The True And Hidden Menace of Liquidations
Today’s radical reversal in stocks notwithstanding, the continuing hits of liquidations are not achieving their settled ends. In purely financial terms, the entire process of liquidation is to renew a settled state. Local imbalances force restriction of financial resources (what used to be money but now is something recognizable as such but truly not money) which triggers a cascade of repositioning that either relieves the immediacy of the imbalance or takes another step further in repositioning. This is the snowball effect predicated a great deal on the scale and, perhaps more importantly, the composition of the leverage in support of it all.
Local liquidations are a troubling nuisance, but when they cannot solve the imbalance from within a localized financial space the risk of a more general liquidation rises. An analogy of our current circumstances would be junk bonds. Since junk bond positions are tied to other asset classes through the vehicles which have “invested” in them, should junk bonds as either collateral or information signals (about the soundness of the vehicle holding them) produce too much of a counter-effect, the tendency to liquidate just junk bonds is surpassed and spills over into other assets. Obviously, that has already happened.
This is why Ben Bernanke’s assertion to Congress on March 28, 2007, was dead wrong before it was ever uttered:
At this juncture, however, the impact on the broader economy and financial markets of the problems in the subprime market seems likely to be contained. In particular, mortgages to prime borrowers and fixed-rate mortgages to all classes of borrowers continue to perform well, with low rates of delinquency. We will continue to monitor this situation closely.
The point of liquidation is not credit loss or even credit risk, it is exclusively volatility. Loss is not predicated upon only default, as markets often demand current prices take those same qualities. Repo provides the perfect example of just such a reset. Prior to Bernanke’s statement, it was typical for high-rated private label MBS to trade in repo with a 3-4% haircut – meaning that for every $100 in par put up as collateral you could receive $96 or $97 in overnight funding. The haircut was in place for purely liquidity reasons having less to do with default than Bernanke could apparently imagine.
In other words, as a cash owner your goal in a collateralized lending program is to remove any risk of loss. If I give you $97 and you provide $100 in private MBS, with no subprime mortgages in it whatsoever, my concern is not whether there will be loan defaults within that MBS but rather if you don’t give me my cash back tomorrow at what price can I surely sell it to restore my cash position? A 3% margin of safety overnight didn’t seem particularly unreasonable since market-based measures of volatility prior to 2007 rarely demonstrated “tail risk” kinds of moves.
The danger of subprime in early 2007 wasn’t truly default – it was volatility knock-on effects in markets far and wide. Once liquidity started to drop purely in subprime, that created more volatile pricing which downstream models started to perceive. It didn’t matter that a particular MBS security had no subprime mortgages in the structure, what mattered is that 3-4% haircuts were now modeled as perhaps not enough safety in the overall MBS collateral segment. That is exactly what happened from 2007 through 2009.
According to ICMA, a prime, AAA-rated agency MBS traded at a 4% repo haircut in June 2007, just before the fatal shift in eurodollars. By June 2009, that same MBS would find instead a 10% haircut. That was a massive change, which caused selling to beget selling and so on and so on. For unrated agency MBS, again prime, the haircut that in June 2007 was 10% had moved to 30% or even 100% by 2009. And that was not the full extent of the collateral/liquidity problem, either, as certain strains of even prime MBS collateral became non-negotiable on any haircut terms.
For the collateral owner, some kind of investment vehicle, maybe a bank portfolio, rollover funding wasn’t an option it was the entire business model. That meant when haircuts were adjusted on your prime MBS from 4% to even 5% it left you needing to post another 1% of your own “equity” in order to roll that next day. If you didn’t have that, and many (most) did not being leveraged as far as possible, in every manner possible, you had to sell something. As the subprime MBS market became increasingly illiquid, that meant selling something other than subprime MBS – transmitting liquidity contagion far beyond subprime.
It, at some point, became self-reinforcing outward from localized liquidation into generalized liquidation (then the full version; crash) as selling creates more measured volatility which made repo counterparties demand greater haircuts (or different collateral altogether) creating more selling, and so on. Once it becomes so self-reinforcing it attains that ultimate downward spiral regardless of the kind of security. Repo haircuts on AAA-rated, IG corporates in June 2007 were 1%; even though these were survivor companies in little danger of default and having nothing at all to do with subprime “toxic waste”, haircuts by June 2009 were 8%! Even short and medium-term government bonds (G7) which had traded with zero haircuts before 2007 suddenly found at least 1% haircuts by the time it was all over.
The point of liquidation is to find a settled state where the cascade of illiquidity-driven negative processes end; they stop being self-reinforcing, an equalization where liquidity-providers no longer need additional collateral or whatever kind of assurances in prices, modeled volatility or qualitative factors that allow the liquidation wave to runoff. I bring up the MBS repo example not because I expect so much of junk bonds in the past few years have been used in repo collateral regimes and arrangements similar to subprime mortgages, though that can’t be discounted given the renewed rise in CLO’s and leveraged loan syndications of all varieties, rather it offers a useful and historical example by which to examine possible liquidity structures and processes under duress.
These liability chains do not, as Bernanke supposed then and Yellen likely does now, behave as purely assets. The repo implosion was in every way like a “run” where unblemished, primary collateral was the derivative “currency.” The week after Lehman, FRBNY reported $3.544 trillion in UST repo fails, up from $577 billion the week before; by the third week in October, $5.305 trillion. Not coincidentally, the stock market crash occurred within that period.
The problem wasn’t so much “toxic waste” as it was artificially low volatility in the years before the crash. That was taken as the liquidity baseline and used to set vast liquidity and funding regimes as if it were all real. And that is what concerns me now, as that was exactly the behavior in especially junk, but not limited to it, in the past few years – particularly since QE3 in 2012. I believe the world was awash in artificially low volatility regimes which are being forced, right now, into re-assessment. That includes not just junk bonds but the very structures of the “dollar” and money dealing itself; from IR swaps to almost any chained bank balance sheet liability traded , thus providing liquidity, under low volatility assumptions.
What is even more concerning is that even after all these liquidations, and August was not the only general version, it was just the first that hit the US, there are still more coming. In other words, past liquidations were not enough to find that settled state to end the run. The longer it takes to trudge through self-reinforcing illiquidity, the greater the probability of “tail risk” – which is nothing like a tail risk.
I wrote on December 11 that (subscription required):
The net or sum total of all these indications [“dollar” conditions and proxies] appears more and more unified in the direction of another liquidation. This is not to say that it is a given or that such an event is the only possible outcome; far from it…What we can interpret of these “dollar” conditions now, especially those along the Asian “dollar”, is that the probability of more extreme volatility, including something like August, is perhaps as high as it has been since August.
From that perspective, the “dollar” warning level is quite high.
After all that has happened so far in 2016, and that is a lengthy catalog, the warning level remains quite high by my current judgement (subscription required). It’s as if nothing has yet been solved even at these greatly affected general price points and liquidity levels. That is what should preoccupy us from this point forward. One factor that shouldn’t is all these continued calls that oil is about oil; it isn’t. Black gold is also “dollar” collateral.