intro: correlation tables intro: correlation tables Current month

 

diversity

With so many sectors to choose from most investors cannot include funds from every sector in their portfolio (even if this was desirable). However it is a good idea to build a diversified portfolio, i.e. select funds from different sectors so that all the funds do not respond in exactly the same way to economic and market forces. If two sectors can be shown to act independently then either sector can be included or not on its own merits. If two sectors respond to events in the same way they can both be considered part of a larger sector grouping. Correlation Tables give a measure of how every sector behaves relative to other sectors.

correlation

Correlation is a simple statistical method for deciding the similarity between two series of numbers. For our purposes the series of numbers needed are the monthly average performance values for each sector.

meaning

Correlating the performance values of two sectors results in a single number that can vary between -1 and +1, or -100% and 100% in percentage terms. A correlation result of 100% means there is a relationship between the past performances of the two sectors. A correlation of 0% means the two sectors are acting independently; any similarity over a single month is just random chance.

The general case of a correlation result between 0% and 100% may have a number of causes. For a correlation of 50%, say, it is simplest to assume that half the economic and market forces acting on both sectors are the same and half are different. A negative correlation means the two sectors are being driven in opposite directions by the same forces.

the table

Graphic Investor's correlation table is constructed by taking the monthly average performance values for each sector and then correlating each sector with every other sector. A small section of the table is shown below.

Example

The example shows some of the results for two sectors, 'Far East Excluding Japan' and 'Global Emerging Markets'. The shortened names for the two sectors are shown in the column on the left. The shortened names for the sectors they have been compared to are shown in the columns on the right.

The sectors being compared are arranged in decreasing correlation value, with the highest correlating sector on the left. The correlation percentage result is shown in red. For example, the 'Far East Excluding Japan' sector has its highest correlation with 'FarE IncJ' ('Far East Including Japan') with a correlation of 96%.

groups

The correlation table shows that many sectors have high correlations with other sectors. By arranging the table with related sectors listed together it becomes clear that sectors fall into natural groups. The Unit Trust groups are:

Japan and Far East Group N. America, UK and Global Equity Group
Europe, Small Cos. and Specialist Group UK Gilt and Corporate Bond Group
Remaining Individual Sectors   

In the correlation table all the sectors from the same group are given the same background colour.

funds

Individual funds may vary a great deal from the sector average. (See the 'Fund Performance by Sector' charts from the drop-down menus on the left). Nevertheless the relationships between sectors is one of the first factors to consider when building a portfolio.

relationships

A high correlation between two sectors means there is a 'linear' relationship between them. To explore this further follow the 'Linear Relationships' links on the left.