3rd CEPR-Imperial-Plato Market Innovator (MI3) Conference 2019 -The benefits of Periodic Auctions beyond MiFID II Dark trading caps

  

 

3rd CEPR-Imperial-Plato Market Innovator (MI3) Conference 2019

The benefits of Periodic Auctions beyond MiFID II Dark trading caps

 

Paul Besson (Kepler Cheuvreux)
Matthieu Lasnier (Kepler Cheuvreux)
Antoine Falck (Kepler Cheuvreux)

 

Paul Besson is the Head of Quantitative Research at Kepler Cheuvreux. His main areas of research are market microstructure and execution on equities. Prior to that, he worked as a fund manager in quantitative arbitrage, for both hedge funds and institutional funds. He also has three years’ experience as a buy-side Head of Quantitative Research and lectured in finance for five years at Sciences-Po and HEC in Paris.

 

Matthieu Lasnier is a Quantitative Analyst at Kepler Cheuvreux. He studied for a postgraduate degree in economics, finance and statistics at ENSAE Paris.

 

Antoine Falck is currently a Research Associate at Capital Fund Management in New York. Prior to that, he was a Quantitative Researcher Intern at Kepler Cheuvreux and has a particular interest in algorithmic trading and asset management.

 

About the paper

 

The paper takes a highly analytical approach to analysing the design, market share, trade size, timings, reversion trends and market impact of periodic auctions, with specific focus on the effect of MiFID II’s dark trading caps.

 

It divides the above lines of enquiry into six key points, concluding key findings on Periodic Auctions, and offering an outline of and key next steps following ESMA’s Final Report from 11th June on its Call for Evidence on Periodic Auctions.

 

It begins by tracing where periodic auctions came from, citing regulatory constraints and the problems associated with Dark trading. It also provides outlines of their key features.

 

Main Takeaways

 

The key findings of the paper are that:

  • Periodic Auctions’ popularity indicates that they aren’t just a substitute for Dark trading.
  • Random end time prevents participants from being gamed.
  • Much like Dark trades, they do enable Market Impact reduction.

 

According to their findings, they conclude that the advantages outweigh the potential threats that represent 2% of non-contributing liquidity to the price formation process.

 

Periodic Auctions were born as a result of two key factors, according to the paper: to provide a Dark trading alternative that respects regulation, while mitigating against the problems of Dark trading like information leakage and adverse selection. Key characteristics of the Periodic Auction include the existence of a succession of auctions between 8:00 to 16.30 (UK Time) with random call durations, less than or equal to 100ms.

 

Surprisingly, despite the rebasing of Dark volume caps 6 months after implementation, the data shows that the market share of PAs continued to rise, even on uncapped stocks. This suggests that PAs have benefits besides simply offering an alternative to Dark trades.

 

The paper also finds that, according to analysis of average trade sizes, most PAs settle at European Best Bid Offer midpoint (over 96%); PAs liquidity imports prices as opposed to making prices as in Lit markets.

 

Furthermore, in terms of timing and trading intensity, the PAs liquidity curve follows continuous liquidity patterns throughout the day, except that they rise less towards the end of the trading day than Lit markets. They are also less likely to reduce the number of preceding trades – after a Dark trade, there is often immediate future aggressive trading, however after a PA trade, the participants that agreed mostly on prices have emptied making future trades less likely.

 

In terms of reversion of PA price, no reversion is observed after a trade – the coinciding trading imbalance was found to be only -20%: the random duration is neither beneficial nor detrimental to participants, with neutral timing setting a level playing field.

 

Furthermore, much like Dark trading, PAs reduce market impact according to a two-factor linear model to estimate amplification factors.

 

Images: Muhammad Ashraf ©2019