Clarus Financial Technology

CAD Swaps – What Does the Data Show?

Amir took a look at CAD swaps when we first added the new Canadian DTCC data source at the beginning of this year. Now that we have more data and reporting has bedded down a bit, I thought it worthwhile to analyse the data to see if two sources are better than one.

Canadian reporting is all uncleared

For whatever reason, the majority of vanilla CAD IRS being reported by DTCC_CA is coming through as uncleared. To the extent that I find only 74 swaps reported as “Cleared” so far this year, versus 10,230 uncleared swaps! I do not believe this can be accurate, but it has been the same ever since reporting began, as Amir highlighted.

If anyone knows why this might be, please get in touch.

The amount of crossover between US and Canadian reporting is surprisingly small

On a trade count basis, we find that only 1,179 vanilla CAD IRS trades have so far been “double-counted” – i.e. included in both the US and Canadian DTCC data feeds. By trade number we see the following;

SDR Split by Trade Number

The data quality is pretty poor.

There are over 1,800 swaps that have coupons off-market (typically zero with no adjustment to par). These trades are overwhelmingly in the Canadian SDR – over 91% of trades with a zero percent coupon have been reported to DTCC_CA in 2017.

As we saw with the US experience, this is fairly common in the early days of trade reporting as problems are ironed out. As market participants continue to consume the data, we expect this to improve over time.

The average size of trades is significantly different across SDRs

We have 3 SDRs to analyse data from for CAD IRS – DTCC, DTCC_CA and BBG. We can therefore analyse the data by “source” to look at a number of primary economic terms. First – average trade size in notional terms during 2017:

Average CAD Trade Size

Showing;

The average maturity of trades is the same both “onshore” and “offshore”

On a notional-weighted basis, we see a weighted-average maturity at both of the DTCC SDR venues of ~2.5 years. This is somewhat shorter than we see in other markets. Remember that this excludes OIS products and FRAs.

The average daily volume reported to SDRs has doubled in 2017

If you take one thing away from this blog, please note that the new trade level reporting from the DTCC appears to have nearly doubled our coverage of the CAD IRS market on a trade-by-trade basis.

Average Daily Volumes CAD IRS

On the face of it, I would like to think that trade level coverage of the CAD IRS market has therefore more than doubled this year as a result of the Canadian Trade Reporting requirement. Of course, we cannot say for sure that volumes have not just increased naturally, but a leap of this magnitude seems unlikely.

If we consider the BIS triennial survey as the “whole” market we must acknowledge the caveats that we have noted before – namely the inclusion of inter-affiliate trades and the sheer number of reporting dealers at 1200+ sales desk. The BIS survey may therefore have a tendency to overstate the total size of risk transfer that actually occurs on a given day.

The price impact of trades is much larger in the Canadian “onshore” market

I (briefly) traded CAD swaps back in my trading days. Never very well I might add! But the over-arching flavour was of a market driven by “locals” (i.e. the Canadian banks) who had very strong franchises and were very protective over it. They hence were not overly keen to extend any meaningful liquidity to international banks trading all currencies in an effort to deliver a “one-stop shop”. Therefore, the market tended to be bifurcated along these lines – small trades between local banks and internationals, and larger trades that really defined the price action, and tended to occur between locals. I therefore wondered whether this could be witnessed in the data.

I looked at all spot starting five-year CAD IRS that have traded this year. I ran a price dispersion measure on these trades to look at the “price impact” at each SDR. The results are surprising:

Price Dispersion in 5 year CAD IRS

Showing;

Does this suggest that more Client trades are reported to DTCC_CA and they are subject to a larger bid-offer spread than a (presumably interbank) US market? Or is this simply consistent with locals defining the price action? Or simply due to the inclusion of uncleared trades in our sample?

It is an interesting finding that we will revisit in future blogs as the data continues to improve over time.

In Summary

Stay informed with our FREE newsletter, subscribe here.

Exit mobile version