Q: How would a scientist build an investment portfolio?
A: Equally.
The CSIRO is Australia’s national science agency and is one of the largest and most diverse research agencies in the world. So diverse that it is now directing resources to investment portfolio construction.
The CSIRO-Monash Superannuation Research Cluster is researching many aspects of our superannuation system, including investment practices. A recent paper concluded that equal weighting is the “highest performing” structure for a portfolio, better than market capitalisation and better than alternatives such as RAFI’s ‘fundamental indexation’.
Put simply, an equally weighted index aims to allocate the same amount to each company within the index regardless of size. An equally weighted Australian index, for example, would have as much CBA in it, as it did Harvey Norman.
The paper tests US data for the years 1962 to 2009. US data is used because this is the largest reliable data set available.
The results are that investing $1 in 1962 would grow to:
Sophisticated models developed by such finance luminaries as Nobel Prize winning economist Eugene Fama and Robert Merton were used to determine the source of this outperformance.
The conclusion is that equally weighting a portfolio outperforms market capitalisation because of three factors:
What the paper means by market timing is that equal weighting extracts more return when markets are rising and loses less when markets are falling.
Equal weighting was found to outperform fundamental indexation because of the better market timing. This is contrary to the claims made by the proponents of fundamental indexing.
Fundamental indexing is promoted as being superior to market capitalisation on the basis that it uses better measures of the size of a company. Market capitalisation is based on traded prices so is distorted when the market misprices. Fundamental indexing uses data such as book value, revenue, cash flow, dividends, sales and employee numbers to get a less market-distorted quantifiable size-based weighting.
The resulting index is said therefore said to be less susceptible to market sentiment and extreme market phases such as bubbles. The CSIRO have shown equal weighting achieves this but that fundamental indexation does not.
In non-scientific terms, the problem with investing more in bigger companies and less in smaller companies is twofold. When the market overvalues a stock, you buy too much of the overpriced stock. When the market undervalues a stock, you sell too much of the underpriced stock.