REGULATORY REFORM IN THE JAPANESE ELECTRIC POWER INDUSTRY AN EVENT STUDY ANALYSIS IAEE 2017 Conference, Singapore 20 th June 2017 Koichiro Tezuka,

Transcription

1 REGULATORY REFORM IN THE JAPANESE ELECTRIC POWER INDUSTRY AN EVENT STUDY ANALYSIS IAEE 2017 Conference, Sngapore 20 th June 2017 Kochro Tezuka, Nhon Unversty, Masahro Ish, Sopha Unversty, Satoru Hashmoto, Tekyo Unversty,

2 Agenda 1. Introducton 2. Regulatory reform n the Electrcty sector n Japan 3. Event Study Analyss 4. Data and Methodology 5. Results and Interpretatons 6. Concluson and Future Research

3 1. Introducton In ths study, we observe the effects of the recent regulatory reform on the electrcty ndustry n Japan. After 11 th March, 2011 of Fukushma, the government and METI(Mnstry of Economy, Trade, and Industry) changed the electrcty system drastcally. To see ths, we apply an event study analyss to the ndustry.

4 1. Introducton The problems are: When the regulatory reform was substantally decded? Complcated polcy makng process Dffcult to dentfy the tmng When the stock market recognzed the reform? 1 st regulatory reform passed Parlament on 13 Nov Cabnet mnsters decson on 2 nd Aprl 2012 These two had been already factored nto the market.

5 1. Introducton In ths study, By utlzng an event study analyss, we try to dentfy the tme when market recognzed the reform. By utlzng the results, we also provde nterpretatons from the results. when the regulatory reform substantally decded?

6 2. Regulatory reform n the electrcty sector n Japan After Fukushma, The government provded the gudelne about the Electrcty System Reform on 2 nd Aprl, The gudelne conssts of three parts: 1. To expand the cross-regonal coordnaton of transmsson, 2. Fully deregulaton of the retal and power generaton sectors, 3. Legal separaton of power generaton and transmsson/dstrbuton So called unbundlng.

7 2. Regulatory reform n the electrcty sector n Japan After the gudelne was announced, the Electrcty Busness Act has revsed for 3 steps: 1. By the frst revson on 13th November 2013, the Organzaton for Cross-regonal Coordnaton of Transmsson Operators (OCCTO) was establshed. 2. Entres to electrcty retal busness were fully lberalzed n the second revson on 11 th June As the thrd step, t was decded on 17 th June 2015 that the transmsson/dstrbuton sector wll be unbundled legally on 1 st Aprl 2020.

8 Electrcty companes n Japan Source METI Each EPCO vertcally operates power generaton, transmsson/dstrbuton, and retal.

9 Image of Unbundle

10

11 Commttee of Electrcty System Reform

12 2. Regulatory reform n the Electrcty sector n Japan Regardng the problem when unbundle was subtantlly decded?, we pcked up three events before the fnal decson of the reform: th October, 2011: The cabnet decded to set taskforce about regulatory reform: th December, 2011: The taskforce showed the ponts at ssue nd Aprl, 2013: The cabnet decded to carry out the reform.

13 3. Event Study Analyss Event study analyss To observe reactons of the market for events Two Man Approaches: 1. To compare mean return / before and after 2. To calculate CAR(Cumulatve Abnormal Return) Models for measurng :Macknlay (1997) Constant Mean Return Model Market Model others

14 Models for Measurng Cumulatve Abnormal Return Constant Mean Return Model Market Model R t R mt e t, where 2 and Var e E e t 0 t e 11 Rmt denotes rate of return for market portfolo.

15 Market model/ Estmaton of CAR Estmaton wndow: from T0 to T1 Before the event occur Event wndow τ: from T1 to T2 In ths study we use T1=-10, T2=10 (20 days) Post event wndow (from T2 to T3) Estmaton of Cumulatve Abnormal Return

16 By utlzng market model, Estmaton wndow of observaton m T T and R X R R R where e X R,,

17 OLS Estmator : From T0 to T1 Estmaton wndow of ovservaton R X X X ' ' 1 e e e T T ˆ ' ˆ 1 ˆ X R e ˆ ˆ 2 1 ) ' ( ˆ e X X Var

18 We set vector of abnormal proft as follows: Estmaton of Abnormal Profts m t X R R R R ˆ ˆ * * * * , 1, *, 1, * m T m T m T T R R R R R R ˆ, ˆ ˆ

19 Abnormal Proft Under null hypothess, abnormal proft s: e N(0, We set CAR as follows: ~ ˆ* V CAR ˆ ( T1 1 2 T2 Var * 1, ˆ 2 ) ' e CAR ˆ 2 ( 1, 2)] ( 1, ) [ 2 ) ' V

20 CAR: Cumulatve Abnormal Return Under null hypothess H0 CAR ˆ (, ) Standardzed CAR s: SCAR N(0, (, CAR ˆ (, ˆ 1 2 ( 1, 2 ) ˆ 2 If we substtute ˆ 2 to ˆ e, then dstrbuton of SCAR follows t-statstc wth degree of freedom L1-2 ~ )) )

21 4. Data and Methodology Data descrpton We employ 10 EPCOs and 3 Gas Companes. Lsted wth Tokyo Stock Exchange As proxy varable of market portfolo, TOPIX and NIKKEI225 are employed. Data source : NIKKE NEEDS Fnancal Quests. Sample Perod: 1 st Aprl th Aprl 2013 We use only busness days.

22 Event Study We use Constant Mean Return model and Market model. We employ three event dates as mentoned: th October, 2011: The cabnet decded to set taskforce about regulatory reform: th December, 2011: The taskforce showed the ponts at ssue nd Aprl, 2013: The cabnet decded to carry out the reform Estmaton wndow : 6 month To avod the day of 11 th March Event wndow: 10 busness days before/after the event That s from -10 to 10 Post event wndow: 1 month

23 4. Results and Interpretatons The results are followng sldes: 28 th Oct. 2011, 27 th Dec. 2011, and 2 nd Aprl 2013 Constant Mean Return model Market model

24 CMR, 28 th Oct. 2011

25 CMR, 27 th Dec. 2011

26 CMR, 2 nd Aprl 2013

27 Market Model 28 th Oct. 2011

28 Market Model 27 th Dec. 2011

29 Market Model 2 nd Aprl 2013

30 CAR 28 th Oct. 2011

31 CAR 27 th Dec. 2011

32 CAR 2 nd Aprl 2013

33 28 th Oct 2011 Results of Event study, Market model 27 th Dec nd Aprl 2013

34 5. Conclusons and Future Research From the results, all null hypothess could not be rejected. So we could not fnd the effects of (three) events on stock prces and ther return. Except for CMR, 2 nd Aprl 2013 It may be consdered that the stock market had already allowed for the regulatory reform. If that s the case, when the stock market convnced the mplementaton of the reform? Before? After?

35 5. Conclusons and Future Research Whle polcy makng processes n Japan seem to be gradual, n ths case the decson regardng electrcty system reform seems to be fast. However, t s dffcult to specfy the date when the mplementaton of the reform was substantally decded. In ths study, we could not fnd the date. To fnd ths, we wll nvestgate the crcumstances n more detal, and apply more event days for our future research.

36 Thank you for your attenton