MiFIR 2021 Corporate Bond Trade Data Analysis and Risk Offset Impact Quantification | AFME

Share this page
MiFIR 2021 Corporate Bond Trade Data Analysis and Risk Offset Impact Quantification
03 May 2022
Download Links
​ ​

Press release available in English, French, German.

AFME has today published a first of its kind study on fixed income data on trade-out times. This new concrete data will help inform the MiFIR fixed income deferrals calibration policy discussion.

This shows that the majority of fixed income trades could be made transparent in near real-time, but also finds there is a clear need for a longer deferral period for the publication of larger or illiquid trades. Data provided by FINBOURNE Technology for this study demonstrates that an inadequate deferral calibration - as currently proposed by the European Commission - could have potentially significant negative implications for market liquidity.

The AFME paper analyses recent European fixed income trade data from around 5,500 of the most frequently traded securities. The analysis focuses on the corporate bond landscape (rather than government bonds) to identify which types of trades could be subject to near real-time price and volume transparency, and which types of trades should be subject to deferrals.

From the data set studied, AFME and Finbourne find that different deferral periods need to be applied based on the trade size and issuance volume, among other characteristics.  Applying the Commission’s proposed deferral regime to all trades, especially those larger in size or illiquid, risks exposing liquidity providers to potential undue risks, which could negatively impact the amount of liquidity/pricing that market makers are able to provide.

Key findings:

  • Small trades (of less than EUR 500k) account for the majority (c. 70%) of the overall number of trades in the data set and can support being reported in near real-time. Therefore, making these small transactions transparent will significantly improve transparency by almost 10 fold, increasing from 8% of transactions currently being reported real-time to almost 70% of transactions becoming near real-time transparent.
  • The smaller the trade size and the more liquid the instrument, the less risk is associated with rapid dissemination of price and volume information for liquidity providers, with the ‘trade out’ (i.e. moving the risk off the bank’s balance sheet) being less than 1 day for liquidity providers.  
  • However, this 70% reflects 13% of market volume. Therefore these transactions represent a much smaller percentage of market volume than of the number of trades.  
  • Larger transactions (of more than EUR 500k) reflect a relatively small percentage of total transactions, accounting for c. 30% of total transactions but a much larger share of market volume.  The data analysis demonstrates that the larger the transaction, the longer it takes to 'trade out' and clear the market. For trades larger than EUR 1 million, it takes on average 6 business days to ‘trade out’ of positions. For trades over EUR 5 million it takes on average 19 days to trade out, while larger trades take even longer.
  • The deferral regime should have a conceptual link between trade size categories (i.e. near real-time transparency), bond liquidity and deferral periods (i.e. for a regime with a higher trade size, or deemed illiquid the deferral period should be longer);
  • At the same time, longer deferrals for the small number of large transactions should limit the risk of liquidity reduction in the market for institutional investors.  

AFME therefore opposes a hardwiring of price and volume deferral calibration in primary legislation (as is currently proposed).  Since each fixed income asset class will include a significant number of illiquid bonds, AFME urges the co-legislators to adopt a range of deferral periods, going beyond the Commission’s proposal for maximum deferral period for prices (by the end of the day) and volume (within two weeks). ESMA will then be able to calibrate the details of which bonds should go into the various deferral categories, this should be based off detailed and high-quality data.