Merger Review Regimes in the ASEAN Region and Case Analysis of Grab-Uber Merger

September 3, 2021 Vol. 11 No. 39 ISSN 2233-9140

Merger Review Regimes in the
ASEAN Region and Case Analysis
of Grab-Uber Merger
Yungshin Jang Associate Research Fellow, Southeast Asia and Oceania Team, New Southern Policy Department (ysjang@kiep.go.kr)
Gu Sang Kang Associate Research Fellow, Americas Team, Advanced Economies Department (gskang@kiep.go.kr)

I. Introduction
ASEAN Leaders adopted the ASEAN Eco- worked as hurdles in integrating the regional
nomic Community (AEC) Blueprint 2025 in economy. This divergence could be more prob-
2015, which guides the broad trajectories of lematic in reviewing cross-border M&As (mer-
ASEAN economic integration from 2016 to gers and acquisitions) because if each competi-
2025. Grounded on the second initiative of the tion authority in the region reaches a contradic-
AEC Blueprint 2025, “A Competitive, Innova- tory conclusion, investors targeting the inte-
tive, and Dynamic ASEAN,” ASEAN estab- grated ASEAN market would be more reluctant
lished the ASEAN Competition Action Plan to invest in the cross-border M&As.
(ACAP) 2025 via the ASEAN Experts Group
on Competition (AEGC).1 This plan elaborates In 2018, the largest yet cross-border M&A deal
strategic measures related to promoting harmo- between digital platforms in Southeast Asia
nization of competition policy and law in the was reached, namely the Grab-Uber M&A case.
ASEAN region. As of August 2021, nine of the The local digital platform Grab consolidated
ten ASEAN Member States have introduced the regional operations of San Francisco, Cali-
their national competition law regimes, with fornia-based Uber, a development which had
Cambodia still reviewing the draft. However, significant effects on competition and con-
detailed competition laws of the member states sumer welfares in the SEA digital market. The
are relatively diverse in their regimes while competition authorities in the region inde-

basic principles enforcing competition laws pendently initiated their investigation and
 started to deliberate the merger case to deter-
display similarities. The institutional differ-
 mine the anti-competitive effects on their do-
ences considering each state’s enforcement en-
 mestic market, and to decide whether this tran-
vironments are natural, but they could be

1 %20(Website)%2023%20December%202016.pdf
 https://asean-competition.org/file/post_image/ACAP

September 8, 2021

saction should be restricted or approved. Even and abuse of dominance, merger review re-
though the two merging and merged firms com- gimes show significantly heterogeneous insti-
pleted their transactions, each authority applied tutional characteristics. In regulating anti-car-
different logic and imposed different remedies tel behaviors, all four countries’ competition
in deciding the case. Authorities in some mem- authorities commonly separate “per se illegal
ber states such as Singapore and the Philippines rule” and “rule of reason” when enforcing
decided that the Grab-Uber merger was anti- their laws. And, all four countries assume that
competitive, while others such as Indonesia and a firm has a dominant position if the firm's
Viet Nam considered the merger not anti-com- market share is above a certain threshold and
petitive. explicitly regulate the abuse of dominance.

We review the competition policies and laws The requirements for reporting mergers and
of four major ASEAN countries – Indonesia, acquisitions show a wide spectrum. Table 1
Singapore, Viet Nam, and the Philippines – compares the institutional divergence among
from institutional and legal perspectives, focus- the four states. Singapore has a post-review
ing on M&A review regimes. Then, we briefly process where competition law requires firms
introduce how these competition authorities de- to report the M&As afterwards but they do not
cided on the Grab-Uber merger case, also ana- have to report consolidations in advance.
lyzing the competition effects of the case on the However, Viet Nam and the Philippines have
ride-hailing market in the countries. Based on a mandatory pre-review process where merg-
the analysis results, we propose overseas com- ing firms are obligated to both report M&As
petition policies for Korea. to the authorities and obtain approval in ad-
 vance. In between the two regimes, Indonesia
 runs voluntary pre-review and mandatory
II. Merger Review Regimes in post-review regimes. Also, these countries
 the ASEAN Region have substantive differences in assessing the
 anti-competitive effects of mergers as well as
While Indonesia was the first country to intro- procedural differences in the review periods
duce a competition law in ASEAN in 1999, and thresholds for notification.
Singapore and the Philippines are equipped
with relatively advanced forms of competition These institutional differences can impose
policy. Viet Nam started to accept global stand- burden and cost on firms conducting cross-
ards since the comprehensive amendment of its border M&As in the ASEAN region. That is,
competition law in 2018. the divergence by states within ASEAN can
 disturb market integration by increasing com-
Compared to the regulation of cartel conduct petition law risks in the region.

 2

 Merger Review Regimes in the ASEAN Region and Case Analysis of Grab-Uber Merger

September 8, 2021

 Table 1. Merger Review Regimes: Indonesia, Singapore, Viet Nam, and the Philippines
 Indonesia Singapore Viet Nam Philippines
 Chapter 5, Law No. 5 Competition Act (Chap- Law No. 23/2018/QH Philippines Competi-
 /1999 ter 50B) 14 tion Act
 Government Regula- CCCS Guidelines on Decree No. Rules of Procedure of
 tion No. 57/2010 merger procedures 35/2020/ND-CP the Philippine Compe-
 KPPU Regulation 2012 Decree 75/2019/NC- tition Commission
Competition
 No.3/ 2019 CCCS Guidelines on CP Rules and Regulations
 Law the substantive as- to Implement the Pro-
 sessment of mergers vision of Republic Act
 2016 No. 10667
 Merger Review Guide-
 lines
 Voluntary pre-merger Voluntary post-merger Mandatory pre-merger Mandatory pre-merger
Notification consultation and Man- notification notification notification
 datory post-merger no-
 Regime tification
 Turnover-based No compulsory thresh- Total domestic assets Basically depending
 threshold of IDR 5 tril- old for pre-notification or sales-based thresh- on nominal GDP
 lion (approx. USD 340 old of VND 3 trillion
 million); or However, firms are en- (approx. USD 129 mil- As of March 2020,
 Asset-based threshold couraged to notify if lion); or Turnover or asset-
 of IDR 2.5 trillion (ap- 40% of combined mar- Transaction value- based threshold of
 prox. USD 170 million) ket share (MS) or based threshold of PHP 6 billion (approx.
 more; or VND 1 trillion (approx. USD 120 million); or
Thresholds a 20–40% combined 43 million USD); or Transaction value-
 MS and70% of CR3 or 20% of combined MS based threshold of
 for more Financial sector-spe- PHP 2.4 billion (ap-
 cific thresholds prox. USD 48 million)
Notification
 *Due to Byanihan Act 2
 to recover from COVID-
 19, transaction values
 below PHP 50 billion are
 exempted from compul-
 sory notification for two
 years from 15 September
 2020
 In voluntary pre-mer- Phase I with maximum Phase I with maximum Phase I with maximum
 ger consultation, 30 working days after 30 working days after 30 working days after
 Phase I (preliminary) complete document complete document complete document
 with maximum 30 submission and Phase submission, Phase II submission and Phase
 working days and II with maximum 120 with maximum 90 II with maximum 60
 Phase II (comprehen- working days days, and extended days
 sive) with maximum Phase II with addi-
 Review 60 working days tional 60 days

 Period In mandatory post-
 merger notification,
 only one Phase with
 no longer than 90
 working days after
 complete document
 submission

 3

 Merger Review Regimes in the ASEAN Region and Case Analysis of Grab-Uber Merger

September 8, 2021

 Indonesia Singapore Viet Nam Philippines
 Market concentration Unilateral effects, co- The 2018 new compe- Unilateral effects, co-
 (HHI), entry barriers, ordinated effects and tition law introduced ordinated effects and
 possibility of anti-com- foreclosure effects global standards as- foreclosure effects
 petitive behaviors analysis sessing the merger’s analysis
 post-merger, efficiency anti-competitive ef-
 effects, failing firms Justification test with fects Justification test with
 (bankruptcy), etc. efficiency in demand efficiency effects
 and supply sides, and *Before 2018, mer-
 dynamic efficiency gers generating 50% MS
 Substantive were not allowed
 Assessment Assumed not anti-
 competitive if com- Mergers related to na-
 bined MS is below tional strategic indus-
 40%, or 20–40% with tries, growth of SMEs,
 less than 70% CR3 and promoting com-
 petitiveness of domes-
 tic firms are positively
 assessed to be ap-
 proved
 Corrective orders with Corrective orders with Corrective orders with Corrective orders with
 behavioral and struc- behavioral and struc- behavioral and struc- behavioral and struc-
 tural remedies tural (preferred) reme- tural remedies tural remedies
 dies
 IDR 1 billion of fines Financial penalty not Financial penalty be-
 Remedy everyday (not exceed- Interim measures exceeding 5% of an- tween 1–5% of trans-
 ing IDR 25 billion) nual domestic sales action amount
 against failing to sat- Financial penalty not amount
 isfy filing deadlines exceeding 10% of an- Consent order
 nual domestic sales
 amount

III. Analysis on the Impacts of that the Grab-Uber M&A was anti-competi-
 Grab-Uber M&A on Market tive. They argued that the consolidated Grab’s
 Competition market power after the merger would be
 strengthened due to the elimination of its
 strong competitor, Uber. The merger had the
 1. Competition authorities’ effect of easing intense competition pressure
 decisions in the digital platform market. Nonetheless,
 due to the institutional limitations of the vol-
We compare each of the four competition au- untary post-review regime in the country, the
thorities’ decisions on whether to approve a merger was approved by Singapore’s compe-
representative cross-border M&A case in the tition authority with certain behavioral reme-
region in 2018 – the Grab-Uber M&A case. dies, such as restrictions against raising prices.
The competition authorities of Singapore and The Philippines’ competition authority ap-
the Philippines both independently decided proved the case as well, but chose a slightly

 4

 Merger Review Regimes in the ASEAN Region and Case Analysis of Grab-Uber Merger

September 8, 2021

different method from Singapore’s by finaliz- provided by a specific platform, user ratings,
ing it by way of consent order. the number of daily passengers for the plat-

In Viet Nam, the Vietnam Competition and form, and the average waiting time for the
Consumer Authority (VCCA) delivered its in- platform) in the estimation procedure. With
itial opinion that the M&A should not be ap- these features of the platform, we combine a
proved due to the potential competition re- generalized method of moments (GMM) with
strictive effects. However the Vietnam Com- instrumental variable (IV) estimation as the
petition Council (VCC), as the final decision empirical methodology. The benchmark
commission, chose not to accept the proposal model of the demand estimation is the follow-
and approved the merger without any condi- ing.
tions. The Indonesian competition authority
did not apply the competition law, arguing that = − + + , + 
the merger did not show any changes of con-
trol rights regulated in the law and only was
 where represents a passenger i’s utility
considered as sales of an asset.
 level obtained by utilizing a ride-hailing plat-
 form j in a market m, describes unob-
 2. Empirical Analysis servable characteristics of the platform j in the
 market m, indicates a price of the service
We empirically analyze the economic effects provided by the platform j and represents
of the Grab-Uber M&A case on market com- observed platform characteristics varying with
petition using a data set provided by Allied distributions of unobserved consumer charac-
Market Research. The data set includes infor- teristics ( , ). indicates an unobserved
mation on the features of consumers and three error term. Table 2 shows the empirical results.
ride-hailing applications, Grab, Uber, and Go- Price is negatively associated with the ride-
jek, from 2008 to 2019. The empirical results hailing service demand in every column. In
show that the anti-competitive effects from the addition, the second and the third column
merger are weaker in Indonesia where the con- show that the number of different types of
solidated Grab still has a strong competitor, platforms provided by an individual platform
Go-jek. has positive impacts on consumer demand at
 the significance levels of 1% and 0.1%, re-
We utilize a structural econometric model to spectively. As predicted above, except for the
estimate the demand and supply function for second column, user ratings have positive ef-
the ride-hailing service following Berry, Lev- fects on consumer demand, but are not statis-
insohn, and Pakes (1995). For the purpose of tically significant. Lastly, we need to note that
estimating the market demand function, we the effects of the average waiting time of ride-
use four characteristics (diversity of services hailing platform users on demand is different
 5

 Merger Review Regimes in the ASEAN Region and Case Analysis of Grab-Uber Merger

September 8, 2021

from our expectation. Previously, we pre- changes in market share due to changes in the
dicted that consumer demand for the platform price of specific platform services from 2013
would decrease as users wait longer after hail- to 2016, before Grab’s acquisition of Uber’s
ing a vehicle. However, according to the sec- Southeast Asian business in 2018, Go-jek’s
ond and third columns of Table 2, we find that market dominance is gradually weakening, af-
the average waiting time has positive impacts ter enjoying a monopolistic position in the ex-
on consumer demand, and the estimated coef- isting market. For example, when Go-jek low-
ficients are also statistically significant. As ers its service price by 1%, the increase in Go-
Go-jek established itself early in the Indone- jek's market share is 0.530%p (2013),
sian ride-hailing platform markets, it can be 0.513%p (2014), 0.455%p (2015), 0.441%p
estimated that a lock-in effect has occurred for (2016), showing a decreasing trend. Regard-
users who are accustomed to the Go-jek ser-
 ing this result, we can conjecture as follows.
vice.
 Go-jek was the first to enter the Indonesian
 ride-hailing platform markets and held its po-
We also calculate the own- and cross-demand
 sition as a monopolist. However, as new com-
elasticities in response to changes in charac-
 peting platforms such as Uber and Grab en-
teristics of ride-hailing platforms. Using the
 tered, existing Go-jek users began to experi-
estimated coefficient indicating the effect of
 ence multi-homing or completely changing
the price of the ride-hailing platform service
 the platform being used. This is in line with the
on the average utility of consumers, we can
 fact that Uber's own-demand elasticities have
find the change in demand for a given platform
 been increasing until 2016.
(own-demand elasticities) and change in de-
mand for a competitive platform (cross-de- The second characteristic can be derived from
mand elasticities), according to changes in the the own-demand elasticities of Uber and Grab.
price of a specific platform service. Here, we In other words, when both companies lowered
use the change in market share of each plat- their service prices by 1%, Uber's market share
form as the proxy variable representing the growth showed a large increase up to 2016,
change in quantity demanded for the platform while Grab's share decreased significantly
following the existing literature. Table 3 from 2016 (2015: 0.402%p, 2016: 0.297%p).
shows the calculation results of the own- and These results may have had an impact on the
cross-demand elasticities from 2013 to 2019 in merger between the two platforms in order to
the Indonesian ride-hailing platform markets. stop competition in the same market and in-
With the results, we can summarize some crease efficiency, such as combining customer
characteristics observed in the Indonesian information and sub-services provided by
ride-hailing platform markets. First, looking at each platform.

 6

 Merger Review Regimes in the ASEAN Region and Case Analysis of Grab-Uber Merger

September 8, 2021

Finally, after Grab's acquisition of Uber's need to show whether competition between
Southeast Asian business, we find that the platforms is increasing after Grab's takeover.
competition between Go-jek and Grab has Table 4 shows the marginal cost and markup
been increasing. However, in a situation where calculated using the solution derived from the
Grab's market share is still lower than that of profit maximization problem of each ride-hail-
Go-jek, we find that the effect of the service ing platform. Some characteristics that can be
price change on each platform on the market drawn from the results are as follows. First, as
share change of both platforms is still asym- shown in Figure 1, Uber's marginal cost has
metrical. For example, when Grab lowered the been increasing, while Grab's marginal cost
price of its services by 1% in 2017, Go-jek's has been decreasing. Moreover, Uber’s
market share fell by 0.040%p. However, in markup, which increased until 2015, turned to
2018 and 2019, after Grab acquired Uber, the a decreasing trend starting in 2016, while
same rate of change in Grab's prices leads to Grab's markup has steadily increased (see Fig-
decrease in Go-jek's market share by 0.063%p ure 2). Based on this, it can be inferred that the
and 0.058%p, respectively. Also, when Go-jek deterioration of the business environment in
lowered its service price by 1%, we find that the region due to the decrease in Uber's
Grab's market share in 2018 and 2019 also de- markup had an impact on Grab's acquisition of
creased by 0.278%p and 0.275%p. However, Uber's Southeast Asian business, which had a
as pointed out above, the fact that there is a trend of increasing markup.
large difference in market share changes be-
tween the two platforms in response to Second, we find that Go-jek, which had a high
changes in the service price of individual plat- market share, continues to experience a rise in
forms is still a result of Go-jek's dominance in marginal cost and drop in markup in a situa-
the Indonesian ride-hailing platform markets. tion where the platforms expand their market
 share after Uber and Grab enter the market. In
Now, let's look at the impact of Grab's acqui- addition, Go-jek's markup decreased less than
sition of Uber's Southeast Asian business on before the acquisition of Grab's Uber business,
competition in the Indonesian ride-hailing but continues to decrease after the acquisition.
platform markets. To this end, we utilize a spe- On the other hand, we find that Grab is gradu-
cific estimation method first assuming the ally increasing its markup after acquiring Uber.
competition model of the Indonesian ride-hail- These results can be seen as evidence that
ing platform markets, solving the profit maxi- Grab's acquisition did not adversely affect
mization problem of individual platform com- competition by forming an oligopoly, but ra-
panies to derive a new competitive equilib- ther promoted competition in the form of
rium after the acquisition, and calculating the checking Go-jek, the existing dominant mar-
marginal cost for each platform. Here, we ket operator.

 7

 Merger Review Regimes in the ASEAN Region and Case Analysis of Grab-Uber Merger

September 8, 2021

 Table 2. Results of GMM-IV Estimation for Ride-hailing Service Demand
 Explanatory
 (1) (2) (3)
 Variable
 -5.750* -4.717* -4.749*
 α Price
 (2.756) (1.720) (2.026)
 The number of Platforms
 -0.123 0.663** 0.681***
 provided by the same
 (0.347) (0.256) (0.246)
 ride-hailing platform
 0.034 -0.085 0.008
 User ratings
 (7.233) (4.109) (2.827)
 
 The average number of daily 0.002
 0.017
 passengers for a platform (8.853)
 (1.119)
 0.433*** 0.449†
 The average waiting time
 (0.107) (0.246)
 Platform dummy variable ○ ○ ○
 Minimum value of
 1.6014 1.0626 1.0512
 GMM objective function
 Observations 19 19 19
Notes: 1) Standard errors are in parentheses.
 2) Significance level: † p

September 8, 2021

 Table 4. Marginal Cost and Markup of Ride-hailing Platform Service

 Ride-hailing
Year Marginal Cost Pre-M&A Price Mark-ups
 Platform
 Uber -0.01048399 0.210 0.2204840
2013 Grab 0.25403876 0.470 0.2159612
 Go-jek -1.07968508 0.675 1.7546851
 Uber 0.13335557 0.355 0.2216444
2014 Grab 0.25180082 0.470 0.2181992
 Go-jek -0.82901578 0.675 1.5040158
 Uber 0.13218285 0.355 0.2228172
2015 Grab 0.25066437 0.470 0.2193356
 Go-jek -0.72874806 0.675 1.4037481
 Uber 0.15451601 0.375 0.2204840
2016 Grab 0.12335557 0.345 0.2216444
 Go-jek -0.56360123 0.675 1.2386012
 Uber 0.15566437 0.375 0.2193356
2017 Grab 0.11612803 0.345 0.2288720
 Go-jek -0.46317411 0.675 1.1381741
 Ride-hailing
Year Marginal Cost Post-M&A Price Mark-ups
 Platform
 Grab 0.09433070 0.345 0.2506693
2018
 Go-jek -0.32767719 0.675 1.0026772
 Grab 0.09131059 0.345 0.2536894
2019
 Go-jek -0.24048787 0.675 0.9154879
Source: Author’s estimation based on AMR Database (2020).

 Figure 1. Changes in Marginal Costs

 Source: Author’s estimation based on AMR Database (2020).

 9

 Merger Review Regimes in the ASEAN Region and Case Analysis of Grab-Uber Merger

September 8, 2021

 Figure 2. Changes in Markup

 Source: Author’s estimation based on AMR Database (2020)

IV. Conclusion tion with the ASEAN member states and mini-
 mizing competition law risks of Korean compa-
Considering the legal and institutional gaps nies that are conducting or planning to do busi-
across countries and divide in their capacity to ness in the region. First, in line with the New
enforce competition laws, it is important to syn- Southern Policy, which emphasizes multilateral
chronize the ASEAN member state’s regimes by cooperation with ASEAN, Korea’s competition
reducing the heterogeneity in competition policy authority should build a multilateral cooperation
within the region. Also, capacity-building pro- channel with the ASEAN Experts Group on
grams to enhance enforcement skills should be Competition as well as develop bilateral cooper-
addressed. In particular, due to the complicated ation with an individual member state. Second,
nature and convergence features of the digital the authority should promote demand-based and
economy, competition authorities in ASEAN customized collaborative projects in the areas of
should consider a new regulation paradigm in competition policy through joint investigation,

competition policy more suitable to the changing research, and sharing best practices in the digital

digital competition environments. platform economy. This cooperation could in-
 clude capacity-building programs on cultivating
 competition environments and enforcing compe-
On this backdrop, we provide the Korea com-
 tition laws. Third, Korea’s competition authority
petition authority with four policy suggestions in
 should establish a cooperation network with the
the context of promoting international coopera-
 ASEAN Competition Enforcer’s Network to

 10

 Merger Review Regimes in the ASEAN Region and Case Analysis of Grab-Uber Merger

September 8, 2021

prepare for the increasing demand of cross-bor- Korean companies with more detailed infor-
der competition law enforcement in the digital mation on the local competition law and the au-
economy. Finally, considering the increasing thorities’ enforcement standards to minimize
trend of business opportunities in the ASEAN competition law risks by helping them avoid
region, the Korean government should provide violating the local competition laws.

References
Jang, Yungshin, Gusang Kang, Seung Kwon Na, Jegook Kim, Jaypil Choi, and Suruyn Kim. 2020.
 Competition Policy and Law in the ASEAN Countries: Focusing on Digital Platform M&A.
 KIEP Studies in Global and Regional Strategies, no. 20-09 (in Korean).

 11

 Merger Review Regimes in the ASEAN Region and Case Analysis of Grab-Uber Merger