We got some questions following last Monday’s post on cash vs. non-cash collateral on CCPs and did some further investigation. Here’s what we’ve got so far.
A summary of last week’s post:
– End-users who post non-cash collateral under the LSOC model have the risk of not receiving back the same collateral they post. Update: this appears to be true primarily in case of a default, not as a day to day concern under typical business conditions. We’ve heard that this might be just under European versions of LSOC or US – that’s not confirmed to us yet. Still, end-users in the US and Europe are concerned.
– Concerns about non-cash collateral may hamper the development of the collateral transformation trade insofar as government bonds are lent in exchange for lower credit assets.
The questions came regarding the economics. We used some numbers we have seen in the market that were dated and did not line up to what market participants were seeing on last week’s screens. Here are the updates:
We said last week that an investor lending out high grade corporate bonds would pay 20 bps. Our digging finds this is true on the dealer side funding in tri-party for short dates. A non-purpose loan for term would pay higher rates.
We said last week that the securities lending rate for government bonds lent with corporate bonds as collateral was in the 50 to 70 bps range. That was true a few months ago but not in the current market. We are now hearing in the 25-40bps range depending on the term.
We said last week that we saw a pricing mismatch, whereas our further investigation shows no mismatch for a corporate bond lent for a non-purpose loan. We still think that further demand to fund corporate paper would push repo or non-purpose rates higher.
In any case, we still think that rates will move around as there is more demand to transform assets (corporates into cash, equities into governments, etc) to post on CCPs. Right now demand is low but we see that changing as Category 3 derivatives clearing begins in September, EMIR kicks in next year, and the LTRO shuts down later on.
For the record, we really appreciate these questions and feedback. It helps us do a better job.