Many investors view cash positions in a unit trust as a drag on performance. Asset consultants typically advise investors to minimise cash holdings in multi-asset mandates whenever possible, citing the low long-term average returns achieved by cash, relative to the other asset classes.

We have at times had significant cash holdings, even in our higher-risk multi-asset funds. This is not the result of a top-down macro forecast, but due to a scarcity of investment opportunities that meet our criteria in the riskier asset classes. We are willing to hold cash because it effectively expands our opportunity set from risk assets at today’s prices, to the prices that will become available on risk assets in the future. Viewed like this, cash plays a crucial long-term role in our portfolios.

Conventional wisdom: minimise cash to maximise returns

It is clear cash has the lowest long-term average returns of all the asset classes. The table shows that over the past 16 years, South African equities have provided an average 6% per annum additional performance compared with cash.

The simplistic response is to conclude that any mandate with a long-term investment horizon, should aim to minimise its cash holdings and have a  larger holding in risk assets (in this example equities and bonds).

In fact, a typical asset allocation approach is to use the long-term average returns for each asset class and an estimate of risk based on the historical volatility of return. Holdings in each asset class are then optimised to produce the highest risk-adjusted return for the portfolio. This process – which is widely used to determine asset allocation in multi-asset mandates – is almost guaranteed to restrict cash holdings to the minimum.


Source: Credit Suisse Global Investment Returns Yearbook 2017

A key assumption underlying the return-optimised asset allocation process, is that the future returns for each asset class will be similar to their historic long-term averages. This assumption is, however, deeply flawed.

It shows little regard for the importance of entry multiples or yields in determining long-term returns, particularly for equities. Future long-term returns can differ substantially from long-term average returns.

We illustrate this using FTSE/JSE All Share Index data from 1995 onwards. We compare the rating of the index initially with the subsequent five-year total returns. The returns are expressed relative to the average for the whole period.

FTSE/JSE All Share Index, five-year subsequent total returns relative to the average for the whole period, September 1995 to February 2012, rolled monthly.

The rating was determined using the price-to-sales ratio. This ratio is less distorted by the effects of loss-making companies – and the margin fluctuations of the more cyclical index components – than a price-to-earnings ratio.


The table above shows that if we invest in equities at a time when they are relatively expensive (a price-to-sales ratio greater than 1.6x), returns are unlikely to significantly exceed those achieved by cash and will probably not compensate investors for the additional risk incurred. By contrast, investments in equities when the index is relatively cheap (a price-to-sales ratio of less than 1.2x), are likely to deliver much stronger returns than the average.

 In periods when we find fewer shares that are available at a price that provides a margin of safety, we far prefer to accumulate cash rather than overpay.

Historically, there is a strong link between the prevailing valuation and subsequent long-term equity returns that the asset allocation optimisation model ignores.

Thus we believe there is a serious flaw in the asset allocation processes used by many managers of  multi-asset mandates.

Rather than using top-down or macro-economic forecasts to determine positioning, we advise bottom-up, or company-specific research.

Our equity team searches for companies that meet our ‘3 Ms’ criteria:

  • a sound economic moat;
  • management with a proven track record and an alignment of interests;
  • and, most importantly, the stock must trade at a sufficient discount to our estimate of its intrinsic value to provide a margin of safety for our investment.

While it is relatively easy to find shares with attractive economic characteristics and sound management, they are seldom available at reasonable prices. Thus the requirement to invest with a margin of safety can often restrict the opportunity set.

In periods when we find fewer shares that are available at a price that provides a margin of safety, we far prefer to accumulate cash rather than overpay.

Hence our asset allocation is effectively driven by our bottom-up stock selection process, in contrast to the majority of our peers who typically are bound by a tight target range.

The crucial role of cash  

A cornerstone of our investment philosophy is having the patience and discipline to wait for attractive prices before investing. This is often not without a temporary cost, as peer group funds with higher equity weightings may outperform in the short term.

However, cash plays a crucial long-term role in our portfolios, in that it effectively expands our opportunity set from the prices available on risk assets today, to the prices that will become available in the future.

By providing this intertemporal choice, cash in fact becomes a very valuable holding, enabling us to take full advantage of future periods of investor pessimism (or even panic). This would not be possible with an investment process that prevented substantial cash accumulation, by tightly restricting us to predetermined asset allocations.

Article credit: Kevin Cousins, fund manager at PSG Asset Management ;