Competition and Market Strategies in the Swiss Fixed Telephony Market An estimation of the incumbent’s dynamic residual demand curve Roberto Balmer Bundesamt für Kommunikation Disclaimer: The views presented here are those of the author and do not reflect those of BAKOM. Biennial Conference International Telecommunications Society 1 December 2014 Link to working paper
15
Embed
Competition and market strategies in the swiss fixed telephony market
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Competition and Market Strategies in the
Swiss Fixed Telephony Market
An estimation of the incumbent’s
dynamic residual demand curve
Roberto Balmer
Bundesamt für Kommunikation
Disclaimer: The views presented here are those of the author and do not reflect those of BAKOM.
• Retail fixed telephony regulation lifted in many European States
• Sufficient competition?
• Demand/supply model for Switzerland 2004-2012
The challenge:
• Highest frequency data available is usually quarterly (few data points)
• Complexity of market model strongly limited
Strong simplifications:
1) No vertical integration issues:
Identical network origination costs are assumed across competitors (access at efficient marginal costs for access seekers and infrastructure-based operators). LRIC prices therefore assumed to be close to marginal cost (exogenous).
2) Two-way access:
a) Regulated fixed termination rates (LRIC) are close to zero (<1€cent/min). No fixed termination costs & revenues modeled.
b) Mobile termination rates are unregulated in Switzerland. Rate modeled as exogenous cost shifters for fixed operators (>10€cent/min).
1. Introduction 2. Model 3. Variables 4. Conclusions
3
Market and Assumptions (2/2)
3) Marginal costs are assumed to increase with output (customer care, capacity constraints in backhaul) in the segment under consideration
4) Fixed-mobile substitutionis assumed to be limited (mobility, prices) &fixed divisions of fix-mobile operators act to independently
5) VoBBis assumed to not be a substitute (QoS)
6) Market for Accessis abstracted from (inelastic demand, high penetration, no price changes).
Fixed fee is abstracted from as not critical. No waterbed effects expected as FTR is assumed to be zero.
Only a average linear tariff for national fixed originated traffic is considered (market ARPM).
1. Introduction 2. Model 3. Variables 4. Conclusions
4
Traditional model assuming a particular way how firms compete:
• Dominant firm takes the reaction of the fringe firms into account when deciding on quantity and price
• Fringe firms are assumed to be pure price takers ( )
• Fringe supply is therefore given by sum of fringe marginal cost curves.
• Incumbent Swisscom maximizes profitability with Q and P such that
MR(residual demand)=MC (Swisscom)
Fcp
Dominant firm – Competitive Fringe (1/3)
1. Introduction 2. Model 3. Variables 4. Conclusions
DM is the market demand curve (Q is total
demand)
𝑆𝐹 is the fringe supply curve (𝑄𝐹 is fringe supply)
𝐷𝑅𝑒𝑠𝑖𝑑𝑢𝑎𝑙 is the residual demand curve that the dominant
firm faces (𝑄𝐷 is dominant firm demand)
𝑀𝑅𝑅𝑒𝑠𝑖𝑑𝑢𝑎𝑙 is the residual marginal revenue curve the
• Fringe supply depends on marginal costs which in turn depend on the level of output & possible cost shifters (unique to fringe firms or common with incumbent).
Supply relation not structurally changed with λ<1 (constant)
• Dominant firm residual demand:
expressed in log-linear (isoelastic) form:
1. Introduction 2. Model 3. Variables 4. Conclusions
dominant firmgeneralised, λ<1 is lessdominant behaviour
6
Model equation 1 (first stage):
incumbent cost shifters include its number of staff. Instrumented prices are then used to obtain the incumbents residual demand
Model equation 2 (second stage):
λ cannot be estimated in traditional models (need demand rotator or similar). The particular estimates of this model allow, however,to infer a higher bound.
Dominant firm – Competitive Fringe (3/3)
1. Introduction 2. Model 3. Variables 4. Conclusions
7
Variable Definition
𝑝 Price; average revenue per outgoing minute (subscription revenues excluded), deflated
by CPI (100=2006), traditional and proprietary VoIP
𝑞𝐷 Swisscom’s fixed national outgoing calling minutes (to fixed and to mobile)
𝑥1 Income; real GDP per capita (100=2006)
𝑥2 Number of active telephony lines (PSTN)
𝑥3, 𝑥3, 𝑥4 Quarterly dummies for quarters 2, 3 and 4
𝑤𝐹1 Number of wholesale ADSL lines sold by Swisscom
𝑤𝐹2 Number of ULL access lines sold by Swisscom
𝑤𝐶1 Average regional fixed voice origination prices per minute deflated by consumer price
index (CPI) (100=2006)
𝑤𝐶2 Weighted average of mobile termination rates deflated by CPI (100=2006)
𝑤𝐶3 Interest rates on 30 year bonds of the Swiss Confederation
𝑤𝐶4 Exchange rate EUR/CHF
𝑤𝐷1 Number of staff working for Swisscom
𝑤𝐷2 Number of active retail ADSL lines (Swisscom)
Demand shifters:
Fringe supply shifters:
Common supply shifters:
Dominant firm supply shifters:
Variables (1/2)
1. Introduction 2. Model 3. Variables 4. Conclusions
8
Econometrical problems
• Variables p and q decline over time (non-stationarity)
• Errors are serially correlated
Solution
To avoid spurious regression results: ARDL(1,1)
• Sufficient level of cointegration
• No serial correlation
Variables (2/2)
1. Introduction 2. Model 3. Variables 4. Conclusions
9
• ARDL(1, 1) implies that a change in an independent variable has a direct same period effect (coefficient; „impact multiplier“) and longer and long term effects („long term multiplier“). A dynamic residual demand function of the incumbent is therefore estimated:
• The long term multiplier can be interpreted as long term cumulative elasticities (log linear function).
• Multiplier effects from t to t+k:
-0.300
-0.200
-0.100
0.000
0.100
0.200
0.300
0.400
q t+1 quarter t+2 quarter t+3 quarter t+4
Results (1/2)
exchange rate
mobile termination
ADSL wholesale lines
Price
Regulated origination ratesInterest rates
Long term effects on Swisscom‘s residual demand:
- Price (-0.124)
Demand shifters
- Income elasticity (1.5) – not on graph
- PSTN lines (2.4) – not on graph
Fringe cost shifters
-ULL lines (-0.002)
-ADSL wholesale lines (-0.158)
Common cost shifters
-Mobile termination rates (0.141)
-Interest rate (-0.035)
-Exchange rate (0.473)
-Origination price (-0.051)
ULL lines
1. Introduction 2. Model 3. Variables 4. Conclusions
10
• Most importantly price elasticity for Swisscom voice traffic is shownto be very low (-0.124)
• 10% market price increase corresponds to a decline in Swisscom traffic by 1.24% (longterm).
• Such a low elasticity is incompatible with a profit maximising dominant firm (λ=1) asMR is negative (MC positive).
• When λ<0.12, MR is positive again (conduct is compatible with profit maximisation)
Conclusion:
• Incumbent Swisscom acted largely competitively (λ=0 is price taking)
• Todays very limited price regulation (Swisscom F2F calls) could be lifted.
Variable
ARDL (1,1) estimates
Coefficient
estimate
Robust Std.
err.P>| t |
p -0.044 0.044 0.321
L1 -0.102 0.049 0.038
LR -0.124
Results (2/2)
1. Introduction 2. Model 3. Variables 4. Conclusions