MIFID II Data Day Two
Yesterday’s blog was very well received, so we’ve decided to do it again today. We’ll take a look at;
- Yesterday’s data sources again. Will they work today? Will we get any historical end of day files? Will I finally be able to use some data in Excel?
- New sources of data – we are still waiting for a definitive list of APAs. Come on ESMA!
- Different data – I see there is some Compression data now. We’ll probably start with that.
- I’ll try to make sense of the bond data as much as possible. Ten most common ISINs traded or similar.
This is a live blog, so remember to refresh throughout the day (hit F5 in your browser) and scroll down.
Why do we care about MIFID data?
Some of you may be wondering why we are doing these blogs? We have a successful data business, part of which involves consuming the public SDR feeds in the US. The value from this data (once it is cleansed, augmented and made easily accessible) is clear. I think one of the great examples is our FREE Daily Report for USD swaps. We will produce similar for EUR and GBP if the right data ever becomes available and usable.
It is FREE to subscribe – just register here.
Live blog continues below….
I think this will be the final post on our live blogs until early next week. We are expecting to see more data come through at the start of the week, particularly for deferred transactions. That will give us a better idea of the overall scale of transactions being captured.
MIFID II data could have been delivered via a standardised format from all providers, delivered in a CSV file that was truly “machine readable”. With better regulations in place, this could have happened. Look at the US.
Instead, we have seen a lot of different approaches from different operators.
The project teams and IT professionals who have put together these solutions deserve a lot of credit for getting them done on time, and most data has been accessible for a good amount of up-time.
I am left wondering what will happen with the paid-for data feeds. Are they delivered in the same format as the (free) public data feeds, but 15 minutes earlier? This seemed to be the intent from the regulations. Will customers pay a fee for data that they cannot aggregate or manipulate? From our experience running a succesful data business, that would be a “no”!
ESMA have updated their website with the following:
The link is here. Lots of useful links, search tools to use ( we are going to be busy). But still only one APA in the DRSP file. Frustrating.
Take a look at the GFI OTF pre-trade transparency right now:
I have nothing bad to say about this. It is clear, usable and very exciting data to have. They should be credited for following the spirit of the law and not just the letter.
However, I don’t see any other MTF or OTF providing such pre-trade transparency in GBP swaps (or any swaps to be fair!). When I look at SEFView, BGC (who own GFI) have a tiny market share in GBP IRS. What figures?
Or not. Back to Earth with a bump. The ESMA register of APAs includes….ONE APA. A single APA? Really? Come on….this is becoming farcical.
I finally have it! The ESMA register of APAs. Of course it wasn’t listed as “news” or with the rest of the registers. Lots more data to come now…
A few APAs seem to be struggling now. I have a couple of the websites down right now.
No matter, we still have the daily files from yesterday to interrogate. One “easy” win is to look at all liquid German Government bonds. From one trading venue alone, we saw over €2.6bn in notional trade and 200+ trades. Some simple pricing metrics included as well. The simple average price shows there is some bad data in there.
Here’s our first microservice for MIFID data. A simple ISIN look-up, but it tells you which trading venues are registered to trade it, a short description as well is provided.
Remember that our Microservice architecture can be accessed in a number of ways – python, R, and even Excel. Should prove useful in the coming days….
Apologies for the brief hiatus. Lots of data still to get through!
Daily files from yesterday are available. Despite being listed as “post trade”, one file includes 1,460 trades for the EUR IRD EZW2GD95CKV8, which according to the DSB is the ISIN for a 9 year swap. There are 2,035 EUR IRS in total reported, totalling €726bn. It seems to me that this is unlikely? Total EUR IRS reported to the US SDRs yesterday was about €40bn….this daily file from Europe claims to be from only one trading venue!
Typical teething problems can be seen in the data. Not least in the ISIN generation itself. Check out this ISIN which has pre-trade quotes being published by an APA:
Swaps do not trade versus Euribor 9 month. Or Euribor 9 year. It is an obvious mistake on the generation of the ISIN and highlights that:
- Quotes (and/or trades) may be reported against an erroneous ISIN.
- If this ISIN is corrected, there will be a duplicate ISIN in existence for the same instrument.
- You cannot aggregate the data if the ISINs are wrong.
- There is clearly no quality control check at the DSB looking to see if the ISIN creation makes sense. Euribor 9 year doesn’t exist…..
- It is a simple config change to correct. Let’s hope these errors are being picked up….
We are still getting excited about actual trades. The long road of downloading activity files for the 3rd and checking ISINs is upon us. But we are seeing the odd trade reported in EUR IRDs today as well. Two such trades in EUR100m just went through:
We now need to cross-reference those ISINs with a (usable) APA providing pre-trade quotes to see if there are matches. Much as we cross-referenced SDR data yesterday (for EUR swap prices…because the data is clean, usable and understandable). These are nice use-cases already for the data.
I wondered about this. If an APA makes only slice files available AND the website goes down (assuming people are grabbing the slice files)….how is that a viable way of publishing data? This is the problem with not persisting the data. Oh, and look, an APA that chooses to only publish slice files is currently unavailable….
Making the data usable – this is what this exercise is all about. And we have our first tape:
We’ve got work to do. But this is a snapshot of close-to-streaming quotes for EUR IRDs. It is instantly recognisable and usable for anyone used to consuming market data. Now, to decipher those pesky ISINs….
From the sublime (see 08:55) to the ridiculous. At least they are honest….far less frustrating than downloading empty zip files!
To be fair, the above screenshot is for a single Systematic Internaliser. I’ve clicked on quite a few and still returned nothing so far though.
Oh my, oh my, oh my. Someone has actually obeyed the spirit of the law! Five minute slice files, with Bids AND Offers! This is more like it. Where is that “excited GIF” when you need it?!
A bit of a later start this morning, so we are expecting lots of trades and quotes to be waiting for us as we go about remembering our log-in details from yesterday.
Still, I think we are up and about before ESMA. Still no register for APAs or ARMs that I can see, and the Transparency calculation results (it would be handy to have the ISINs right now….) doesn’t even work this morning. As we said yesterday, this was all a little bit predictable. Who gets to fine the regulators for failing their MIFID II implementation?!
5 thoughts on “MIFID II Data Day Two”
Many thanks for the great updates. Where do you actually find the post-trade data online? Haven’t found anything on the websites of different APAs
Thanks Matthias, glad it is proving useful. We’ve had to call a lot of the APAs to get the actual public web address. Woould be happy to discuss off-line, feel free to reach out.
Chris, Amazing blog series as usual. Any plans in the cards to produce an analysis on which venues, APAs % of market share in terms of reported volumes ?
Thanks for the positive feedback! We will produce any analyis that is actually possible given the data available. At the moment, that is proving very difficult. Best I can say at the moment is to subscribe to the newsletter to keep up to date with what we are able to do with the data.
Thanks Chris, facing same kind of issues you are – given your extensive knowledge assumed you must have landed on the holy grail which gives you the unique perspective 🙂 that we mortals don’t get to see. Will keep a watch out on your newsletter, blogs.
Comments are closed.