Demand Analysis and Penetration Forecasts for the Mobile Telephone Market in the U'K'

1
Demand Analysis and Penetration Forecastsfor the
Mobile Telephone Market in the U.K.

• Presentation to the 18th Annual ICFC
• Seattle, Sept. 26 - 29, 2000
• Chris Dineen
• Teligen Ltd
• cdineen_at_storm.ca

2
The U.K. Mobile Market
3
Data

• mobile telephone data are available from Oftel,
  the U.K. regulator
• quarterly series, which include calling minutes,
  new connections, number of subscribers and
  revenues for each of the competing operators
  (Vodaphone, Cellnet, One2One and Orange) are
  issued with a time lag of about 8 months
• also, shortly after quarterly results are
  tabulated, each operator usually issues a press
  release detailing its gains in net subscriptions
  split by contract and prepaid

4
Demand Model Form

• Minutes1 ?1 ?1Price1 ?1GDP ?1 Subs1
  ?1
• Minutes2 ?2 ?2Price2 ?2GDP ?2 Subs2
  ?2
• Minutes3 ?3 ?3Price3 ?3GDP ?3 Subs3
  ?3
• Minutes4 ?4 ?4Price4 ?4GDP ?4 Subs4
  ?4
• Price is real average revenue per minute (ARPM)
• Subs is number of subscribers
• SURE model ?j ? ?k, ?j ? ?k etc.
• Fully pooled model ?j ?k, ?j ?k etc.
• Hybrid model some coefficients constrained,
  others unconstrained

5
Demand Model Results

• Dependent Variable LOG_MINUTES
• Method GLS (Cross Section Weights)
• Sample 19963 19994 
• Variable Coefficient Std. Error t-Statistic
• LOG_REAL_PRICE -0.4651 0.0505 -9.2144
• LOG_SUBS_VOD 0.6572 0.0299 22.0082
• LOG_SUBS_CELL 0.7435 0.0383 19.4000
• LOG_SUBS_ONE 0.4942 0.0250 19.7518
• LOG_SUBS_ORA 0.7305 0.0412 17.7461
• LOG_REAL_GDP 2.3899 0.4711 5.0730
• Other variable Prepaid Dummy
• R-squared 0.9980
• Adj R-squared 0.9974
• S.E. of regression 0.0298
• Durbin-Watson 1.7447
•  
•  
•  

6
Demand Model Discussion

• Constant Price elasticity -0.47
• Static price lagged price insignificant
• Well behaved model with good fit and reasonable
  coefficients

7
Penetration Model Form

• Subscribers1 ?1 ?1Price1 ?1Advertising
  ?1 Time1 ?1
• Subscribers2 ?2 ?2Price2 ?2Advertising
  ?2 Time2 ?2
• Subscribers3 ?3 ?3Price3 ?3Advertising
  ?3 Time3 ?3
• Subscribers4 ?4 ?4Price4 ?4Advertising
  ?4 Time4 ?4
• Price is Real ARPM
• Adv is total industry advertising expenditure
• Time is an exponential time trend

8
Penetration Elasticity Model Results

• Dependent Variable SUBSCRIBERS
• Method Seemingly Unrelated Regression
• Sample 19963 19992
• Variable Coefficient Std. Error t-Statistic
• LOG_REAL_ARPM -485.502 20.236 -23.992
• LOG_TOTAL_ADV_VOD 553.022 259.868 2.128
• LOG_TOTAL_ADV_CEL 348.949 156.966 2.223
• LOG_TOTAL_ADV_ONE 536.336 222.723 2.453
• LOG_TOTAL_ADV_ORA 366.993 155.380 2.362
• Other variable exponential time trend
• R-squared 0.9458
• Adjusted R-squared 0.9273
• S.E. of regression 375.769
• Durbin-Watson stat 1.669

9
Penetration Elasticity Model Discussion

• Price elasticity at mean value of dep var
  -0.050
• Price elasticity at max value of dep var
  -0.029
• Unstable model

10
Forecasting Penetration

• Bi-Logistic Curve
• y ?1 / (1 exp-?1(t - ?1))
• ?2 / (1 exp-?2(t - ?2))
• ? is the upper asymptote or saturation level, the
  value which penetration approaches over time
• ? indicates the growth rate
• ? indicates the time of maximum growth
• Hypothesis based on examination of subscription
  chart contract subscriptions represented by
  first logistic curve, prepaid represented by
  second

11
Penetration Forecasting Model Results (1)

• Data Range 1995 Q1 to 2000 Q2
• Coefficient Estimates
• ?1 18.1 ?2 42.7
• ?1 0.178 ?2 0.529
• ?1 14.0 ?2 10.0
• Implications
• saturation level of penetration for contract
  subscriptions is about 24.9 of the population
• saturation level for prepaid is about 42.7 of
  the population
• ultimate total penetration level for mobile
  phones in the U.K. is about 67.6 of the
  population

12
U.K. Mobile PhonePenetration Forecasts
13
Penetration Forecasting Model Results (2)

• growth times from 10 to 90 of saturation 
• Contract 24.7 quarters
• Prepaid 8.3 quarters
• Fixed-Line Penetration Forecasting Model
• Growth has followed a simple linear trend
• Y 48.29 0.485time
• R2 0.997
• Mobile penetration will overtake fixed line
  penetration in 2001 Q1

14
Forecasts of Fixed-Line and Mobile Penetration in
the U.K.
15
Some Definitions

• CPP (calling party pays)
• the initiator of a call to a mobile phone pays
  the full cost
• in effect in the U.K. and Europe
• RPP (receiving party pays)
• both the initiator and the receiver of a call to
  a mobile phone pay a portion of the costs
• in effect in Canada, U.S. and many other
  countries in the Americas
• Contract - subscribers pay a monthly fee for
  network access and a per-minute rate for outgoing
  (and incoming calls in RPP regimes) calls. The
  monthly fee usually includes an initial allotment
  of calls
• Prepaid - customers pay in advance for a given
  number of minutes and has no contract with the
  service provider. A credit check is not required