TELECOMMUNICATIONS ORDINANCE (CAP.106) Supplementary Determination Made By the Telecommunications Authority under Section 36A of the Telecommunications Ordinance on 30 November 2001 Concerning the Terms and Conditions for the Provision of "Point of Interconnection" Capacity for Interconnection between the Fixed Telecommunications Networks of Wharf New T & T Limited and PCCW-HKT Telephone Limited Short Title 1. This determination may be cited as the Supplementary Determination. Interpretation 2. In this Supplementary Determination unless the context otherwise specifies: "Determination" means the determination issued by the Telecommunications Authority ("TA") under section 36A of the Telecommunications Ordinance on 16 March 2001 concerning the terms and conditions for the provision of 'Point of Interconnection' ("POI") capacity for interconnection between the fixed telecommunications networks of New T & T Hong Kong Limited (now re-named "Wharf New T & T Limited") and PCCW-HKT Telephone Limited; "the parties" means PCCW-HKT Telephone Limited ("PCCW-HKTC") and Wharf New T & T Limited ("New T & T") (previously named "New T & T Hong Kong Limited"); "External Telecommunications Services" has the same meaning as in the determination made by the TA under section 36A of the
Telecommunications Ordinance on 30 December 1998; "Value-Added Services" means all services licensed under Public Non- Exclusive Telecommunications Service Licence except External Telecommunications Services; "forecast period" means the period ef the 12 months starting from 1 March of the current year; and "past 12 months" means the period of the past 12 months immediately preceding the forecast period. The Determination 3. The TA, in exercise of his power under section 36A of the Telecommunications Ordinance, having been satisfied that the parties have been afforded reasonable opportunity in accordance with section 36A(4) to make representations to him as to why a determination should not be made; having considered all submissions made by the parties concerning this Determination; having considered the Preliminary Analysis issued by Office of the Telecommunications Authority ("OFTA") on 4 September 2001; having been satisfied that the parties have been afforded reasonable opportunity to consider the Preliminary Analysis and to make representations to the TA in response to the Preliminary Analysis; having considered the Final Analysis annexed to this Supplementary Determination and the recommendations therein; hereby determines that a new paragraph 11A shall be inserted between paragraph 11 and paragraph 12 of the Determination for the avoidance of ambiguity and for the purpose of elaborating on the method for the calculation of the "normal" level of POI capacity. -2-
4. The new paragraph 11A reads: "11 A. In applying the method given in paragraph 11, for each of the five sets of variables available, a linear regression line shall be computed based on the historical data in the past 12 months. This linear regression line shall be projected by extrapolation to the end of the forecast period. The percentage change of the variable in the forecast period shall be determined based on the values of the variable on the regression line at the commencement and the end of the forecast period. The highest percentage change of the five shall be taken. The "normal" level of the POI capacity at the end of the forecast period shall be equal to the level of POI capacity at the commencement of the forecast period on the linear regression line for POI capacity supplied multiplied by (1 + highest percentage change)." Implementation 5. The above paragraph 11A shall form a part of the Determination issued by the TA on 16 March 2001. The "normal" level of POI capacity between the fixed networks of the parties should be calculated according to paragraph 11 of the Determination and the new paragraph 11A specified above. 6. According to paragraph 11 and the new paragraph 11 A, the "normal" levels of POI capacity for Value-Added Services ("VAS") and External Telecommunications Services ("ETS") during the period 1 March 2001-28 February 2002 are calculated as follows: Level of POI Capacity at the Commencement of the Forecast Period "Normal" Level of on the Linear POI Capacity Regression Line for Growth during the period POI Capacity Supplied Rate 1.3.2001-28.2.2002 VAS [ ]T1 [ ] [ ]T1 ETS [ ]T1 [ ] [ ]T1-3-
7. Apart from the insertion of the new paragraph 11A to the Determination, there are no other changes to the content of the Determination. The interconnection framework established by the TA in the Determination should continue to apply. Publicity 8. This Supplementary Determination will be made public. WONG Sik-kei Anthony Telecommunications Authority 30 November 2001.4-
Supplementary Determination Made By the Telecommunications Authority under Section 36A of the Telecommunications Ordinance on 30 November 2001 Concerning the Terms and Conditions for the Provision of "Point of Interconnection" Capacity for Interconnection between the Fixed Telecommunications Networks of Wharf New T & T Limited and PCCW-HKT Telephone Limited Annex Final Analysis Introduction On 16 March 2001, the Telecommunications Authority ("TA") issued a determination ("the Determination") under section 36A of the Telecommunications Ordinance on the terms and conditions for the provision of point of interconnection ("POI") capacity for interconnection between the fixed networks of New T & T Hong Kong Limited ("New T & T")(now renamed "Wharf New T & T Limited") and PCCW-HKT Telephone Limited ("PCCW-HKTC"). Under the Determination, the requesting operator is required to make commitments on the level of utilisation of the POI circuits, which exceeds the "normal" level of POI capacity. The method for the calculation of the "normal" level of POI capacity was set out in paragraph 11 of the Determination. Different Interpretations on the Method for the Calculation of "Normal" Level of POI Capacity 2. After the issuance of the Determination, both parties had a different understanding of paragraph 11 of the Determination in respect of the method for the calculation of the "normal" level of POI capacity. They sought the assistance of the Office of the Telecommunications Authority ("OFTA") in calculating the "normal" level for the 12-month period following the Determination (the "forecast period") -5-
using the method set out in paragraph 11 of the Determination. 3. After reviewing the submissions of PCCW-HKTC and New T & T, the TA observed that paragraph 11 of the Determination might not be sufficiently clear. The ambiguity of paragraph 11 of the Determination had led to two possible methods of interpretation, which were respectively adopted by PCCW-HKTC and New T & T to be the correct method for the calculation of the "normal" level of POI capacity. 4. In PCCW-HKTC's opinion, the "normal" level of POI capacity at the end of the forecast period should be computed by projecting the respective regression lines (based on the data of the 12-month period preceding the forecast period) to the end of the forecast period. The growth rate of the forecast period should be based on values on the regression lines ("Method 1"). 5. In New T & T's view, the "normal" level of POI capacity at the end of the forecast period should be computed by adding to the current level of POI capacity the growth value obtained by multiplying the current level of POI capacity by the highest of the growth rates (out of the five types of growth rates as stated in paragraph 11 of the Determination) ("Method 2"). Actions Taken by the Interconnection Determination Committee 6. On 15 June 2001, the Interconnection Determination Committee ("IDC") invited PCCW-HKTC and New T & T to make a representation to the TA on the methodology that should be adopted for the calculation of the "normal" level of POI capacity. Both companies made their representations to the TA on 21 June 2001. The representations of the two companies were exchanged for comments on 27 June 2001. Their comments on each other's representations were received by the TA on 4 July 2001. 7. Having studied the views of PCCW-HKTC and New T & T, the IDC presented its preliminary views in the Preliminary Analysis issued on 4 September 200L In response to the IDC's invitation to comment on the Preliminary Analysis, both parties made their submissions to the TA on 18 September 2001. -6-
Considerations of the IDC Legal basis of the amendment 8. New T & T submits that the TA does not have the power to amend the Determination. In New T & T's opinion, section 36A of the Telecommunications Ordinance gives the TA the power to determine terms of interconnection but it does not give the TA the power to amend an existing ruling. In addition, the TA had failed to follow the due process and procedures for any variation. 9. The IDC considers that New T & T's objection is misconceived. Section 36A has unequivocally empowered the TA to make determination and the ambit of such statutory power is elaborated upon in section 46 of the Interpretation and General Clauses Ordinance (Cap. 1), which provides that: to amend or suspend such proclamation, order, notice, declaration, instrument, notification, licence, permit, exemption, register or list; (b) to substitute another proclamation, order, notice, declaration, instrument, notification, licence, permit, exemption, register or list for one already made, granted, issued or approved; (c) to withdraw approval of any proclamation, order, notice, declaration, instrument, notification, licence, permit, exemption, register or list so approved; and (d) to declare the date of the coming into operation, and the period of operation, of any such proclamation, order, notice, declaration, instrument, notification, licence, permit, exemption, register or list " [emphasis added] Accordingly, the powers identified are sufficiently ample to enable the TA to "amend" the determination made by him. "Where any Ordinance confers power upon any person to make, grant, issue or approve any proclamation, order, notice, declaration, instrument, notification, licence, permit, exemption, register or list, such power shall include powerfa) -7-
10. Regarding New T & T's allegation that the TA has failed to follow the due process and procedures for making the variation as depicted in paragraph 31 of the Procedures for Making Determination on the Terms and Conditions of Interconnection Agreement (the "Determination Procedures") revised on 27 September 2001, paragraph 31 is, indeed, an illustration of the power of the TA to amend a determination. Although, strictly speaking, New T & T may argue that request for the Supplementary Determination was not formally submitted by a party to a determination, the Supplementary Determination was initiated due to requests for assistance respectively made by New T & T and PCCW-HKTC to OFTA in calculating the "normal" level of POI capacity. According to section 39 of the Interpretation and General Clauses Ordinance, which says: "(7) Where any Ordinance confers any power or imposes any duty, then the power may be exercised and the duty shall be performed from time to time as occasion requires" An occasion has arisen in the particular circumstance of this case, which makes it necessary for the TA to amend the Determination. Section 36A(2) has also expressly empowered, amongst other things, the TA to make a determination (with the inherent power to "amend" a determination under section 46 of Cap.l), in the absence of request, if he considers it is in the interest of the public to do so. In considering the public interest, the TA has also given regard to the factors under section 36A(10) of the Telecommunications Ordinance: The clarification by the Supplementary Determination of the methodology for calculating the "normal" level of POI capacity would facilitate the enforcement of the Determination, which was intended to assist operators in Hong Kong to continue to offer a wide range of quality telecommunications services at a reasonable cost and in the most economically efficient manner possible, promote the liberalisation policy objectives of the Government and enhance consumer interests. The Supplementary Determination aims to devise a model for fair sharing of the risks in the investment of the network equipment for the provision of interconnection capacity which exceeds the capacity that could be agreed by both parties. The model would promote the fair competition between New T & T and PCCW-HKTC and facilitate the effective investment in telecommunications
infrastructure. The Supplementary Determination is particularly necessary on this occasion based on the justifications as elaborated in paragraphs 12-25 below. 11. Moreover, the TA has followed the due process and procedures in making the Supplementary Determination. The IDC has sought the views of both parties several times, arranged the exchange of their views and issued a preliminary analysis to solicit the opinion of both parties before making a recommendation to the TA. Both parties have in fact been given sufficient opportunity to make their representations to the TA and the views expressed by both parties have been fully considered by the TA before the TA makes his decision. Justifications for making amendment to the Determination 12. New T & T considers that there is no reason for the TA to insert a new paragraph 11 A. New T & T disagrees that paragraph 11 of the Determination is not "sufficiently clear". It considers that the method regarding the calculation of the normal level of POI capacity has been stated very clearly in paragraph 11 of the Determination. New T & T's argument, however, is not substantiated by what had happened since the making of the Determination. If paragraph 11 were "sufficiently clear", it would not lead to the emergence of two interpretations submitted respectively by PCCW-HKTC and New T & T. Paragraph 11 of the Determination states that: ' '11. The "normal" level of POI capacity referred to in paragraph 8(b) for each type of traffic should he calculated by applying the highest of the following growth rates to the current level of POI capacity:" However, there is no elaboration on how the growth rates should be "applied". While New T & T is of the view that the word "applying" should be interpreted as "multiplying", PCCW-HKTC disagrees and considers that the word "applying" should be interpreted as "putting into use". After considering the views of both parties, the TA accepts that paragraph 11 of the Determination is not "sufficiently clear", in view of the fact that both New T & T's and PCCW-HKTC's interpretations are compatible with paragraph 11 of the Determination in the literal sense. 13. In the absence of a detailed methodology specified in paragraph 11 of the
Determination for the calculation of the "normal" level of POI capacity, the TA realises that paragraph 11 is open for different interpretations. Such an ambiguity will not only make New T & T and PCCW-HKTC unable to reach an agreement on the "normal" level of POI capacity but also make the Determination difficult to be enforced. The purpose of the Supplementary Determination is to remove the ambiguity by specifying and elaborating on the methodology that should be applied for calculating the "normal" level of POI capacity. In making the Supplementary Determination, the TA is mindful of giving a sensible and reasonable meaning, which would best construe the letter, intent and spirit of paragraph 11 of the Determination and paragraph 25 of the Final Analysis to the Determination: "25. The IDC proposes to adopt a model which aims to achieve fair sharing of the risk in the investment of the network equipment for the provision of interconnection capacity which exceeds the capacity that could be agreed by both parties. The model also aims to institute some form of discipline on the operator which requests interconnection capacity so that it would bear the cost consequence of making excessive forecast The objective is to ensure that New T & T could have all the capacity that it needs for its development but at the same time the model provides a safeguard to PCCW-HKTC in the event that it is proved that the order of New T & T is not justified in terms of demand for service. " [emphasis added] 14. Having studied the views of PCCW-HKTC and New T & T, the IDC's preliminary view (as stated in the Preliminary Analysis) was that Method 2 is not a sufficiently reliable mathematical method for the purpose of forecasting and is not a common method for such forecasting purposes. Method 2, in particular, would yield an unreasonably high forecast when the starting base value is low because it makes the assumption that the growth will increase exponentially. The IDC considers that Method 1 should be more reasonable because it applies a best-fit line to the past 12 months' data and derives the forecast result from this line by extrapolation. Unlike Method 2 which assumes there is an exponential growth in the forecasting period, Method 1 assumes that the growth in the forecasting period will be constant at a slope same as that in the past 12 months. 15. New T & T disagrees to the IDC's preliminary view that Method 2 is not a sufficiently reliable nor a common mathematical method for the purpose of -10-
forecasting. It submitted two papers written by academics, one from Dr. Philip Chan 1, another by Dr. Vincent Cho and Dr Eric Cheng 2, to substantiate that Method 2 is not inferior but is superior to Method 1. 16. Dr. Philip Chan submits that the problem of Method 1 is that it applies a best-fit straight line to the historical data and derives the forecast result from this line by extrapolation. As the actual growth trend (i.e. f(t)) may not be best reflected by a straight line, Dr. Philip Chan considers that Method 1 would have the risk of underestimating the actual growth rate "if f(t) is convex (that is, curves up or has increasing rate of change), and over-estimating the actual growth rate if f(t) is concave (that is, curves down or has a decreasing rate of change)". 17. With reference to the historical data of New T & T in the past years, Dr. Philip Chan observes that the actual growth trends of New T & T's data (in number of lines installed, busy hour traffic and installed circuits for the VAS and ETS etc.) are convex functions. Therefore, he holds the view that Method 1 is inappropriate because it would "seriously under-estimate the true growth rate". Dr. Philip Chan comments that, comparatively speaking, Method 2 is superior to Method 1 because the forecast made by Method 2 "appears to be a more reliable estimate of the true growth rate, no matter whether f(t) is a straight line, or whether f(t) curves up or curves down". 18. The IDC notes that Dr. Philip Chan is relying on New T & T's historical data of VAS and ETS routes (including busy hour traffic and number of circuits installed) during March 1998 - June 2001 to characterise the growth trend of these data points as one in convex shape. It is on this basis that Dr. Philip Chan holds the view that Method 1 is inappropriate because it has the drawback of under-estimating the growth rate of a convex curve. However, the IDC observes that the convex shape curve as presented by Dr. Philip Chan is, to a large extent, attributed to the inclusion of the June 2001 figure (a month of exceptionally strong growth in busy hour traffic and number of circuits installed). The IDC notes that, had the June 2001 figure not been included, the historical growth trends of busy hour traffic and number of circuits installed would not display a significantly convex shape. As there is no data supporting that the convex curvature of the trends would continue after June 2001, the IDC considers that it would be too early to conclude, by relying on one 1 2 Associate Professor, Department of Industrial and Manufacturing Systems Engineering, The University of Hong Kong. Both are Assistant Professors, Department of Management, The Hong Kong Polytechnic University. -11-
exceptional month's data in June 2001, that the growth trends of the VAS and ETS traffic in terms of busy hour traffic and number of circuits installed are in convex shape. 19. The IDC observes that Dr. Philip Chan has explained at length the "weaknesses" of Method 1 but he has not provided any explanation to substantiate the "strengths" of Method 2 and the basis on which he concludes that the forecast made by Method 2 "appears to be a more reliable estimate of the true growth rate, no matter whether f(t) is a straight line, or whether f(t) curves up or curves down". The IDC also notes that Dr. Philip Chan has commented that the values of "growth rates" calculated using Method 1 are much lower than those calculated by Method 2. However, he has not explained why he opines that the lower growth rate forecasted by Method 1 is less representative than that calculated by Method 2. Also, Dr. Philip Chan has not explained why he considers that the growth rate forecasted by Method 2 reflects the actual growth rate more accurately. 20. Dr. Vincent Cho and Dr. Eric Cheng accept that both Method 1 and Method 2 are "basic models capable of predicting a time series". However, they hold the view that Method 2 is better than Method 1 for forecasting purpose, on the ground that Method 2 appears to be more suitable to reflect the dynamic telecommunications market of Hong Kong. Though taking this view, Dr. Cho and Dr. Cheng caution that Method 2 has a drawback that it would have a larger prediction error if the growth trend is volatile and, in general, the reliability of both methods cannot be guaranteed in the sense that time series forecasting is based on extrapolation. Dr. Cho and Dr. Cheng suggest to use a longer period of data and to use other forecasting methods for comparison 3. 21. Dr. Cho and Dr. Cheng use the Mean Absolute Percentage Error ("MAPE") to illustrate that Method 2 is better than Method 1 for forecasting purpose. However, the IDC notes that Dr. Cho and Dr. Cheng have used a set of data which is different from that presented by the IDC at the Preliminary Analysis to calculate the MAPE. Dr. Cho and Dr. Cheng have not explained why they have chosen not to use the historical data presented in the Preliminary Analysis. The IDC discovers that, if the historical data presented in the Preliminary Analysis are used for the calculation of MAPE, the result will be opposite - Method 1 will yield a smaller MAPE than 3 The IDC had proposed to use a longer period of data but such a proposal was not supported by both New T & T and PCCW-HKTC [see paragraphs 98-100 of the Final Analysis annexed to the Determination. -12-
Method 2 in most situations. In view of this, the IDC considers that Dr. Cho and Dr. Cheng's argument that Method 2 is a better than Method 1 for forecasting purpose is not well substantiated. 22. After examining the views of Dr. Philip Chan and Dr. Vincent Cho/Dr Eric Cheng, the IDC notes that their advice is focused on assessing whether Method 1 or Method 2 is a better forecasting technique from an academic angle. However, the IDC would like to point out that the decision on whether Method 1 of Method 2 should be adopted for the calculation of the "normal" level of POI capacity is, in fact, a policy decision. The mechanism for the calculation of the "normal" level is not intended to forecast the future trend of the POL It is just a part of the arrangement established in the Determination made by the TA on the terms and conditions for the provision of POI between the fixed networks of New T & T and PCCW-HKTC. While the Determination is the core issue, it is not appropriate to put the mechanism for the calculation of the "normal" level in isolation for consideration without giving due regard to the objective and content of the Determination. 23. Under the arrangement established in the Determination, the "normal" level is not the forecast future level of POI capacity between New T & T and PCCW- HKTC. It is just the triggering point for requiring the requesting operator to make traffic commitments on the requested POI capacity. According to the Determination, New T & T may request as much POI capacity as it needs from PCCW-HKTC (even if such POI capacity is beyond the "normal" level) and PCCW-HKTC must meet New T & T's request (notwithstanding PCCW-HKTC's disagreement that the requested POI capacity is necessary). 24. New T & T and PCCW-HKTC have disputes on the demand forecasts for POI capacity between their networks for a long period of time. The Determination is intended to achieve a fair sharing of the risk between New T & T and PCCW-HKTC of the investment in POI capacity for meeting the disputed demand forecasts. According to the Determination, if New T & T's forecast turns out to be true, New T & T will not need to pay more in terms of usage charge for the POI capacity acquired. But if New T & T's forecast turns out to be wrong, New T & T will need to compensate PCCW-HKTC for the over capacity by paying the shortfall in traffic volume. The Determination is fair to both parties, as it would let New T & T acquire the POI capacity it requests (even if such capacity is considered to be excessive from PCCW-HKTC's point of view) and enable PCCW-HKTC to get a reasonable "insurance" on the investment made for meeting New T & T's request. -13-
25. In considering whether Method 1 or Method 2 should be adopted, the IDC is mindful of the respective consequences of an over-estimation and an underestimation of the "normal" level of POI capacity to both parties. The IDC notes that the consequence of under-estimation is less undesirable. This is because, even when the "normal" level is under-estimated, neither New T & T nor PCCW-HKTC will suffer. New T & T will still get the POI capacity from PCCW-HKTC and will not suffer from any loss as long as the usage of the POI capacity is not under-utilised. Conversely, if the "normal" level is over-estimated, there will be a higher risk in excessive investment in POI capacity. An unreasonably high forecast of "normal" level will remove the pressure for a serious consideration of the need for the POI capacity before making the request. The providing operator would not be fairly compensated for the investment which is under-utilized. Under-utilization of capacity is also incompatible with the Government policy objective that "telecommunications services should be provided in the most economically efficient manner possible". Injustice due to abrupt change of the TA '$ decision 26. New T & T considers that the abrupt behaviour of the TA to amend his earlier ruling in such a short space of time is disturbing since the Determination was just issued some 6 months ago. New T & T opines that the sudden amendment to the Determination would adversely affect the industry's confidence on the TA. As an administrator to promote the Government's policy objectives, the TA should provide certainty to its decision that shapes the regulatory framework. Business decisions involving significant financial commitments are made in reliance on the TA's decisions. The TA should not change its decision on a flimsy moment without due consideration and proper justifications. The IDC disagrees that the amendment represents an abrupt change of the TA's decision. Rather the amendment seeks to remove the ambiguity by adding paragraph 11A to specify the methodology for calculating the "normal" level that would best construe the letter, intent and spirit of the Determination of fairly allocating risks between the providing operator and the requesting operator. 27. New T & T is also concerned that the TA's proposed amendments to the Determination, if implemented, will significantly increase the cost structure of New T & T and reduce the flexibility for New T & T to offer better and more attractive terms to retail customers. New T & T also submitted that it had placed numerous orders -14-
for capacity subsequent to the issuance of the Determination by the TA in March 2001. The IDC fully understands New T & T's concern but would like to point out that New T & T's concern is not warranted in this particular case. The amendment to the Determination aims only to clarify the ambiguity of paragraph 11 and establish a method for the calculation of the "normal" level of POI capacity that could best achieve the intent and spirit of the Determination. As New T & T has submitted, the POI capacity was ordered according to New T & T's business needs. So long as New T & T has not ordered excessive POI capacity and under-utilised such capacity, it will not suffer from any loss as a result of the Supplementary Determination. New T & T would pay for the utilisation of the POI capacity at the normal usage rate. The model as depicted by the Determination and the Supplementary Determination would require the requesting operator to bear the cost consequence only if it has ordered excessive POI capacity and under-utilised such capacity. This principle had been clearly explained in paragraph 54 of the Final Analysis to the Determination as follows: "54. In the Preliminary Analysis, the IDC considered that the proposed model has achieved sharing of the risk between PCCW-HKTC and New T &T of the investment in capacity to meet the disputed demand. There are two scenarios to consider: (a) If the growth of the narrowband Internet traffic would continue as New T & T has predicted during the lifetime of the investment, or if New T & T's growth forecast for ETS traffic turns out to be accurate, New T &T would obtain the capacity at the normal usage rate while PCCW-HKTC would recoup the investment through the payment of normal usage charges throughout the lifetime of the investment. (b) Data for the Calculation of "Normal" Level of POI Capacity If the narrowband Internet traffic would decay as PCCW- HKTC has predicted during the lifetime of the investment, or if New T & T's growth forecast for ETS traffic turns out to be excessive, New T & T would pay the shortfall of the interconnection payment for the Prescribed Period." -15-
28. New T & T submits that it will be more equitable to use the data from 1 January to end of December for the calculation of the "normal" level of POI capacity. 29. PCCW-HKTC proposes to adopt the data for the period from 1 April 2000 to 31 March 2001 to calculate the "normal" level of POI capacity for the forecast period of 1 April 2001 to 31 March 2002. PCCW-HKTC also proposes that the normal level of POI capacity in subsequent years should also be calculated in cycles of period commencing on 1 April and ending on 31 March. 30. The IDC considers that neither New T & T's nor PCCW-HKTC's proposal is reasonable. Both parties have not put forward any convincing arguments to substantiate that the period they proposed is more equitable. The IDC is therefore of the view that it is more appropriate to adhere to the arrangement as specified in paragraph 11 and paragraph 11A of the Determination that the data input for the calculation of the "normal" level of POI capacity should be the data "in the past 12 months". As the Determination was made on 16 March 2001, and data up to the end of February 2001 was then available, the IDC considers that the appropriate data input for the calculation of the "normal" level of POI capacity for 2001/2002 should be the data in period from 1 March 2000 - end of February 2001. The data input for the calculation of the "normal" level of POI capacity in subsequent years should be collected based on the same cycle. Conversation Erlang Loss Formula 31. In respect of the variable in paragraph 11 (a), PCCW-HKTC submitted that a conversion Erlang loss formula of 0.5% grade of service ("GOS") should be used to convert the projected Busy Hour traffic Erlang at the end of the 12-month forecast period into the number of POI trunks in order to calculate the percentage change as compared with the level of POI capacity at the commencement of the forecast period on the linear regression line for POI capacity supplied. 32. The IDC does not see there is any need to implement an additional step to use the conversion Erlang loss formula of 0.5% GOS to convert the projected Busy Hour traffic Erlang into the number of POI trunks. The Busy Hour traffic Erlang is already a true and reliable reflection of the volume of traffic passing through the POI capacity. The adoption of the additional step to convert the forecast Busy Hour traffic Erlang into POI trunks by the conversion Erlang loss formula will not provide any improvement to the accuracy and the reliability of the forecast. The IDC -16-
HKTC to adopt the conversation Erlang loss formula to convert the forecast Busy Hour traffic Erlang into the number of POI trunks. The Supplementary Determination 33. For the purpose of avoiding ambiguity, the IDC recommends that the TA should issue a supplementary determination to clarify the calculation method for the "normal" level of POI capacity as envisaged in paragraph 11 of the Determination. The supplementary determination will add a new paragraph 11A to the Determination to elaborate on the method for the calculation of the "normal" level of POI capacity. The text of paragraph 11A should read: "11 A. In applying the method given in paragraph 11, for each of the five sets of variables available, a linear regression line shall be computed based on the historical data in the past 12 months. This linear regression line shall be projected by extrapolation to the end of the forecast period. The percentage change of the variable in the forecast period shall be determined based on the values of the variable on the regression line at the commencement and the end of the forecast period. The highest percentage change of the five shall be taken. The "normal" level of the POI capacity at the end of the forecast period shall be equal to the level of POI capacity at the commencement of the forecast period on the linear regression line for POI capacity supplied multiplied by (1 + highest percentage change)." The "Normal" Levels of POI Capacity 34. According to paragraph 11 and the new paragraph 11 A, the "normal" level of POI capacity during the period 1 March 2001-28 February 2002 should be as follows (detailed calculation is shown in the Appendix): therefore considers that there is no need to take the step as suggested by PCCW- -17-
Level of POI Capacity at the Commencement of the Forecast Period Normal Level of on the Linear POI Capacity Regression Line for Growth during the period POI capacity Supplied Rate 1.3.2001-28.2.2002 VAS [ ]T1 [ ] ETS [ ]T1 [ ] Interconnection Determination Committee Office of the Telecommunications Authority 30 November 2001-18-