C. Beta

As mentioned above, the determination of the systematic risk of risk assets quoted on the stock exchange is based on the CAPM methodology. Systematic risk corresponds to the general risk of the market, representing the risk related to all aspects (e.g. political, economic, etc.) which can impact the behaviour of investors. This risk is different from the individual risk of each of the listed securities, being a non-diversifiable risk of the market considered as a whole.

The share’s systematic risk is defined by calculating its beta, which in the context of defining the company’s cost of capital corresponds to the equity beta.

Two alternative methodologies would be valid for calculating the beta: i) to estimate the beta for PTC based on a benchmark of companies with similar activities or, ii) to seek to infer the beta for PTC on the basis of the beta of PT SGPS.

As referred in the determination of 2010, as regards the second alternative, the calculation of the beta of each activity of PT SGPS would be required, with a careful economic analysis of each one in the different geographies in which they operate, an exercise which, in the case of operations not covered by the stock market (for example TMN’s activity) would inevitably call for the use of benchmarking.

Subsequently, it would be necessary to assess activities mentioned above in order to weigh their impact on the beta of PT SGPS. Most assessments carried out by analysts, based on the most common methodologies, such as Discounted Cash Flow and the Enterprise value/EBITDA ratio, showed that results can vary depending on the assumptions used.

Accordingly, given the inherent discretion embedded in the definition of the market value of a company’s activities, the alternative which could be considered would consist of the evaluation of each activity of Grupo PT, at book value, which however could have several limitations arising, for example, from the fact that assets have different maturities and are often shared by different activities.

Accordingly, the benchmark was deemed to be the most plausible and consistent alternative for the definition of the beta, as laid down in the determination of 2010.  As regards the benchmark methodology,  the use of comparable companies is proposed, given that, although regulatory precedents are deemed to be a good reference (see table 6), they may distort final results, given that in most cases, the methodology used by regulators is also based on a  benchmark.

Source: Website of the respective regulators and BT report

As far as the benchmark of comparable companies is concerned, the methodology for calculating the beta, as in the determination of 2010, has the following underlying characteristics:

- Use of the Harris and Pringle model¹ to calculate the equity betas of comparable companies. The formula allows the asset’s beta to be calculated, i.e., the beta without the effect of capital structure, later leveraged with the capital structure defined as optimal for PTC;

-  Frequency of observations: beta can be estimated through observations made on a daily, weekly, monthly or quarterly basis. The number of observations matters because it contributes to the reduction of estimate uncertainty. In fact, just as with the risk-free interest rate, the option for using monthly observations was taken;

- Period of time: taking short series into consideration may distort results and remove relevant information. In fact, the most recent observations contain effects which may not represent future expectations appropriately, thus series of a period long enough to correct the volatile effects that can be felt in the short term should be used. The beta shows fluctuations over the business cycles of the company and it is noted that, in fact, PT has been undergoing since 1995 significant changes with respect to its activities’ structure, with the liberalization of the sector, the internationalization of the group, technological innovation and the diversification of services provided. In this context, it is considered that the series’ period should incorporate relevant observations to ensure that the result is robust and representative of the risks inherent in the company’s current structure. European regulators clearly seem to prefer 5-year periods, which allow a high level of robustness and certainty of results obtained;

- Beta data are taken from Bloomberg and correspond to the values resulting from the Bayes formula, that is, the adjusted Beta². This adjustment provides a more robust estimate and less volatile fluctuations.

Methodology to be applied from 2012
 
In the light of points listed above, the beta must be calculated using the benchmark of comparable companies.
 
Data correspond to adjusted betas of comparable companies, provided by Bloomberg - historical series for the five years preceding the decision year, with monthly observations. Moreover, data from Bloomberg must be unleveraged from the financial structure and later leveraged, using the Harris and Pringle model (equity β = asset β * (1+ D/E)), using the gearing defined for PTC for the decision year.