Clarus Financial Technology

SEF Trading on US Holidays

Happy Thanksgiving!

If you logged on to our website on Friday, you will have seen something fairly unusual:

Friday Dashboard
Clarus Dashboard showing summary of the USD, EUR, JPY and GBP swap markets on 26th November 2015

Showing;

USD Holidays

I’m not surprised that we saw almost no volume trading in USD swaps last Thursday, due to the Thanksgiving holiday in the US. That is consistent with the volumes we see reported to SEFView by each individual SEF on all USD holidays – i.e. USD swaps just do not trade when the UST market is closed.

However, what I did find surprising were the non-USD volumes reported by the SEFs on Thanksgiving:

SEFView daily non-USD volumes in November

Here we see that;

SEFView Market Share Percentage per day November 2015

That last point got me thinking – because it really surprised me. My interpretation of SEF-trading on US holidays has typically been that it is volumes recorded by London-based US banks. But if that were the case, why don’t we see more volume in D2D space on US holidays and zero volumes on the D2C platforms? In addition, I had expected BSEF to outperform Tradeweb, because common wisdom states that many of the smaller interbank tickets now trade with a simple click on a Bloomberg terminal. Strange.

Volumes vs ADVs on USD Holidays

Using SEFView, I downloaded daily data for both EUR and USD swaps for 2015. I then calculated Average Daily Volume each month, and finally calculated the percentage of notional that trades on a US holiday compared to these monthly averages.

For USD swaps, the story is consistent – little to no volumes trade on a US holiday. This applies to both Dealer-to-Dealer and Dealer-to-Client SEFs, although D2C venues did see slightly more volume on Good Friday (a half day, according to SIFMA recommendations).

But for EUR Swaps, we see a completely different story. First for Dealer volumes:

On-SEF D2D EUR Swap Volume vs Average Daily Volume on each US holiday this year

Showing that on-SEF EUR Swaps traded almost as normal on Martin Luther King day, Veterans Day and even Independence day this year. This kind of activity, as I said at the top, may broadly be attributed to US banks operating EUR franchises out of London, plus some Australian and Canadian banks who prefer to trade on-SEF. Generally speaking, whilst these market participants are free to trade off-SEF as well, maybe they have become more ambivalent as time wears on?

Now, what about EUR on-SEF Client trading on US holidays? Surely there are no US clients in the office to trade with? So there might be prices, but who would choose to trade in a reduced liquidity pool compared to Off-SEF trading? A lot of people, it turns out!

On-SEF D2C EUR Swap Volume vs Average Daily Volume on each US holiday this year

Showing;

ISDA Liquidity Fragmentation

Looking at what else has been written around these ideas, ISDA are very keen on the concept of bifurcated liquidity pools – namely a US-persons EUR swap liquidity pool versus a Europeans-only pool. There is plenty to read on this in their 3 or 4 reports on the ISDA Research pages. In summary, they make a very fair assumption that LCH is the global heart of all cleared EUR swap liquidity and look at trades between US Persons, Europeans only and US Persons with Europeans.

Also noteworthy about the data-set that ISDA use is that they only consider dealer volumes. So I wondered if we saw anything similar using CCPView for Client trades.

Unfortunately, we don’t have the detailed counterparty information that ISDA are privy to, but we can look at how cleared EUR swap volumes at LCH change on a US holiday as a rough proxy. This should give us some idea of how many clients are US-based (SPOILER: I thought clients would be much more US-centric than the data suggest, particularly given the absence of a Clearing Mandate in Europe).

Volume traded on a US holiday in EUR swaps as a percentage of Average Daily Volume within the month

Showing;

In Summary

Stay informed with our FREE newsletter, subscribe here.

Exit mobile version