In developing its value and growth indices, Frank Russell has adopted a methodology with a long-established pedigree, albeit one that is not without its critics.

Russell has emphasised that its indices are meant to help pension funds and managers understand the market better, while accepting some purist style managers could take issue with their methods.

But Russell's methodology does receive implicit support from MSCI which plans to use the same method for calculating the 22 developed markets and 26 emerging style and growth indices covering Europe, Australia, Asia, the Far East and Japan that it plans to launch in the autumn.

MSCI describes the price to book ratio method as the only secure measure, adding that as far as possible it will attempt to divide the index 50/50 between value and growth.

Some managers and consultants, however, strongly back alternative systems, for example drawing conclusions about style characteristics from a survey of unit trusts or from leading portfolios.

Much of the argument concerns the growth component and whether it can be defined as those shares in an index that are not value".

David Morris, an associate at John Morrell & Associates assessing the Russell methodology, which he acknowledges is the standard practice, says: "Ranking a list of stocks on an analytical measure like price to book and/or yield and calling the top segment 'value' is sensible but I have a quarrel with people who call the bottom half growth, because both measures have got price in them. You are not any smarter for having the second index."

In the US such indices are used in the "least squares regression technique" to define style attributes of portfolios. The process involves taking the monthly returns of a portfolio and comparing them to value, growth, small cap, large cap, cash and bond indices and calculating coefficients that determine what proportion of returns are due to each style in a given fund thus labelling the portfolio.

Morris says: "We reject that fully and we believe that the fund management industry rejects it too. "In our holdings-based system, we look at each style element in the portfolio through time and we measure its attributes based on the stocks held. We can take a snap-shot at any single point in time and look at the style biases making a composite of a manager's strategy."

This criticism of definitions is added to by Karl Heinz Thielmann, senior fund manager at Deutscher Investment Trust (DIT). DIT defines growth stocks as companies expecting to earn 8% per annum over the next five years. Value, he says, "is a bit more complicated".

"We found that in Europe there were no universal value criteria, so you have to look for value companies by sector and country and we combine this with a scoring method."

He also takes issue with the price to book method. The only sensible way to construct these indices, he says, is to use an index of leading funds for both value and growth, though he concedes that such a system could not be employed in Europe as yet.

However John Stannard, Russell's director of data services, sets out a case for indices on the basis of their utility value: "This is a way of identifying trends. An informed investor can use the information this provides to make better judgements."

He concedes that some managers may have fears about being type-cast but he does not think these will be borne out, believing that work of this nature will simply help the market to evolve as it has done elsewhere.

"It's important to evaluate the construction methodology in terms of how the index will be used. The UK style indices are intended to provide insights into general trends in growth and value stocks. Definitions of growth vary widely, based on individual beliefs, and a single index can't be a unique benchmark for every growth strategy."

However, Dutch pension fund ABP believes that indices are a valuable tool for managers. Assessing the difficulties of constructing such indices, Robbert Coomans, who manages all ABP's external equity managers, adds: "I think that for value it is very easy to get a decent number of ratios. On the growth side it is a little bit more difficult. One possibility is that you will determine what is value and then say everything else is growth or you determine value, with growth based on excess earnings growth, leaving a grey area." John Lappin"