The FSB has advocated for central clearing for Repo, but only for widely traded securities. The Fed has taking a similar approach on tri-party fire sale solutions – focus on the more liquid instruments. But where does this leave “risk assets”?
We define risk assets as those likely to lose liquidity in a stress situation and/or become more volatile:
- ABS, Investment and Non-Investment Grade
- CMO Private Label, Investment and Non-Investment Grade
- Corporates, Investment Grade and Non-Investment Grade
- Equities
- Money Markets (now incorporated into Other)
- Municipals (now incorporated into Other)
- Other (CDS, International, and Whole Loans)
(The categories that the Fed tri-party reports use have been a moving target – hopefully it is now stable.)
In the latest figures (August, 2013) we see nearly $300 billion of risk assets being financed in tri-party, just over 19% of the total. Equities is the largest single category, coming in at 40% of the risk assets and 7.3% of the tri-party total. Median haircuts ranged from 5% for investment grade ABS & corporates to 8% for non-investment grade ABS & corporates. The median haircut for equities was 8%.
So if these asset classes are orphaned from either the CCPs or tri-party fire sale solutions, what will become of them? How will cash lenders respond to the assets not supported by any of the tri-party fire sale or CCP solutions being talked about? The first thing that should happen is for haircuts to go up – potentially by multiples. Spreads should rise. Limits go down. But we know that tri-party haircuts tend to be pretty sticky and cash lenders aren’t always the most pro-active bunch (until they turn the spigot off).
But making matters worse, these categories of assets tend to have higher dealer market share concentrations. In the most recent data, the top 3 dealers in corporate or non-investment grade ABS were both over 50%. Equities market share for the top 3 dealers was 44.5%. Only corporate paper — IG and non-IG – had numbers that looked more like the liquid categories, 28.8% and 38.7% respectively. This was not always the case, especially for non-investment grade corporates. In Jan. 2012 the share for the top 3 dealers in non-IG corporates came in at 53.1%, in Jan. 2011 the share for the top 3 dealers was 54.8%. In comparison, US Treasuries excluding Strips came in at 35% (Aug. 2013) top 3 concentration.
What does this tell us? The high concentration numbers in risk assets financed in tri-party implies that a shock to any of those markets could cause a great deal of pain for a handful of dealers. (It might be interesting to know something about the concentration on the cash lender side too.) The potential for systemic risk driven by this high market share is scary. Like most systemic risks, the big picture across dealers is only something the Fed could know about. The concentration numbers have been published as part of the tri-party data and we applaud the Fed for doing this (same thing goes for the haircut stats). But is this the end of the story? Should the Fed be taking steps to discourage high concentration and if so, how?
Here is a link to the Fed tri-party reports.