This presentation by UK’s Competition and Markets Authority was delivered during a workshop on “Methodologies to measure market competition” held virtually for competition authorities officials on 23 February 2021. More materials on the topic can be found at http://oe.cd/mmkts.
This presentation was uploaded with the author’s consent.
The 3rd Intl. Workshop on NL-based Software Engineering
Methodologies to measure market competition – UK CMA – Feb 2021 OECD Workshop
1. The State of UK Competition
OECD Methodologies to measure market competition
Richard Havell
Federico Mor
Callum Miller
1
2. Introduction
● We estimated a range of metrics for the UK:
- Concentration (structure of industries, dominance of large firms, partial ownership)
- Dynamic competition (rates of entry and exit, stability of firms’ positions in the economy)
- Profitability and mark-ups (levels, distribution and persistence of profits and mark-ups)
- Consumer surveys (measures of trust and satisfaction with businesses)
- Real time indicators including business formation and closure data
● Main findings tend to show a long-term weakening in competition:
- Concentration rose across the economy following the 2008 financial crash – while recovery
slightly it remains higher than before the crash
- The number of “younger” large firms has increased – suggesting some entry and exit – but churn
at the very top remains lower than 20 years ago
- Average mark-ups have risen by 7% over the last 20 years and there has been some small
increase in EBIT margins for the most profitable firms. Return on capital employed seems to
have fallen
2
4. Concentration results – Economy-wide
● Economy-wide concentration shows
an increase in 2008/09 following the
financial crisis, following by a gradual
decrease since 2013/14
- Weighting by Gross Value Added rather
than total turnover shows flatter trend
● Concentration in 2018 is still 3
percentage points higher than it was in
1998
● These findings are in line with previous
work on UK concentration
4
Average C10, whole economy
Finance and Fuel wholesaling are excluded from the whole-economy
figures as they would otherwise have disproportionate weight
5. Concentration results – Sector level
● Sector level results mirror economy-wide
trend
● Finance and insurance, and wholesale
and retail trade both saw concentration
increase, especially in the years leading
up to 2010
● Manufacturing also has an upwards trend
over the period
- However, many manufacturing industries have
high and increasing levels of imports. This may
alleviate competition concerns in these
industries
5
Average C10, higher-revenue sectors
These sectors account for 86% of total turnover on the BSD in 2018
6. Entry and exit rates
● Entry rate is calculated by dividing the
number of new firms in each year by the
total number of active firms in that year
- Exit rate is calculated likewise, using the
number of exiting firms each year
● Main trend is cyclical – the financial
crisis caused an increase in exit rate
and a decrease in entry rate
● As entry and exit rates include all
businesses and weight them equally,
they are likely to dominated by smaller
firms
6
Entry and exit rates, whole economy
7. Rank persistence
● The assailability of the positions of
the very largest firms in the
economy is an important part of a
competitive economy
● This metric identifies the ten largest
firms in each sector of the economy,
and counts the number of them
which were also in the ten largest
firms three years ago
● Trend shows a move to a high
plateau of stability at the top of all
large sectors since 2008
7
Rank persistence, higher-revenue sectors
9. Theory and evidence
● Ownership links can reduce incentives to compete.
● They can also facilitate collusion by facilitating communication between
firms, increasing incentives to coordinate and making collusion more stable.
● Various academic papers found that common shareholdings lead to
higher prices or other outcomes that suggest a reduction in competition.
● No systematic, economy-wide assessment of ownership links in the UK.
● Existing evidence focuses largely on listed firms, with little research
on private companies. For example, in the US, the combined stake of
BlackRock, Vanguard and State Street in S&P500 stocks is larger than that
of any other shareholder in 438 (88%) of firms.
9
10. CMA analysis
● We used data from company filings to gauge the potential extent of
ownership links not captured in the ONS data used to calculate
concentration.
● Our own analysis is a first attempt at identifying ownership links
among private companies across the economy.
● The links identified appear to decrease the number of fully
independent competitors by 42,000 (out of 2.6 million enterprise groups).
While not material at the whole economy level, the impact on concentration
in a minority of industries is significant.
● For example, the number of fully independent groups in the ‘retail sale of
hearing aids’ drops by nearly a third. Of 672 industries with 10 or
more enterprises active in them, 35 register a drop of 3% or more.
10
12. Measures
● A firm’s mark-up (the ratio of price to marginal cost) is taken by economists
to be a measure of its market power.
● Mark-ups have to be estimated and there is an active academic debate
about how best to do this.
● Profitability is directly observable and measures of it are used by the CMA to
assess competition in specific markets, albeit after making case-specific
adjustments that cannot be replicated in a whole-economy analysis.
● The two metrics we have used for this analysis are:
- EBIT margin = (Earnings Before Interest and Tax) / Turnover
- ROCE = (Earnings Before Interest and Tax) / Capital Employed
12
13. Existing literature
● The majority of the literature has focused on mark-ups. Generally these papers find:
- A trend of increasing mark-ups either globally, in advanced economies or in
specific countries including the UK (e.g. Aquilante et al. (2019)).
- That the increase in mark-ups was concentrated among firms in the upper part
of the mark-up distribution.
● A number of papers complement these results with an assessment of profitability
because mark-ups miss fixed costs. Therefore, markups may need to rise to cover
fixed costs. On the other hand, an upward trend in both mark-ups and profitability
is more likely to indicate a reduction in competition.
● As findings on profitability are mixed, there is some debate about how this should
all be interpreted, and the relative contribution of different hypotheses:
- Technological changes have increased fixed costs, which need to be covered by higher
mark-ups.
- Global ‘superstar’ firms have higher mark-ups because they are more efficient (e.g. due to
global supply chains and economies of scale).
- Competition has decreased. 13
14. CMA Analysis: Trends – mark-ups
● We find that:
- Average mark-ups have
increased by 7%.
- This is concentrated in firms in
the upper part of the
distribution: 9% increase at the
90th percentile.
● Consistent with past work.
● Further work is needed to test
these results using different
approaches.
14
15. Trends – EBIT margin
● Broadly flat trend over time.
● At the 90th percentile there has
been a small upwards trend since
2008, albeit within the historical
range of our data.
15
16. Trends – ROCE
● We have seen a downward trend in ROCE
at all levels of the distribution since 2007.
● However, the cost of debt (which we use as
a proxy for the cost of capital) has also
decreased over the same period.
● So it is not clear if the fall in ROCE reflects
stronger competition or cheaper capital.
Cheaper capital may mean that the actual
economic profits earned by firms have not
in fact fallen.
● More work is needed to reach firmer
conclusions.
16
● In theory, ROCE may be the 'best' measure because it takes into account both fixed cost
and capital intensity of firms. But harder to intepret in practice, as our analysis shows.
17. Trends – Conclusion
● We have rising mark-ups, mostly flat margins and an inconclusive ROCE
picture.
● Overall, we take our analysis as tentative evidence that there may have
been a weakening of competition at the top, where the rise in mark-ups is
accompanied by a gentle rise in margins too.
● However, since margins are broadly flat overall, the fixed costs story may
play a role too.
17
18. Persistence - results
● Persistence seems to have
increased overall.
● This suggests there is less scope for
firms to break into (or be displaced
out of) more profitable positions.
● In terms of the absolute level, ROCE
is the better indicator, as it is the
most comparable across different
firms.
18
The proportion of companies in the top decile that were in
the top decile three years before
20. Key findings (1)
● Assessing the outcomes
consumers and businesses
believe markets are delivering
for them. Structured our search
considering: choice and
shopping around, switching,
satisfaction, trust and consumer
problems.
● UK sits in 11th place out of 30
European states included in the
study, below Germany (fifth) and
France (sixth).
● UK goods markets consistently
outperform UK services markets
whether comparing markets
directly to each other within the
UK or to the EU average.
20
Source: CMA analysis of European Consumers Markets
Scoreboard (2018).
21. Key findings (2)
● The UK’s better performance in goods markets is linked to stronger scores
on: consumer trust in providers to respect consumer rights, satisfaction with
the choice of providers in markets and how far markets have met consumer
expectations.
● Multiple surveys show UK consumers experience a relatively high level of
consumer problems worthy of complaint (for service markets in particular).
● Surveys show consumers who are less financially secure and/or on lower
incomes tend to have a poorer experience of a wide range of markets.
● Transport, utilities and telecommunications/media consistently feature lower
down the rankings when measuring different consumer and business
outcomes.
21
23. ● Wave of business closures we might have expected has not been seen.
● Figure 1 shows fewer closures in Q2 and Q3 v 2019, and figure 2 shows majority saw no change in
number of competitors.
23
Figure 1: Year-on-year percentage change in total business closures
-30%
-20%
-10%
0%
10%
20%
30%
40%
50%
Q1 2018 Q2 2018 Q3 2018 Q4 2018 Q1 2019 Q2 2019 Q3 2019 Q4 2019 Q1 2020 Q2 2020 Q3 2020
UK
Source: Office for National Statistics, Business demography, quarterly experimental statistics, UK: July to September 2020.
Figure 2: Number of businesses selling similar products/services
Has the number of businesses selling similar goods or services as your businesses
changed since the start of the coronavirus (COVID-19) pandemic?“ Source: BICS survey
(wave 12) August 2020. The BICS survey is unweighted and, rather than being nationally
representative, may only reflect the characteristics of those who responded
Summary of findings (1)
24. ● Some signs that businesses being
created following the first wave of the
pandemic are smaller than the historical
average and more likely to be in
industries less affected by the current
pandemic. (Not yet seen a significant
impact on total no. creations)
● Clear survey evidence of an impact on
expansion plans (Figure 3)
- Four in ten businesses reported having had
to postpone or scale back expansion plans
as a result of the pandemic
24
Figure 3: Changes to expansion plans following the pandemic (by wave 12
and wave 13 of the BICS survey)
Has the (COVID-19) pandemic affected your business's plans to expand the business?
Responses of ‘Not Sure’ are excluded
Source: BICS survey, Waves 12 and 13, reporting period August and September 2020
Note: The BICS survey is unweighted and, rather than being nationally representative, may only reflect the
characteristics of those who responded.
.
Summary of findings (2)
25. ● We found evidence that since the
pandemic outbreak, consumers had
shopped around less than normal
- Over 40% said they shopped around a lot
or a little less than normal
- This was true for both food and toiletries,
as well as non-essentials
● Concerning trends surrounding who
shopped around less – those with
activity-limiting health conditions and
elderly less likely to shop around
- This could be linked with eg shielding
- Reasons given for shopping around less
were mostly (hopefully) temporary eg
safety concerns and practical issues of
how to compare quality
25
Figure 4: Percentage shopping around more or less than before – by
activity limiting health problems or disability
Source: CMA analysis of OPN survey (Waves W and X), Fieldwork dates August and September
2020
Base: All respondents. Base size: N = 3,176 (Food/Toiletries), N =3,066 (non-essential items)
Summary of findings (3)