reducing volatility by diversification is not always the best step for investors who are looking to make good long-term returns
an diversification ever be bad? I believe that it not only can, but indeed often is for many a private investor's portfolio. Diversification is supposed to reduce risk, but there are situations where it merely lowers returns. Here are the reasons, but first a philosophical point.
Many equate risk with volatility, but that just shows how even the language of investment has been subverted by our industry's sales-driven culture. Risk means risk of loss, volatility is merely one mathematically definable formulation, and one with its own pernicious pre-conception, namely that variability in performance equates with tracking error, a.k.a. divergence from an average.
Some may consider this to be an exercise in semantics and pedantics, but the point is that divergence from the average is the tool with which investors strive to maximise performance, whereas approximation to the average is the tool with which fund managers strive to retain their funds and jobs.
That item of pedantry cost investors across the world tens of billions of dollars in capital destruction. To be precise, the loss was $12,245bn between the Millennium and the end of 2002, according to figures supplied by FIBV, the trade association of world stock exchanges.
An easy mistake
Furthermore, volatility analysis is dangerously seductive, for volatility varies with the state of the market. Typically in the early stages of bull markets capital gains are accompanied by low volatility. As the market matures, so volatility increases while capital gains become harder to achieve. As the bear market deteriorates so volatility increases even further. Indeed volatility is such a variable factor itself, that there are derivative contracts for it in Chicago and Frankfurt.
Even if one talks that language, high volatility can be good news. Look at table 1 below. Here is an example where a global growth portfolio is compared with the FTSE World Index converted into sterling, over the past six years. Over the entire period, the standard deviation of monthly returns is similar. However when this is segregated into months of positive and negative returns, volatility is higher for the growth portfolio in the positive months but correspondingly lower in negative months. This example is not theoretical. It represents the actual results of a pension fund.
The explanation lies in the declining returns from increasing diversification. The reduction in volatility by increasing the number of holdings from one to two may be huge. Further reductions in volatility are increasingly hard to achieve as the number of holdings increases to 10, 50 or more. Theoretically, this can be shown in the following chart.
Once again it would be wrong for individual investors to ape fund managers. If their focus funds find that 50 is the ideal number of holdings, individuals may find that a much smaller number of direct investments is appropriate.
On the one hand that is because of the beneficial effects of size on costs for individual investors. Commission rates initially fall sharply as the size of transaction rises. For instance, Charles Schwab charges a flat commission of $20 for the first 1,000 shares, with the rate rising $0.015 for every additional share. Where annual administration charges are levied, these are typically related to the number of holdings.
On the other hand that is also because of the negative effects of size on liquidity for institutional investors. Disposing of a position may take several trading days of dribbling out small portions or accepting unsolicited offers in the market. In my experience as a stock-broker, it was not possible to do more than about a quarter of a third of daily trading in a share for any extended period, without the market realising that a tap was on. For the largest institutional investors, that problem applies even if they merely want to reduce the size of a holding.
Psychological factors also come into play. Too few holdings and one risks falling in love with one's favourite stock. As a rookie registered representative, when putting forward lists of stocks, I would often be asked which was my favourite. Foolishly I met the client's expressed need, but did them a dis-service, because inevitably that subsequently turned out to be the idea that was too good to be true.
Equally, too many holdings and one risks losing track of one's investments. Every holding costs time in tracking company announcements and industry trends. As an experienced institutional salesman I therefore learnt to focus on just one stock in each sector.
For those who prefer funds to direct investment, the case for concentration is even more compelling, because each fund itself represents a diversified portfolio of holdings within its mandate. The number of funds that are required may be as low as four. While that may seem unduly risky to many advisers, it is in fact the actual number of funds held in the portfolio whose volatility figures are given in the example above.
The rationale is simple. A typical family of funds might contain 20 individual funds, between which investors can switch at little or no cost. Of these, one might judge that 10 are fundamentally unattractive for a variety of reasons. Among the rest, one might find that five are boring. That leaves five funds worth purchasing. If an equal amount is invested in each, the risk of falling in love with any one of them can be reduced.
There is, however, a very big caveat to such a concentration of investments and that is correlation. For example, an investor splitting his portfolio 50/50 between a UK and a US equity fund reduces his risk very little, because the UK is the most international of markets and the US is the largest.
The fourth dimension
In an increasingly global market place, what happens on one side of the world increasingly affects what happens of the other side. Falling trade barriers have been a trend of the global economy, ever since the European Coal and Steel Community was founded by the Treaty of Paris in 1951. Flexible exchange rates have been increasingly common since the Bretton Woods Agreement collapsed in 1971. Alternatives to the capitalist economic model have become unfashionable since the collapse of the Soviet Empire in 1989.
Among equities, it is therefore more useful now to think in terms of four-dimensional diversification, specifically geography, size of company, investing style and economic sector. Table 2 illustrates this point by showing the difference in performance measured in US$ between the best and worst in each category in the five years since the Millennium.
While picking the right region made a big difference, picking the right size of company made a bigger difference and picking the right investment style made an even bigger difference. However by far the biggest difference of all was produced by picking the right economic sector. 100 represents perfect correlation of 100%, while negative figures represent inverse relationships.
To help investors diversify effectively, see the large table entitled Correlation Matrix, which shows the principle asset classes against each other, based on monthly returns I have calculated over the past two decades.
As can be seen from the correlation table, there is little point in broadening a growth portfolio by adding a technology fund. The same applies to a value portfolio adding a fund of financial shares. Equally an emerging markets portfolio gains little from adding an Asia ex-Japan fund, as there are extensive overlaps also between them.
To conclude, resorting to a mathematical definition of volatility may appear to be a simple solution. However the best way of reducing risk is by diversifying into a small selection of assets judged most likely to maximise returns. Making those judgements well requires dedication and courage. Claims that risk can be easily contained by reducing volatility may be little more than cowardice wrapped in jargon that guarantees nothing other than mediocrity.
Our weekly heads-up for advisers
'Nothing can prevent scammers developing workarounds'
Stalwart Scottish Mortgage takes third place
Consistency and compliance vs. slower reaction time
Search for replacement to begin imminently