Slide 1: The impact of COVID-19 on European Market Microstructure
The impact of COVID-19 on European Market Microstructure PAUL BESSON: EURONEXT, HEAD OF QUANT RESEARCH 6 August 2020 Paul Besson Head of Quant Research pbesson@euronext.com
Slide 2: The leading pan-European exchange powering access to capital markets
THE LEADING PAN-EUROPEAN EXCHANGE POWERING ACCESS TO CAPITAL MARKETS HELSINKI TALLINN ▪ SINGLE ORDER BOOK B ERGEN ▪ SINGLE COLLEGE OF REGULATORS STOCKHOL M ▪ HARMONISED RULE BOOK OSLO ▪ FEDERAL MODEL ▪ 8.2 BN EUROS AVG DAILY VOLUME COPENHAGE N ▪ 69% MARKET SHARE ON OUR LISTED STOCKS (2019) BERLIN DU
Slide 3: Contents
CONTENTS Part 1: EUROPEAN COVID-19 CRISIS TIMELINE & MARKET PERFORMANCE Part 2: THE ORDERBOOK REACTION TO COVID -19: BID-ASK SPREADS, AVAILABLE DEPTHS AND THE RESULTING COST OF TRADING - Spreads from 2008 to 2018 in Europe - Spreads and volatility from 2008 to 2018 Part 3: TYPES OF TRADING (LIT/DARK
Slide 4: Part 1: European COVID-19 crisis timeline & market performance
PART 1: EUROPEAN COVID-19 CRISIS TIMELINE & MARKET PERFORMANCE
Slide 5: Continental European COVID-19 crisis timeline
CONTINENTAL EUROPEAN COVID-19 CRISIS TIMELINE Period 1: Period 2: “Financial Market “Real Economy shock”, Drop”, 18 February 17 March to 10 May: to 18 March Belgium and France under confinement, Netherland and Germany under much stricter rules │5
Slide 6: Continental European COVID-19 crisis timeline
CONTINENTAL EUROPEAN COVID-19 CRISIS TIMELINE ▪ While Covid fatalities rose until 10 April in France, Belgium and the Netherlands, the worst market drop was observed from 18 February to 18 March, when our main indices dropped by −35%. At the end of July, main indices were down −15% compared to pre -
Slide 7: Continental liquidity during the European COVID-19 crisis
CONTINENTAL LIQUIDITY DURING THE EUROPEAN COVID-19 CRISIS ▪ During the market drop, daily turnover more than doubled on continental European indices. ▪ In addition to this rise in turnover, the nature of this liquidity has become different after the crisis. Daily turnover on main continental Europea
Slide 8: Continental liquidity during the European COVID-19 crisis
CONTINENTAL LIQUIDITY DURING THE EUROPEAN COVID-19 CRISIS ▪ Paradoxically, while daily turnover doubled during the worst of the crisis, the elementary average trade size almost halved from 10k to 6k Eur, and has remained at a low level since then. ▪ This means that the number of trades per day almos
Slide 9: Continental liquidity during the European COVID-19 crisis
CONTINENTAL LIQUIDITY DURING THE EUROPEAN COVID-19 CRISIS From 18 February to 18 March 2020: ▪ Continental indices lost −35% ▪ Daily turnover doubled ▪ Average trade sizes halved ▪ Number of daily trades quadrupled From 18 February to 30 th June 2020: ▪ Continental indices lost −15% ▪ Daily turnover
Slide 10: Part 2: The orderbook reaction to COVID-19
PART 2: THE ORDERBOOK REACTION TO COVID-19: BID-ASK SPREADS, AVAILABLE DEPTHS AND THE RESULTING COST OF TRADING
Slide 11: The orderbook reaction to COVID-19: basic definitions
THE ORDERBOOK REACTION TO COVID-19: BASIC DEFINITIONS The liquidity of the orderbook can be represented by 2 key characteristics: ▪ The bid-ask spread reflects the cost for trading immediately (aggressively) versus passive trading ▪ The depth of the orderbook displays the average liquidity available
Slide 12: European bid-ask reaction to COVID-19
EUROPEAN BID-ASK REACTION TO COVID-19 ▪ On large caps, bid -ask spreads multiplied by approximately 4 between 18 February and 18 March, from 4 to 18 bps, on both the French and the Dutch large -cap indices CAC 40® and AEX -Index®. The same relative increase was observed on smaller capitalisations. ▪
Slide 13: European bid-ask long-term fluctuations
EUROPEAN BID-ASK LONG-TERM FLUCTUATIONS From 2007 to 2018, between MIFID I and European bid-ask spreads from 2007 to 2018 (Stoxx600) MIFID II, European bid -ask spreads tightened overall, but widened during all crises ▪ Late 2008-end 2009: Great financial crisis Covid worst spread levels ▪ May 2010
Slide 14: Bid-ask spreads widen with volatility over 2020
BID-ASK SPREADS WIDEN WITH VOLATILITY OVER 2020 ▪ Greater volatility translates into a greater risk for the market maker and therefore a greater reward is required. That is why larger volatility is often associated with wider bid -ask spreads. ▪ In late June, the bid -ask spread on the CAC 40® reach
Slide 15: Bid ask spreads widen with volatility over the long term
BID ASK SPREADS WIDEN WITH VOLATILITY OVER THE LONG TERM ▪ Looking at larger European indices such as European bid-ask spreads and volatility (Stoxx600) the STOXX® Europe 600 from 2007 to 2018, we observe a strong relationship between volatility and the corresponding bid -ask spread. ▪ Whatever the
Slide 16: The COVID-19 crisis persistently harmed the depth of orderbooks
THE COVID-19 CRISIS PERSISTENTLY HARMED THE DEPTH OF ORDERBOOKS ▪ Before the Covid -19 crisis, a little less than €50k on average was available on the first limit of the CAC 40® components. This depth reduced to €15k on 18 March at the worst of the crisis. At the end of June the depth amounted to €3
Slide 17: Orderbook depth decreases with rising volatility
ORDERBOOK DEPTH DECREASES WITH RISING VOLATILITY Looking at larger European indices such as the Available depth at first limits vs volatility (Stoxx600) STOXX® 600 from 2007 to 2018: orderbook depth clearly decreases with increasing volatility as market makers reduce their exposure in order to maint
Slide 18: Transaction costs rose on all venues during the COVID-19 crisis
TRANSACTION COSTS ROSE ON ALL VENUES DURING THE COVID-19 CRISIS ▪ The effective spread displays the average cost of trading €25K immediately in a given set of stocks on a given market. This metric combines the spread and the depth available on different markets. ▪ On our indices, Euronext represents
Slide 19: The orderbook reaction to COVID-19
THE ORDERBOOK REACTION TO COVID-19 From 18 February to 18 March 2020: ▪ Bid-ask spreads were multiplied by 3 to 4 times ▪ Available depth at first limits was divided by 3 approx. ▪ Short term volatility on main indices was multiplied by 4 to 5 times From 18 February to 30 th June 2020: ▪ Bid-ask spr
Slide 20: Part 3: Types of trading, market fragmentation, and trading phases
PART 3: TYPES OF TRADING (LIT/DARK/OTC), MARKET FRAGMENTATION, AND TRADING PHASES DURING THE COVID CRISIS
Slide 21: Liquidity favoured lit trading during the COVID-19 crisis
LIQUIDITY FAVOURED LIT TRADING DURING THE COVID-19 CRISIS ▪ During the worst of the Covid -19 crisis, the market share of OTC trades and of Systematic Internalisers (SI) diminished while Lit trading market share rose. With rising risks, it makes sense that proprietary desks should be less inclined t
Slide 22: Liquidity favoured primary markets during the COVID-19 crisis
LIQUIDITY FAVOURED PRIMARY MARKETS DURING THE COVID-19 CRISIS Market share of the primary markets rose on Euronext and Xetra but dropped for the LSE. Nevertheless, since 2017, Primary Markets market share has risen steadily. Source: Reuters Eikon (daily data, Jan to June 2020) │ 22
Slide 23: European fragmentation increased from 2007 to 2018
EUROPEAN FRAGMENTATION INCREASED FROM 2007 TO 2018 European Primary Markets market share 2007 to 2018 ▪ Market fragmentation in Europe appeared in November 2007 with the implementation of MIFID I. Thus the market share of Primary Markets dropped from 100% to 65% on largest European capitalisations.
Slide 24: Part 4: How market participants reacted to the COVID-19 crisis
PART 4: HOW MARKET PARTICIPANTS REACTED TO THE COVID-19 CRISIS
Slide 25: How market participants reacted to the COVID-19 crisis
HOW MARKET PARTICIPANTS REACTED TO THE COVID-19 CRISIS On Euronext 4 types of participant constitute the overall turnover: ▪ Liquidity Providers (33%): registered market makers ▪ Proprietary traders (33%): mostly from large banks and other trading firms ▪ Global brokers (23%) and Regional brokers (7
Slide 26: Participants' reaction during the COVID drop
PARTICIPANTS’ REACTION DURING THE COVID DROP (18/02 TO 18/03) ▪ During the market drop from 18 February to 18 March, the largest buyers were retail brokers (+15%) followed by regional brokers (+11%). Global brokers were almost neutral and proprietary desks were net sellers. ▪ Contrarian retail flows
Slide 27: Participants' reaction during the rebound
PARTICIPANTS’ REACTION DURING THE REBOUND (19/03 TO 30/06) ▪ Only retail investors were net buyers, while global brokers were net sellers. ▪ Retail investors were particularly strong buyers in Real Estate (+18%), but also in Consumer Discretionary, Energy and Utilities. Source: Euronext │ 27
Slide 28: Participants' flows: correlation with performances
PARTICIPANTS’ FLOWS: CORRELATION WITH PERFORMANCES ▪ Global brokers clearly exhibit a positive correlation with market performance as well as all sectors outperformance. ▪ Retail investors’ negative correlation with the performance of sectors accounts for the specifically contrarian nature of their
Slide 29: How market participants reacted to the COVID-19 crisis
HOW MARKET PARTICIPANTS REACTED TO THE COVID-19 CRISIS From 18 February to 18 March 2020: ▪ Market-maker market share rose from 30 to 40% ▪ Global brokers market share dropped from 25 to 20% ▪ Retail brokers displayed the strongest trading imbalance +15% followed by regional brokers +11% From 19 Mar
Slide 30: Conclusion: market resilience and heterogeneity of market participants
CONCLUSION: MARKET RESILIENCE AND HETEROGENEITY OF MARKET PARTICIPANTS