Given the rigour and hard work required in an evaluation of intangible assets conducted by the company itself, which has full access to all relevant information, what should we make of the annual ‘hit parade’ charts which appear in the economics press, giving new values for the top worldwide brands? Why such big differences between valuations?
The Interbrand research company, which is overwhelmingly the main producer of such data, has used two methods over time. Historically, it has attempted to derive values for brand EVA from public information in the annual reports of stock-exchange-listed companies and a variety of other public sources. Not being able to work with a business plan, given the confidentiality of company plans, Interbrand instead analysed data from the last two years.
So how does it make the leap from EVA to brand value? It used an estimation of the share of EVA attributable to the brand, multiplied by a figure (the ‘multiple’), itself derived from a statistical model based on the analysis of the price/earnings ratio (p/e) for stock-exchangelisted companies such as Gillette. The price/earnings ratio is actually a multiple itself. It compares the stock value with the profits associated with that stock: this will indicate, say, that a stock is worth 10 times its dividend price.
Interbrand configured its statistical model using stock-exchange-listed companies. Knowing the multiple (p/e) for each company, it performed a strategic analysis of its brands, following a method similar to the one we have described for our strategic audit of the brand. The end result of Interbrand’s strategic evaluation of the brand is an overall score for the brand, measuring the strength of the brand (the ‘brand strength index’).
This is the sum of the partial scores obtained from each of the individual audit criteria. The criteria are leadership, stability and so on. It is then easy to identify the statistical relationship between the recalculated strength of the brands and the virtual multiple approximated by the price/earnings ratio (p/e) on the stock exchange. This statistical relationship has never been published, but has been represented as shown in Figure below.
Having produced an external estimate of the EVA for each brand, it was then easy for Interbrand to calculate the brand strength index which, when factored into the statistical model, identifies the virtual multiple. All that remained at this point was to measure this virtual multiple as the share of estimated EVA allocated to the brand.
Several remarks can be made about this external procedure, which is used to produce the published ‘league tables’ of global brand value. The tables are based on this logic, except that they are not in possession of all of the relevant information (as opposed to, say, an auditor appointed by the company to value its brands).
They are thus obliged to obtain an external estimation based on the accounts published by stock-exchange-listed companies, and the figures are subject to a wide margin of error. Furthermore, these league tables cannot measure the value of brands belonging to family-run companies such as Mars, Levi’s and Lacoste, which do not release public figures.
Nor can they include brands belonging to companies producing consolidated accounts that are not broken down by brand. Lastly, they exclude cases in which sales may be attributable to factors other than pure demand. Consider air transport, for example, where the policy of alliances means that it is possible to end up flying with Delta Airlines after having bought an Air France ticket. Also, a significant part of demand is influenced by exit barriers such as frequent flyer cards: this is not pure demand driven by customer preference.
Other critical remarks may be made about this approach, as we have already seen, including sensitivity to variations in the multiple, and the validity of the graph. Recently, Interbrand has changed its method of producing its ‘global brand value’ league tables, moving towards a more conventional financial and economic approach. Although its methodology has not been explicitly published, reference has been made to ‘net present value of future brand earnings’, which would be more in line with our recommended nine-step process.