The analysis of the competitive environment within which firms operate has traditionally been in terms of three features: structure, behaviour and performance (Ferguson et al. 1993). This is illustrated in Figure 78 along with the main parameters affected by this model.
Market Structure
Number and size of firms, product differentiation, ease of
entry etc.
Behaviour
Pricing policy, promotional activity, product differentiation
etc.
Performance
Ability to earn and sustain supernormal profits
Number of firms, FTTH or
DSL
Price, Deployment
strategy
Pre-tax profit
Figure 78, Conceptual model for competition model
Market structures differ on two important dimensions: the number of firms and the nature of product differentiation. After years of homogeneous services offered through a monopoly provider, telecommunications of today is a competitive market with several competing firms providing differentiated services. However, due to the substantial investment
required for building wired telecommunications access network, that market is at best an oligopoly market110.
The behaviour of firms within a given market is described in terms of their power to determine price and the efforts these firms will devote to excluding potential competitors. Tirole (2003) classifies these behaviour according to the speed at which they can be altered. In the short run, price and production quantity are often the main instrument that a firm can change easily. Changes to either of these parameters are called strategic interactions and can be systematically analysed as through the theory of non-cooperative games. On a longer time scale, cost structures and product characteristics can be altered.
The performance of firms within markets is reflected in terms of their profitability and their ability to maintain this profitability in the long run.
In traditional economic theory, under the profit maximisation hypothesis, a company increases its production level as long as each additional unit adds more to total revenue than it does to total cost. However, when analysing broadband deployment the high sunken cost of the infrastructure results in low marginal cost and therefore the most profitable state is always 100%
market penetration / take-up rate.
An objective in economic modelling is determining the equilibrium positions that a model implies. This stems from the overall goal of explaining and predicting changes in market values such as prices and quantities. A prerequisite to finding equilibrium is to have a theory that explains how an initial condition arises and how a new condition arises when circumstances change111. The challenge for this project was to develop a competition model that can relate the behaviour of firms (under a fixed market size) to their market share, and ultimately the profit.
110 Market condition in which a small number of firms controls the total supply of a given commodity or service.
111 Liebowitz and Margolis (2001) define equilibrium “as a condition such that no change will occur unless some outside vent intrudes.
They further distinguish between stable equilibriums in which the system will return to the equilibrium in the event of small perturbations away from the position, and unstable equilibriums in which a system will not return if there is a small perturbation.
Microeconomic theory provides three main competition models for oligopoly that can aid in this process: Cournot, Bertrand, and Stackelberg (Besanko and Braeutigam 2005), of which the first two are most widely used. In the Cournot model, each firm selects a quantity to produce, and the resulting total output determines the market price. Alternatively, in the Bertrand model each firm selects a price and meets demand for its product at that price, taking as given the prices of other firms.
To successfully apply either Cournot or Bertrand models on residential broadband deployment knowledge of how customers value FTTH and DSL is necessary, in addition to analysis about the willingness of the competitors to lower prices. In the absence of this data, this thesis can only make very rough estimates on how companies will behave and how customers will react. Additionally competitive actions are likely to be different for different geographical areas, given differences in profitability of different strategies.
Using intuition, the most likely competitive situation would be one where the DSL provider would use lower marginal price to compete on price, and FTTH providers would try to focus on the technological superiority of fibre to differentiate the products. One way of solving this kind of a problem is to use a Bertrand price competition with horizontally differentiated products (Besanko and Braeutigam 2005). When successfully implemented, this model can predict equilibrium prices and market shares of each firm. However, to implement and solve the model for the residential broadband market empirical information about the demand function is needed. Gasmi, Vuong, and Laffont (1992) solve this by using statistical methods of estimating residual demand curves for two horizontally differentiated products.
However, if the market under analysis can only sustain one firm in the long run, the competition can be expected to be different. Then the firms would not seek equilibrium of coexistence but rather to prevent the competing firm from getting a foothold on the market. For this kind of a market situation the third competition model, called von Stackelberg models, could be of use. Von Stackelberg models generally pertain to situations in which one firm acts as a leader. For this reason von Stackelbeg models are particularly well suited for analysing entry deterrence, i.e. situations when an existing provider makes non-optimal decisions to reduce the likelihood of entry from a competitor. Kreps (1990) suggests this type of model for analysing competition between a monopoly firm and a potential entrant.