The complexity of balance sheet mechanics, as observed in the prior discussion, offers a greatly altered view of liquidity and the ancient concept of elasticity. These are not just theoretical assertions, as we have seen them in practice on more than one occasion. The behavior of “liquidity” starting in August 2007 was exactly along these lines, as faults in dealer balance sheet mechanics were transmitted as illiquidity in the form of far fewer “risk units” available on market. That meant not just an unwillingness to meet unsecured interbank demand in eurodollars, but also an inward reflection of fewer securities available for loan in repo.
But perhaps the greatest deflection of “risk units” was in, again, derivatives and thus available balance sheet capacity to make trades, add risk and, most of all, keep prices orderly. Higher calculations of expected volatility meant higher VaR, maintaining a contractive impulse on risk absorption on offer. A greater VaR or cumulative portfolio vega means that the bank’s “capital” is at greater risk of a large negative swing (which is why volatility is so important, including its derivative acceleration, as is correlation). If more bank capital is at risk because of higher expected volatility, then the bank is going to be far less willing to take on new trades or risk positions without being “suitably” compensated. In the extreme, risk calculations may even require dumping positions and unwinding prior trades (or at least shifting trading desk mixes).
In practical terms, that meant, during the panic, nobody was willing to provide protection at a reasonable price at exactly the moment everyone wanted more protection not to expand positions but just to maintain existing trades. This had the effect of drawing liquidity into a self-reinforcing spiral, whereby the fewer “risk units” on offer led to pricing and correlation ill-effects, creating still more negative calculations for balance sheets to consider; so the “risk units” on offer contracted still further and so on and so on. In other words, it was a bank run of only banks, with the run not on cash or currency but risk units in various derivative forms (collateral, hedging, risk absorption, etc.). There was nobody left to provide this marginal liquidity, with central banks caught not just behind the curve but on the outside looking in, which is why the whole system completely froze after Lehman.
This capacity for risk absorption, provided by dealers in derivatives (and there are only four banks in the US, Citi, JPM, Goldman Sachs and BofAML, responsible for 92.6% of this business, which is dominated by interest rate swaps), cannot be overstated. I believe that we are right now experiencing another episode of exactly that, relating all the way back to QE 3 and QE 4’s operational issues of early 2013. As I wrote previously on this turn in funding and credit:
In other words, repo dysfunction tied directly to QE’s notorious tendency to strip out collateral created open disorder. Now to those participants in interest rate swaps, especially the dealers sitting heavily imbalanced on paying floating based on the idea that QE was forever, this was a stunning and potentially very upsetting development (though it should not have been based on April 2011; maybe dealers thought the Fed had learned an OTR lesson about interfering in OTR t-bills, which was what Operation Twist was all about). Dealers are well aware of internal plumbing considerations and likely would have connected the dots about QE – if it was going to be a disruption to liquidity it wasn’t going to be around forever as previously thought, or even all that much longer.
Like the mid-2000’s when dealers were providing “risk units” in the form of CDS protection, the one-way trade was built upon the central bank premise of an asset price “put.” In this case, dealers were taking QE in late 2012 to be semi-permanent (“open ended” as Bernanke described) and thus were taking fixed to an unbelievable degree (which is why the 10-year swap spread went negative, again, right at the announcement of QE3). That had the effect of increasing the “risk units” available in the liquidity system, and credit prices reflected exactly that.
By March 2013, the expansion of “risk units” on offer halted and then reversed due to modified expectations of QE effects. Undoubtedly, expected volatility played a central role in reigning in balance sheets. Higher expected volatility as dealers re-evaluated their positions meant less “risk units” available for trade, and thus lower systemic liquidity. By May, the reversal was solidified as Bernanke began to openly discuss what was only feared. So when credit prices, especially MBS, began to fall there was, like a mini-2008 episode, nobody left to provide “risk units” to maintain orderly pricing and positions.
This also gives us a window into 2014 and all the means in which last year was worse than the year before. As volatility begets volatility, “risk units” become more and more scarce making systemic liquidity less than uneven. Even at the dealer level, trying to calm VaR and vega means not just offer fewer “risk units” to the market for trade and liquidity but will also altering their own balance sheet mix – from positions (and desks) with more volatility toward positions with less, such as UST!
As I have said repeatedly, that is why I believe November 20, 2013, was vitally important thereafter in setting the financial course globally. The steepening yield curve was an important element in setting VaR and vega, thus “breaking” the systemic liquidity capacity – that day must have been the end of the road under the “QE is great” paradigm as volatility calculations maxed out for someone or multiple someones. Since that point, with QE still creating “reserves”, the “dollar” world began to “tighten” seemingly unalterable to everything the FOMC said or did (dispositive of the idea that “reserves” are nothing but a monetary byproduct).
What that means is essentially the systemic capacity to provide “risk” into the open market for funding and pricing has been diminished in the same fashion, though not at the same speed and magnitude, as the prior crisis – which was a “dollar” crisis precipitated by a reduction in “risk units” than anything of a traditional bank panic. The means by which that is expressed in 2014 and 2015 is different now than in 2007 and 2008, but the concept of modern “money” and what is “money like” remains.