Serving at the ECB’s pleasure: billions, if not trillions, of euro-denominated assets pledged as collateral for three years of funding.
And while it’s banged up by being pledged at the central bank… it is, naturally, not circulating through the market. For European markets which depend on approximately €6,000bn worth of repo funding, this is potentially a big problem. Every time collateral is stalled, its ability to be rehypothecated, or used again, is impacted. The shortening of rehypothecation chains, meanwhile, not only kills the velocity of collateral but acts as a contractionary force in money markets.
Add post-MF Global fears about the reliability of segregated account systems and the amount of usable collateral available for repo becomes even more vulnerable.
What’s the worst case scenario?
Well, it’s worth remembering that according to Gorton and Metrick, the trigger for the collapse of Lehman was actually a withdrawal of repo funding and an increase in repo haircuts.
We, meanwhile, have argued that the ECB’s very first LTRO, by attracting the cheapest-to-deliver collateral (Greek debt), may have been a key catalyst in destabilising the Greek repo market, a fact which made it harder for Greek banks to get funding and which also encouraged settlement fails — later misinterpreted by the government as signs of a speculative attack on Greek bonds.
Except this time it could all lead to a withdrawal of funding on a much larger level.
Though, it seems, we’re not the only ones who think that central bank liquidity ops might have a lot to do with the problem.
Sandy Chen over at Cenkos Securities has recently taken a much closer look at the issue. As he noted in a January report entitled “The Return of the Grim Repo”:
… in our view, quantitative easing and central bank repos only exacerbate this collateral crunch. None of the central banks rehypothecate the collateral that is pledged against their repos or other secured funding support mechanisms. In particular, the Fed and BOE’s requirements for relatively high-quality collateral (or the discounts/haircuts applied to lower quality collateral, same thing) use up the supply of gilts, US Treasuries etc. And private broker dealers have continued to demand high-quality, liquid collateral to back their repos.
And this matters because:
If rehypothecation chains are actually shortening again, and thus pulling back the repo markets, the immediate effects of this would be lower income from trading/rates businesses as the overall repo market shrinks.
We are more concerned, however, about repo counterparty risk. In the current market environment, we think that many repo counterparties will be facing increased wholesale funding pressures . And non-banks, e.g. hedge funds, can’t access the central banks’ QE/LTRO support funding. This raises the risk of repo counterparty failures – with the initial failures being amongst non-bank financials, but with the bulk of the writedowns and impairments being recognised by the broker-dealer banks who are their counterparties.
This would, in turn, reinforce the repo deleveraging/collateral crunch dynamics outlined above. A self-reinforcing cycle.
Painful indeed. Though, in Chen’s opinion quite necessary, since the contraction will eventually increase the resilience of the system — a weeding out of rehypothecation practices which should never have been allowed in the first place, if you will.
So who’s most exposed to this continued repo deleveraging?
According to Chen, in Britain, it’s clearly the major repo dealers — the biggest being Barclays, followed by HSBC and then RBS.
As Chen notes:
As at end-2010 (the latest reporting date with available data) Barclays reported £423bn of collateral pledged by its clients against reverse repos and other secured lending, HSBC reported $334bn (£210bn) and RBS reported £94bn.
Barclays and RBS, meanwhile, also have the highest degree of repledging (rehypothecation), at 82 per cent for Barclays and 99 per cent for RBS:
*We should point out that FT Alphaville isn’t not sure that dividing securities accepted as collateral with securities repledged is actually that good an indicator, since pledged collateral includes more than repo.
Nevertheless, here’s one other interesting repo metric thrown up by Chen:
As a rough gauge of net dependency on repo funding, we simply subtracted repo funding from reverse repo lending. This showed negative (funding) gaps of £51bn at Barclays and £26bn at RBS, and positive (lending) gaps of £9bn at Lloyds and $9bn at Standard Chartered. HSBC’s data wasn’t available, but we would assume that given their circa 80% loan-deposit ratio there would be a positive (lending) gap.
Given the extent of the banking system’s exposure to contracted rehypothecation chains, is it any wonder then that some institutions are beginning to seek out alternative forms of collateral, like equities and even less-liquid ABS, just to keep those chains going?
Which makes us wonder if the next phase of the collateral shift might be a return to the use of CDO tranches (though this time correctly priced, of course).
Oh the irony.