investment options for investors for illustrative purposes. Clearly investors have wider choice, but most equity benchmarks that are allowed to allocate offshore would have these two indices as part of their benchmarks in differing proportions. The findings of our analysis are included below.

The first point to note is that the Alsi already has large rand hedge and rand leverage components to its market capitalisation.

Rand hedge companies generate the majority of their revenue and profits in global markets (e.g. ABInbev, Richemont, British American Tobacco). Rand leveraged companies have rand input costs and (mainly) dollar-based revenue (e.g. SA gold, platinum and general resource companies).

According to our classifications and research, currently 61.5% of the Alsi is classified as either rand hedge (32.2%) or rand leveraged (29.3%) companies. As shown below, South African investors in the Alsi already have a significant level of offshore diversification.

A large portion of the Alsi made up of rand hedge or rand leveraged stocks:

Over the last 50 years, investors have not had a dramatically different return vs risk experience regardless of whether they were invested in the Alsi or the MSCI World.

This is demonstrated by the statistics in the table. If you were 100% invested in the Alsi, your annualised return of 16.5% would’ve been better than that of the MSCI World return (measured in rand) – despite a 7% annualised depreciation of the rand. Your maximum annual drawdown would also have been lower. However, you would’ve accepted slightly higher levels of volatility in returns. The Sharpe ratios (which show performance in excess of the risk-free rate relative to risk) are similar.

There are obvious benefits to having an offshore equity allocation in a portfolio.

Returns are not the only consideration in an optimal portfolio. Diversification and correlations of returns matter too.

The benefits of offshore diversification are:

• Improved potential annualised return and risk (volatility and maximum annualised drawdown) outcomes, as the two asset classes are not perfectly correlated. This means that there will be times where South African companies perform well when global companies don’t, and vice versa.
• An additional hedge against substantial weakness in the rand.
• Exposure to global developed and developing markets. This gives investors greater choice and frankly, protects them from an uncertain and potentially weak South African economy.

As you’ll see below, adding global equity exposure can improve risk adjusted returns and, hence, Sharpe ratios. A Sharpe ratio measures alpha (excess returns over the risk-free rate) relative to the standard deviation (risk) of the excess returns. The higher the ratio, the better – it indicates that investors are getting a higher excess return per unit of volatility of the excess return.

An optimal offshore allocation has shown mixed results over time.

Some of the findings from our research are shown in figure 3 below.

Over short periods, such as the past year, it has been more beneficial to have higher offshore exposure.

It has also been beneficial over the past five, 10 and 15 years. This is clear by the annualised returns and the volatility of returns over the periods, for differing allocations to the Alsi and MSCI World. These periods are highlighted in blue in the table below.

This is less obvious over a three-year horizon and over longer periods, such as 20 and 35 years (highlighted in orange), notwithstanding the depreciation in the rand.

Over a three-year period, a larger weighting in the Alsi would have produced higher performance. Likewise, over the much longer periods of 20 and 35 years, annualised returns would have been higher (although volatility would’ve been slightly higher too).

Interestingly, over a three, 20- and 35-year time horizon, Sharpe ratios have not been too dissimilar, regardless of local vs. global equity allocation.

Annualised performance, volatility and Sharpe ratio comparison (in rand):

Expanding on the long-term statistics shown in figure 2 (the 50-year data), and running 10 000 random combinations, the combination of the Alsi and MSCI World that maximised the Sharpe ratio (point of tangency of the black capital allocation line to the efficient frontier curve below) was at an allocation of 40% to the Alsi and 60% to the MSCI World Index. This would have resulted in an annualised return of 16.3% with a standard deviation of 15.4%.

The annualised returns and standard deviation at the maximum allowable allocation of 45% offshore are not that different. The annualised return of the portfolio would’ve been slightly higher at 16.4%, but also with a slightly higher standard deviation of 15.9%.

Academics would argue that any combination of the two asset classes that lie on the efficient frontier would be ‘optimal’. In other words, at any point the highest expected return for the defined level of risk (as defined by the standard deviation of returns) or the lowest risk for a given level of return would be optimal. To put it differently, if history repeated itself, investors with a desire to ‘maximise’ returns by increasing their exposure to the Alsi, should be comfortable to take on the increased level of risk.

The data differs over the various periods (as highlighted in the tables earlier in the article). In particular, over the past 15 years investors would’ve been substantially better off with the maximum offshore allocation. This is shown in the chart below.

However, long-term returns are dependent on the price you pay.

The perceived risk of investing in South Africa, relative to global markets is partly reflected in the PE ratios that investors pay. On average, the PE ratio of the Alsi is lower than the MSCI World. The lower the PE ratio, the higher the subsequent returns you should expect over time.

This risk is also embedded in the higher risk-free rate (10-year bonds) and the implied equity risk premium for SA equities relative to the MSCI World (this is depicted in the chart below of the SA 10-year bonds relative to the US 10-year bond (used as a proxy for the MSCI World).

This is one of the main reasons why the long-term annualised returns for SA equities are not too different from the annualised returns for the MSCI World Index. The risks are largely embedded in the higher risk-free rate and the implied equity risk premium.

Long-term returns are dependent on the initial price you pay for assets. The lower the PE ratio, the higher the returns and vice versa. This is illustrated by the data below, showing the subsequent five-year annualised returns from the Alsi based on the PE ratio ‘paid’.

The current PE ratio of the Alsi is 9.6x. The scatter chart below shows the five-year subsequent annualised total return depending on the PE ratio at which you would’ve entered the market. The lower the PE ratio was when investing in the Alsi, the higher the probability of improved annualised returns. Over the period since January 1960, the probability of generating exceptional returns when paying less than a 10x PE multiple for the Alsi was extremely high. The current PE ratio of the Alsi (at 9.6x) is less than half the PE ratio of the MSCI World Index (at 20.3x).

Assuming the current low valuation of the Alsi relative to the MSCI World, as well as the extent of the recent weakness in the rand relative to most developed market currencies, if you were considering increasing your offshore exposure, now might not be the time to do it. 

In summary, the optimal offshore equity allocation depends on the investor’s time horizon as well as the investor’s desire to manage risk.

When we refer to risk (in the context of this article) we mean volatility, maximum drawdowns as well as political and country risk.

Over a three-year period, annualised returns from holding a larger weighting in the Alsi have been more beneficial. Likewise, over much longer periods of 20 and 35 years where annualised returns have been higher (although your volatility does increase slightly).

Over most other periods, it has clearly been more beneficial to have a higher offshore exposure in a portfolio.

The diversification benefits can allow investors to improve their annualised returns, volatility and maximum drawdowns. This is especially evident over the past 15 years (given the poor SA macro and political backdrop) where it has been more beneficial for investors to have maximised their global allocation.

As shown in figure 4, over the full 50-year analysis, the equity portfolio (assuming our two illustrative equity portfolio options) that maximised return per unit of risk (the point of tangency on the efficient frontier) was at an allocation of 40% to the Alsi and 60% to the MSCI World Index. This is at the 16.3% annualised return with a standard deviation of 15.4%. The annualised returns and standard deviation at the maximum allowable allocation of 45% offshore were not that different. At an allocation of 55% Alsi and 45% MSCI World, the annualised return of the portfolio would’ve been slightly higher at 16.4%, but also with a slightly higher standard deviation of 15.9%.

Price is what you pay, value is what you get.

The million-dollar question is how this will play out in the future. We don’t have the answers. However, we do know that longer-term returns are highly dependent on the price you pay for assets at the start. You’re currently paying a lot less for the Alsi than the MSCI World (using the PE ratio as a proxy for value), implying that a lot of macro and political risk is already discounted in the South African index. This has also been priced into the weak rand.

Having some level of flexibility in your SA vs global equity allocation makes sense – depending on your assessment of the relative value of assets, but this does bring in timing risks.

Notwithstanding this, although I may not advocate going to the maximum 45% offshore allowable limit right now, I would certainly have a large percentage of my assets offshore. If you think South Africa will deteriorate into a pariah state, then the answer is simple. But local is still ‘lekker’, so don’t write off SA just yet!