How Did MiFID Affect Systemic Underperformance in the European

Equity Market? A Tracking Error based Approach

Cheikh Niang and Alois Kanyinda-Kasanda

Department of Finance, Neoma Business School, 59 Rue Pierre Taittinger, Reims, France

Keywords: MiFID, Tracking Error, Market Quality, Market Efficiency, Benchmark.

Abstract: Using a new methodology based on the decomposition of tracking error, we show that the implementation of

MiFID improved the quality of European equity markets. The latter was favorable to the European equity

investors’ ability to reach their investment objectives as measured by the systemic downside tracking error.

In other words, after the implementation of the MiFID directive, investors were less likely to underperform

due to unfavorable market characteristics. The results were statistically significant at the 95% significance

level.

1 INTRODUCTION

The implementation of MiFID was supposed to bring

a higher level of market quality through improved

liquidity and higher efficiency by fostering trade

transparency and competition between trading

venues. The implementation of MiFID did foster

competition; however, it seems many of its negative

aspects were unanticipated.

Gomber and Pierron (2010) show that trading

activity reported as OTC activity is very different

from its description in the MiFID. MiFID

characterizes OTC as transactions that cumulatively

fulfill the requirements of being ad hoc and irregular,

carried out with wholesale counterparties, above

standard market size, and conducted outside systems

used for systematic internalization. However, their

results show that a significant share of OTC

transactions are neither above standard market sizes

nor would they face market impact if concluded on

open, public order books. The authors also show that

the adoption of new trading technologies has

dramatically increased the sensitivity of market data.

The reduction of average transaction sizes in the

various liquidity pools and the implementation of

high-frequency trading have reinforced the

willingness of buy-side firms to hide their trading

strategy by limiting information leakage while

capturing as much information about the trading

patterns of their counterparts as possible. This

situation conjugated with the desire to decrease

execution cost explains the rise of dark pools in the

European market.

Degryse et al., (2015), for instance, show that

fragmentation is beneficial in visible order books

through improved global liquidity, whereas the effect

of dark trading is detrimental.

Buti et al., (2011) show that the existence of dark

pools in illiquid markets tends to widen bid-ask

spreads, decrease the market depth, and deteriorate

overall welfare. In more liquid markets, only large

traders see their situation improve while small traders

are still worse off.

Our study aims to investigate whether these

adverse developments after the implementation of the

MiFID did have a negative impact or not on market

participants’ ability to reach their investment goals in

the European equity market. In other words, how did

the MiFID affect the likeliness of fund managers

displaying underperformance only due to market

factors?

In section 1, we review the related literature. We

dedicated Part 2 to the explanation of the

methodology. Section 3 presents the data used for the

study. Section 4 presents the statistical analysis, and

we comment on the results in part 5.

2 LITERATURE REVIEW

Academic research has widely documented

underperformance by equity mutual funds. Many

Niang, C. and Kanyinda-Kasanda, A.

How Did MiFID Affect Systemic Underperformance in the European Equity Market? A Tracking Error based Approach.

DOI: 10.5220/0008772000570062

In Proceedings of the 2nd International Conference on Finance, Economics, Management and IT Business (FEMIB 2020), pages 57-62

ISBN: 978-989-758-422-0

factors are significant in explaining portfolio

underperformance.

Day et al. focused on the impact of portfolio

composition and the excess turnover on fund

performance. Using standard portfolio optimization

techniques, they showed that the portfolio weights for

the stocks selected by fund managers are, on average

inefficient. They suggest that while fund managers

may possess superior stock selection skills, we could

achieve substantial gains by improving the efficiency

of the allocation of mutual fund assets. They also

present evidence suggesting that mutual fund

turnover is excessive and that fund managers may

rely too heavily on stock price momentum.

Cremers and Pareek (2015) confirmed these

findings. They show that among high active Share

portfolios, only those with patient investment

strategies (withholding durations of over two years),

on average, outperform, over 2% per year. Funds

trading frequently generally underperform, including

those with high Active Share.

Gastineau (2004) claims that ETFs underperform

their index fund competitors. Specifically, Gastineau

suggests that, at least in part, the deficiency in an

ETF’s underperformance is due to the ETF managers’

reluctance to adjust the portfolio before the official

moment of the index adjustment.

Hu et al., (2008) found that a fund’s performance

is negatively related to its age. Blitz et al. (2012) find

the explanatory power of dividend withholding taxes

for fund underperformance relative to its benchmark

to be at least on par with fund expenses. Applying

these findings, Blitz and Huij (2012) show that

emerging market equity ETFs’ expected returns are

equal to their respective gross benchmark index

returns minus expense ratio and dividend taxes.

Charupat and Miu (2013) find that all other things

being equal, the higher the expense ratio of a fund, the

more an ETF can be expected to underperform its

underlying index.). Many articles support this view:

Elton et al. (2002), Lin and Chou (2006), Rompotis

(2006, 2011), Agapova (2011), and Blitz et al. (2012).

Plyakha et al. (2015) find that the weighting

scheme is an essential factor in explaining fund

performance. They find that, on average, value-

weighted funds tend to underperform when compared

with equal-weighted funds. They find an excess

yearly mean return of 2.71% for the equal-weighted

portfolio over the value-weighted portfolio.

According to their findings, 58% of the excess return

comes from the excess systematic component, while

42% comes from the difference in alphas.

Additionally, the higher systematic return of the

equal-weighted portfolio arises from its higher

exposure to market, size, and value factors, which is

determined by the equal initial weights. However, the

higher alpha of the equal-weighted portfolio arises

from the monthly rebalancing to maintain constant

loads.

Our work will have as an objective to assess the

impact of MiFID through its possible negative impact

on the systemic component, as described by Plykha et

al.

In fact, since the implementation of the MiFID

Directive, we observed many possible adverse

developments. Some believe that the implementation

of the MiFID has exacerbated opacity in the financial

markets. This opacity materializes in the growth of

dark pools in Europe. Moreover, it fostered

information asymmetries between different market

participants and, most notably, between high-

frequency traders and low-frequency traders (Lenglet

and Riva, 2013).

These developments put into question the

integrity of financial markets and more to the point;

they cast doubt on the information content of stock

prices in the markets where the MiFID directive

applies. Many facts over the past years are suggesting

market manipulation. For instance, the French

financial market authority revealed that the execution

of orders in Europe is at an interval between 1% and

5%.

While one of the primary objectives of MiFID

implementation is to increase transparency,

Bloomfield and O'Hara showed that there are no

discernible effects of transparency on the market

performance, based on a simulation of three markets

that have different transparency levels. In the same

way, Porter and Weaver (2005) examined the effect

of changes in information disclosure rules on the

Toronto Stock Exchange. By making comparisons in

market performances before and after the reform,

they found that increased pre-trade transparency

decreased liquidity, increased execution costs, and

market volatility.

By using a new approach, our paper will assess

equity investor’s likeliness to underperform relative

to their benchmarks due to unfavorable systemic

factors induced by the implementation of the MiFID.

We present our methodology in the following section.

3 METHODOLOGY

In our study, we would like to determine the effect of

the implementation of MiFID on portfolio managers’

ability to reach their investment objectives, through

the impact of MiFID on market quality.

FEMIB 2020 - 2nd International Conference on Finance, Economics, Management and IT Business

Tracking error is the most prominent metric to

measure the deviation of portfolios from there

announced benchmark.

You can find here below the equation of tracking

error.

𝑇𝐸



=𝑅



−𝑅



(1)

Where:

 TEi is the Tracking Error for the period i

 RPi is the return of the portfolio for the period i

 RBi is the return of the benchmark for the period i

The tracking error of a given portfolio is

influenced both by idiosyncratic factors (number of

stocks in the portfolio, number of non-benchmark

stocks in the portfolio, the degree of leverage) and

non-idiosyncratic factors (most notably volatility of

the benchmark and market quality) (Vardharaj et al.,

2004).

Larsen and Resnick (1998) show that the tracking

error is affected by market capitalization. Large

capitalization portfolios have lower volatility and

tracking error than low capitalization portfolios.

Frino and Gallagher (2001) identify the expenses,

the dividend payments arising from the underlying

stocks that compose an index, and the timing of index

rebalancing as being factors affecting the size of

tracking error.

According to Kostovetsky (2003), the tracking

error of index funds is affected by the bid-ask spreads

of the portfolio's underlying stocks, the dividend

distribution policies, and the transaction costs.

Osterhoff and Kaserer (2015) find that daily

tracking error significantly depends on the liquidity

of the underlying stocks.

This result confirms the findings of several studies

suggesting a positive effect of spreads on tracking

error. For instance, Milonas and Rompotis (2006),

Delcoure, and Zhong (2007) all verify that a fund’s

tracking error is positively affected by the bid-ask

spread.

Frino et al. (2004) use monthly data for the years

1994-1999 and show that tracking error in index

mutual funds for the S&P 500 is significantly related

to index revisions, share issuances, spin-offs, share

repurchases, index replication strategy, and fund size.

Gastineau (2002) finds for equity index funds

tracking the Russell 2000 and S&P 500 indices that

changes in index composition have a significant

effect on tracking error due to the transaction cost

involved in the necessary rebalancing of the

underlying portfolio.

Kundisch and Klein (2009) observe the daily

returns and tracking ability of several DAX

certificates and one DAX ETF for the period 2001-

2006 and show that the trading volume of the ETF

negatively correlates with its tracking error.

Elton et al. presented the non-reinvestment of

dividends as a significant factor affecting the tracking

error of SPDRs.

The contribution of each factor to tracking error is

called the “Marginal Contribution to Tracking Error”

(MCTE).

In our model, we hypothesize that the effect of the

MiFID on the investors’ ability to match their

benchmark is included in the Residual MCTE

(RMCTE). RMCTE is the MCTE once we consider

the MCTE of all the factors except the market quality.

Hence, the difference between the RMCTE of the

Pre-MiFID and the Post-MiFID periods will account

for the effect of the directive on market quality. Thus,

our study periods will be October 2003 – June 2007

serving as a reference period before implementation

and January 2009 – November 2011 after the MiFID

entered in vigor. We intentionally skip the interval of

time between July 2007 and December 2008 due to

the subprime crisis and the bias it could introduce in

the data.

The methodology consisting of a comparison of

the period before the implementation of a policy to

the period after the application of the latter is standard

in the literature for assessing newly implemented

regulations (See, for instance, Gresse 2011, Porter

and Weaver (2005), etc.).

To determine the effect of the MiFID directive on

systemic underperformance (the part of managers’

underperformance only related to market factors,

namely volatility and market quality), we will

examine a metric that we call the downside tracking

error that we note “DTE.” The tracking error is a

metric that treats all types of deviation from the

benchmark in the same way. However,

underperformance and outperformance do not have

the same implications in terms of portfolio

management. This is the rationale behind the

calculation of the DTE. We calculate it as the

negative deviations from the benchmark.

We will extract the idiosyncratic downside

tracking error by creating an extensive portfolio of

portfolios composed of 21 different ETFs and Mutual

Funds randomly chosen in the universe of investable

funds in the European Markets and displaying an

average negative tracking error during the study

period.

As in a single portfolio, the risk related to a

portfolio of portfolios is inversely associated with the

number of portfolios included. Hence, by creating

such a collection of portfolios, we will eliminate all

How Did MiFID Affect Systemic Underperformance in the European Equity Market? A Tracking Error based Approach

the idiosyncratic tracking error (Vardharaj et al.,

2004).

Once we have the systemic downside tracking

error, we can then split the non-idiosyncratic part of

the tracking error into two parts. The first part will be

related to volatility, and the remaining one will be

related to market quality, which includes the effect of

the implementation of the MIFID.

To isolate the part related to volatility, we will run

a univariate regression model using the market

volatility as the explanatory variable and the

downside tracking error of the diversified portfolio of

portfolios as the explained variables. We had to

transform the downside tracking error by taking its

absolute value so that we could use its logarithmic

form.

We show the downside regression equation here

below:

𝑙𝑛 ∣ 𝐷𝑇𝐸



∣− 𝑙𝑛∣𝐷𝑇𝐸



∣= µ



𝑙𝑛𝑉



−𝑙𝑛𝑉





+ 𝜀



(2)

Where:

 DTE

is the downside systemic tracking error

 µ is the elasticity of DTE

related to volatility

 V

is the volatility of the benchmark

The part of the downside systemic tracking error

explained by volatility corresponds to the coefficient

of determination in the regression model obtained in

equation 2 above and is noted MCVDTE (Marginal

Contribution of Volatility to Downside Tracking

Error).

We will deduct the part accounted for by market

quality (including the implementation of MIFID) by

subtracting the MCVDTE to 1 (See equation 3

below).

𝑅𝑀𝐶𝐷𝑇𝐸 = 1 − 𝑀𝐶𝑉𝐷𝑇𝐸 (3)

Where:

 RMCDTE is the Residual Marginal Contribution to

Downside Tracking Error.

 MCVDTE is the part of the downside tracking error

of the diversified portfolio explained by the

volatility of the benchmark.

We will use the criteria below to see if the MIFID

has affected the investor’s ability to match their

benchmark. Additionally, we also consider the

elasticity µ to confirm the effect of volatility on

systemic downside tracking error.

We are evaluating the impact of MiFID on

downside tracking error.

If the RMCDTE before the MiFID is higher than

the RMCDTE after MiFID, then market quality issues

played a less critical role before MiFID

implementation in explaining downside systemic

tracking error, which would mean a better market

quality.

If the RMCDTE before the MiFID is lower than

the RMCDTE after MiFID, then market quality issues

played a more critical role in explaining downside

systemic tracking errors after the implementation,

which would mean a lower market quality.

In fact, in a perfectly efficient market (maximum

market quality), all participants should be able to

match their benchmark perfectly (no

underperformance due to market frictions and no

outperformance due to market mispricing).

We made regressions on 167 weekly observations

for the Pre-MiFID period and 179 weekly

observations for the Post-MiFID period.

We will examine the RMCDTE at a 95%

significance level to ensure the statistical significance

of the results.

4 NOTE ON DATA TREATMENT

AND ANALYSIS PROCEDURE

Due to the peculiarity of the study period, we have

performed a CUSUM squared test on the data for the

period running from October 24

, 2003 to November

, 2013, to ensure that our results are not sensitive

to the effects of the subprime crises. The results of the

test showed parameter stability for the whole period.

It seems that the impact of the subprime crises on the

downside systemic tracking error was minimal. The

details of the CUSUM squared analysis are available

in the appendix.

The data for the Post-MIFID period did not

display perfect homoskedasticity. Hence, all

regressions used robust standard errors. Additionally,

the data showed positive autocorrelation. To deal

with it, we first differenced all variables and

performed the Dickey-Fuller test on their respective

first differences. This process revealed that the first

differences were stationary. Hence the data follows

an AR (1) stationary process. This condition is

necessary to apply the Prais Winsten process with

robust standard errors to correct for autocorrelation

and heteroskedasticity.

Finally, we performed the Dickey-Fuller test for

all variables. None of them is stationary except the

tracking error for the Pre-MiFID period. However,

performing the Johansen test of the variables shows

FEMIB 2020 - 2nd International Conference on Finance, Economics, Management and IT Business

that the variables are cointegrated. As a result, despite

the non-stationary variables, the results from our

regression model are valid.

5 COMMENT ON THE RESULTS

We created a diversified portfolio of portfolios

randomly selected from the universe of investable

equity portfolios in the European markets and

displaying an average negative tracking error during

our study period.

Our analysis has permitted to highlight a

significant difference between the Pre MiFID and the

Post-MiFID periods. We observe that factors related

to market quality account for 25% in the explanation

of the systemic downside tracking error in the Post-

MiFID period while they explained all the systemic

downside tracking error in the Pre-MiFID period

since market volatility was not a significant variable

in the Pre-MiFID period.

The coefficient of market volatility confirms this

finding; the latter is significant and equal to 0.63 in

the Post-MiFID period, while it is not significant in

the Pre-MiFID period.

These results would point towards a decisive role

of MiFID and a better market quality after its

implementation.

However, these results do not tell us the effect of

MiFID of systemic outperformance. Its impact on the

latter can be different from the effect found on

systemic underperformance. To have a complete

study, we would need to investigate the impact of

MiFID on systemic outperformance through its effect

on market quality. Our future research will extend the

methodology used in this article on a different data

set.

REFERENCES

A. Frino and DR. Gallagher, 2001, “Tracking S&P 500

index funds”, In The Journal of Portfolio Management

B. Coriat, 2008, «L’installation de la finance en France.

Genèse, Formes spécifiques et Impacts sur l’industrie.»,

In Revue de la régulation.

E. Elton et al., 2003 “Incentive Fees and Mutual Funds,” In

The Journal of Finance

H. Degryse et al., 2015, “The impact of dark and visible

fragmentation on market quality,” In Review of

Finance

G. Rompotis, 2011, “Predictable patterns in ETF’s return

and tracking error,” In Studies in Economics and

Finance.

G. Larsen and B. Resnick, 1998, “Empirical Insights on

Indexing,” In The Journal of Portfolio Management

L. Kostovetsky, 2003, “Index mutual funds and exchange-

traded funds,” In The Journal of Portfolio Management

M. Lenglet, M. Riva, 2013, “The unintended consequences

of financial regulation: why do algorithms create new

risks in European financial markets?”, In Revue de la

regulation

R. Vardharaj, F. Fabozzi, F. Jones, 2004, “Determinants of

Tracking Error for Equity Portfolios,” In The Journal of

investing

R. Bloomfield, M. O’ Hara, 1999, “Market transparency:

Who wins and who loses?”, The Review of Financial

Studies

S. Buti et al., 2011, “Diving into Dark Pools,” In Fisher

College of Business Working Paper No. 2010-03-010

X. Vives, 1995, “Short-term investment and the

informational efficiency of the market,” In The Review

of Financial Studies

P. Gomber and A. Pierron, 2010, MiFID-Spirit and Reality

of a European Financial Markets Directive, In SSRN

C. Gresse, 2011, Effects of the competition between

multiple trading platforms on market liquidity:

evidence from the MiFID experience, In SSRN

M. Cremers and A. Pareek, 2016, Patient capital

outperformance: The investment skill of high active

share managers who trade infrequently, In Journal of

Financial Economics

G. Gastineau, 2004, The benchmark index ETF

performance problem, In Journal of Portfolio

Management

P. Hu et al., 2004, Fund flows, performance, managerial

career concerns, and risk-taking, In Management

Science

D. Blitz et al., 2012, The performance of European index

funds and exchange funds, In European Financial

Management

D. Blitz and Huij, 2012, Evaluating the performance of

global emerging markets equity exchange-traded funds,

In Emerging Markets Review

N. Charupat and P. Miu, 2013, The pricing and

performance of leveraged exchange-traded funds, In

Journal of Banking & Finance

A. Lin and A. Chou, 2006, The tracking error and

premium/discount of Taïwan’s first exchange-traded

fund, In Web Journal of Chinese Management Review

A. Agapova, 2011, Conventional mutual index funds versus

exchange-traded funds, In Journal of Financial Markets

Plykha et al., 2015; Equal or Value Weighting?

Implications for Asset-Pricing Tests, Working Paper,

Frankfurt School of Finance and Management

R. Bloomfield, M. O’ Hara, 1999, “Market transparency:

Who wins and who loses?”, The Review of Financial

Studies

A. Madhavan et al., 2005, Should securities markets be

transparent? In Journal of Financial Markets

G. Larsen and B. Resnick, 1998, “Empirical Insights on

Indexing,” In The Journal of Portfolio Management

A. Frino and DR. Gallagher, 2001, “Tracking S&P 500

index funds”, In The Journal of Portfolio Management

How Did MiFID Affect Systemic Underperformance in the European Equity Market? A Tracking Error based Approach

L. Kostovetsky, 2003, “Index mutual funds and exchange-

traded funds,” In The Journal of Portfolio Management

F. Osterhoff and C. Kaserer, 2016, Determinants of tracking

error in German ETFs – the role of market liquidity, In

Managerial Finance, Vol. 42 Issue: 5, pp.417-437,

N. Milonas and G. Rompotis, 2006, Investigating European

ETFs: The case of the Swiss exchange traded funds, In

2006 Annual Conference of HFAA in Thessaloniki

T. Day et al., 2001, “Investigating underperformance by

mutual fund portfolios,” Working Paper.

N. Delcoure and M. Zhong, 2007, On the premium of

iShares, In Journal of Empirical Finance

G. Gastineau, 2002, Equity Index Funds Have Lost Their

Way, In Journal of portfolio management Vol. 28, Issue

2 Winter 2002

D. Kundisch and C. Klein, 2009, On the pricing of various

issuers of index certificates on the DAX, In Journal of

Banking Law and Banking ZBB

C. Gresse, 2011, Effects of the competition between

multiple trading platforms on market liquidity:

evidence from the MiFID experience, In SSRN

FEMIB 2020 - 2nd International Conference on Finance, Economics, Management and IT Business