Portfolio planning
In this section of PWM we test the performance and volatility of two investment strategies using model portfolios. Each month we look at two baskets of shares – one global and one European.
GLOBAL PORTFOLIO
Who would not like to own the five best possible stocks while being short the worst possible five stocks and by remaining market neutral through an equal long and short market exposure? The strategy presentedis a modified version of an idea that worked exceptionally well until the market recovery of March 2003.
The long portfolio of five worldwide high capitalisation stocks is based on a very simple version of an optimised return and volatility strategy.
The long portfolio contains the best runners, which are all concentrated on US stocks. This is a bit surprising when considering the excellent comparative performance of the emerging equity markets.
The short portfolio of five stocks takes into account a very complex “bad boy” strategy that searches for the worst possible stocks by using six different fundamental and quantitative measures that make up a bad stock to own.
The short portfolio shows a few surprises, such as Microsoft and Suez.
The valuation must be very high and should be much higher than that of the average stock. The stock should capture the most possible of down moves, and comparative and absolute returns must be low.
The long portfolio should be able to easily beat the benchmark (MSCI World Index) while the short portfolio needs a more quiet market environment than the one seen over the last six months to really falter.
Chart 1: As seen in the line graph, the long portfolio (blue line) has done consistently better than the short portfolio (red line) or the MSCI World Index $ (pink line) since spring 2001.
Chart 2: The scatter diagram shows the excellent risk return profile of the long portfolio (blue) against the short portfolio (red) and the MSCI World Index (pink)
Chart 3: The long portfolio is concentrated in the cyclical sectors such as Consumer Cyclicals and Technology. The short portfolio does not contain any Consumer Cyclical but shows exposure in the Utilities sector.
EUROPEAN PORTFOLIO
Benchmarking and simultaneously beating equity indices might appear to be contradictory as the index tracking process does not leave a lot of room for discretion. Are there really any quantitative factors that help gain an advantage in the benchmarking process? There appears to be an element that works well consistently: market capitalisation. The primary objective of this strategy is to remain benchmarked as closely as possible while trying to achieve an advantage against the benchmark when the market situation allows. The result of the benchmark should be respected as a worst-case scenario.
Different ways of benchmarking were tested over the last three years which did not achieve the desired consistently favourable results. The methods tested were a performance oriented benchmarking and a combination of market capitalisation and performance weighting. Quantitative approaches through the optimisation of return and volatility were also tried out. The results look interesting with exciting performances over a few quarters.
The big problem is the occasional negative quarters that take the portfolio too far astray from the main objective – simple benchmarking.
Chart 1: The benchmarking module has taken into account the last 12 months of price history for the calculation of the optimal benchmarked portfolio. The maximum of four stars (see inset) was obtained for three different quantitative measures while the return error has shown the lowest quality.
Chart 2 (at right) illustrates the correlation between the 10-stock benchmark portfolio and the reference index which has started stabilising at the beginning of 2002.
Chart 3: The cake to the left hand side shows the sector composition of the portfolio that should beat the benchmark. Less significant market segments were excluded by giving more weight to sectors such as Financials, Utilities and above all Technology (see sector weights).
For further information on Brainpower’s professional portfolio analysis software, please visit www.brainpowerweb.com or contact
Alan Parmenter on +44 (0) 20 7392 7108