Suffering the Dead Money of QE and LTRO
There still persists in the minds of both investors and policymakers a need to believe in the printing press as universal financial salvation. First it was quantitative easing (QE) in the US. In the past few months the European Central Bank (ECB) has upped the debasement ante by twice engaging in Long-Term Refinancing Operations (LTRO), much to the unimpressively unbridled enthusiasm of asset prices. By creating and dispersing digital euros directly into the banking system conventional wisdom has settled on these 3-year operations being the definitive answer to the ongoing "liquidity" problem in Europe - if banks are full of cash, so the thinking goes, there won't be any need for desperate firesales that transmit the contagion of insolvency to every market around the globe.
Much the same thinking went into the Federal Reserve's second iteration of QE. They believed that if they "printed" $600 billion in brand new digital dollars it would solve the chronic dollar shortage that became a systemic problem in 2007. With so much liquidity in dollars, another liquidity crisis shouldn't have been possible.
What QE did, rather than solve any liquidity problem, was prove that policymakers might not have really understood the mechanics of the tribulations that they have been confronted with (this imbalanced thinking also applies for their understanding of the real economy as well as the financial economy). The Fed simply assumed there would be a wide dispersal of those new dollars into the wider banking system, expecting particularly an infusion of dollars into the eurodollar blackhole. Further, as Federal Reserve board members and a bevy of contemporary research papers attested to, monetary policy in the QE period based its entire expectation for stimulating economic growth on the assumption that these dollars would eventually circulate beyond the banking system into the real economy, ending up as mild, "beneficial" inflation.
That never happened because traditional monetary policy mechanics depend on the primary dealer network to actually accomplish the distribution of any monetary largesse. Primary dealers in early 2011 were not all that willing to lend out money in any market to any financial institution. The lack of monetary transmission was especially acute toward the end of QE 2.0. Judging by repo market and US Treasury bill rates, there was a desperate shortage of "quality" collateral to grease the interbank wholesale money market wheels (quality collateral means both a widely accepted perception of low default probability and, perhaps more importantly, a liquid market for it - in dollar denominations that means shorter-dated, on-the-run US treasury debt securities).
Because of these mechanics of monetary transmission, QE 2.0 created an interesting near-paradox in the dollar segment of wholesale money markets. Since the Fed was the most active "buyer" of on-the-run short-dated US debt through QE, it crowded out other potential users (especially since it appears as if the Fed was overly generous in its terms). Essentially the Federal Reserve was hoarding all the quality collateral, generating negative repo and t-bill rates as dollar market participants scrambled to find suitable replacements. In trying to solve a liquidity problem, the Fed created conditions where that solution was nearly impossible. Thus the near-paradox - the Fed created money, but in doing so limited its own ability to transmit that money beyond its foundation of primary dealers. That is why we got Operation Twist last autumn instead of QE 3.0, a belated recognition that the mechanics of monetary policy was indeed the actual source of monetary impotence.
By summer 2011, the costs of dollars began to rise all over the globe, especially dollar swaps (even LIBOR got into the mix). By September 2011, there was such a diminishment in dollar circulation that the Fed re-activated its swap lines (though at a penalty rate of 1%). That didn't work, and the crisis deepened into December. By December 8, 2011, rumors of imminent failures in large French and Italian banks were largely substantiated by central bank emergency interventions on both sides of the Atlantic, covering both dollars and euros.
The Fed, for the dollar segment, reduced the penalty rate on those dollar swaps by half. Since December 8, swap usage at the Federal Reserve has grown to, and persisted at, more than $107 billion, not nearly the level of 2008, but an embarrassment for a monetary policy that created so many dollars experts and policymakers believed a near re-run of 2008 to be fully impossible.
For its part, the ECB announced the LTRO's also on December 8, 2011, with what looks like an attempt to take into account some of the shortcomings of QE. First and foremost, the ECB money was dispersed directly to any institutions that placed a bid, foregoing the possibility of repeating the primary dealer problem. Second, the collateral terms were intentionally loosened, likely an effort to try and avoid the problem of collateral shortage. From the ECB announcement:
"To increase collateral availability by (i) reducing the rating threshold for certain asset-backed securities (ABS) and (ii) allowing national central banks (NCBs), as a temporary solution, to accept as collateral additional performing credit claims (i.e. bank loans) that satisfy specific eligibility criteria. These two measures will take effect as soon as the relevant legal acts have been published."
By February 9, 2012, in time for LTRO 2.0, collateral terms were, in fact, weakened:
"The Governing Council of the European Central Bank (ECB) has approved, for the seven national central banks (NCBs) that have put forward relevant proposals, specific national eligibility criteria and risk control measures for the temporary acceptance of additional credit claims as collateral in Eurosystem credit operations. Details of these specific national measures will be made available on the websites of the respective NCBs: Central Bank of Ireland, Banco de España, Banque de France, Banca d'Italia, Central Bank of Cyprus, Oesterreichische Nationalbank and Banco de Portugal."
Only seven NCB's chose to enervate collateral terms: Ireland, Spain, France, Italy, Cyprus, Austria and Portugal. That list pretty much defines the PIIGS crisis. Noticeably absent are the Bundesbank in Germany and the De Nederlandsche Bank in Holland (nor do any of the Scandinavian central banks appear). Clearly these mechanical "improvements" were aimed at banks in the troubled area, meaning that market events were making it difficult for those banks to pledge meaningful and accepted collateral in the wholesale money markets. The ECB either knew for sure, or were reasonably assured, that without loosening these collateral restrictions there would have been (or likely were already in the beginning stages of) another run on "quality" bonds akin to the mid-2011 run on US t-bills.
With these measures in place it seems as if the ECB can succeed where the Fed came up short, and certainly asset prices across the globe have responded as such. However, there remain multiple problems with this liquidity approach. Even if we set aside the more basic insolvency conditions at both banks and countries, there is enough reason to believe that this "improved" liquidity measure will still end up being as counterproductive as QE 2.0.
Global banks first bought into the 2010 "bailout" package with the idea that the Eurozone policymakers would be able to avoid a default, and were further convinced that the ECB would always intervene to artificially support specific bond prices. Indeed, that is what happened in 2010 and part of 2011. But to what end? As MF Global found out the hard way, there really was a practical limit to how far these measures could and would go, and that there was an obvious element of intentionally understating the severity of the crisis. What these initial monetary measures really accomplished was to suck the banking system further into the vortex of the general solvency issue of the countries involved. I have not yet seen any estimates on how many PIIGS bonds were purchased simply because of ECB price manipulation, but it is reasonable to assume that it was a non-trivial amount. The general monetary re-assurance that the ECB (through asset price manipulation) and Eurozone (direct, though conditional, fiscal aid) provided turned out to be nothing more than poison to the banks that were gullible enough to swallow it.
Given that the underlying solvency is neither resolved nor has been given a clear path to a believable resolution, how does this latest monetary schematic really differ from these previous episodes? If anything, banks are being herded systematically in the same direction at the same time yet again. Worse, LTRO 2.0 was accomplished through loosened collateral restrictions in the very countries that should be far more concerned about weeding out the weak firms from the banking herd. In short, LTRO 2.0 actually pushed Euro banks into the very securities they should be actively avoiding (as bond yields have shown in advance of the LTRO's), the same kind of poison that cost them the first time when they were suckered into believing in the infallibility of ECB bond price supports and general government bailouts. Rather than alleviating the liquidity situation, the ECB may have just guaranteed that another liquidity crisis is just around the corner.
Given the reality of the collateral situation in Europe, this next possible escalation of the crisis will occur with general liquidity conditions almost certain to be worse, not better. As much as economists talk about velocity in the real economy (the general pace of the circulation of money) there is also a derivative velocity within the interbank money markets that is dictated by the ability of banks to rehypothecate collateral securities. When collateral disappears because it is being pledged at various central banks, it is taken out of the rehypothecation chain (unless central banks are engaging in rehypothecation themselves, not likely, but you really never know in this day and age).
When the Fed engaged in QE 2.0, it not only suppressed the available supply of t-bill collateral, that diminishment was amplified by the "velocity" of collateral that would have been re-pledged elsewhere (rehypothecation is statutorily limited in the US, but banks operating in the eurodollar arena have no restrictions, including subsidiaries of US-based banks, which is why you often see banks in trouble, Lehman and MF Global for example, transfer accounts and collateral to their European subsidiaries in a desperate attempt to squeeze out incremental funding for each collateral security). Take one bond out of that collateral circulation and multiple avenues for liquidity are extinguished all at once.
In July 2010, the IMF attempted to quantify this collateral velocity, estimating that:
"When we add U.S. banks data together with large European banks with significant relations with the hedge fund industry, such as Deutsche Bank, UBS, Barclays, Royal Bank of Scotland and Credit Suisse , the total available pledged collateral was over $10 trillion at end-2007.1 Roughly $1 trillion AUM of all hedge funds were rehypothecated and hedge funds contributed about 40 percent of all pledgeable collateral received by the large banks; thus the churning of collateral could have been around a factor of 4 as of end-2007. More recently, as of end-2009, the churning factor has also declined in line with the total available pledged collateral."
While collateral velocity has fallen recently as hedge funds have become more aware of this particular rehypothecation vulnerability, the velocity is still likely greater than one, and likely some low multiple. I believe that is why QE 2.0 had such an outsized negative effect on the liquidity condition in dollar markets well beyond just the nominal amount. The same is likely true for the LTRO's, some multiple of collateral posted with the ECB in participation with these liquidity programs is being withdrawn in the traditional interbank markets because of the now limited ability to rehypothecate.
Given this likelihood, it is not really a mystery why so much of this newly generated liquidity finds its way right back with the ECB. At latest count, over €800 billion sits idle in the ECB's bank deposit program, having been returned by the banking system directly to the ECB. It's not just that banks don't trust each other and won't lend to each other on unsecured terms, there just isn't enough quality collateral to allow for all this money to move and circulate: interbank velocity has likely fallen in greater proportion to the increase in the stock of money. In other words, these monetary liquidity programs make potential liquidity conditions incrementally worse, not better. The ability to absorb new shocks has been diminished, not strengthened.
That leads to other after-effects, including the hoarding of what quality collateral is left unencumbered, further dampening the vital circulation of money that marks a healthy banking system. As circulation drops, the probability of a systemic event expands dramatically. Couple that with the fact that banks are now herded into the same place at exactly the same time, and the entire system becomes that much more unstable.
The only way the LTRO/QE paradigm can work is if the quantity of money taken up by the system is more than enough to cover any and all contingencies. In the case of QE, the Fed set a hard number that proved wholly insufficient. The ECB has allowed banks themselves to "bid" whatever amount banks estimate, but that is far from a perfect approximation of what banks really might need should conditions further deteriorate. Given the shortage of collateral that existed prior to the LTRO's, there is no guarantee shaky banks were able to acquire sufficient collateral today to fund up to their full estimation, let alone all potential scenarios. On top of that, the LTRO's come with a "penalty" rate attached, a rate that is floating and not fully certain:
"The rate in these operations will be fixed at the average rate of the main refinancing operations over the life of the respective operation. Interest will be paid when the respective operation matures."
While the ECB is giving the banks a free pass on a cash basis (they pay the interest cost at maturity ), they will still accrue that expense as a liability on their balance sheet. But the real matter here is that the LTRO rate, which is replacing lost repo funding, will be significantly above what banks would have paid into the repo market. It is a penalty rate not dissimilar to the Fed's current views of the discount rate (which was changed to a penalty rate 100bp above the Fed funds target rate for "primary" participants, and 150bp for "secondary" participants), designed to explicitly discourage banks from using this workaround method to fund operations above some bare minimum. Judging by the fact that the Fed was forced to reduce the penalty rate on its dollar swaps from around 1% before any foreign banks were actually able to use them, we can infer that the similar ~1% cost of the LTRO's is also a penalty to participants. In that event, it is not likely that banks used the LTRO's to fund every contingency (and perhaps not even mildly adverse contingencies - but that is speculation on my part), leaving the banking system still short of "full liquidity" (just how short might be determined in the coming month[s]).
Where this really becomes dangerous is in the context of diminished interbank velocity. The interbank market, when functioning well, is essentially money that can and does flow to where it is needed most. Flow is what is important, especially in times of stress. ECB money, just like the Fed's QE money before it, is dead money . It does not flow anywhere, hoarded like any inventory of leftover quality collateral. Worse, the growing proportion of dead money encourages even more hoarding, a self-feeding loop that leads to the liquidity desert - money everywhere and not a drop to circulate.
From a wider perspective, what these central banks have accomplished through these massive interventions is to remove the key to monetary flow (locking up quality collateral) in exchange for dead money stock. Dead money does nothing for liquidity conditions. A liquid, healthy system is one where money is both plentiful and available to move, a high degree of systemic flexibility. Any bank that pledged collateral in either LTRO no longer possesses the ability to rehypothecate in an emergency, a significant loss of funding flexibility (I am certainly not defending the practice, just making an observation about the system as it is currently construed).
In an October 2009 working paper, the Bank for International Settlements analyzed some of the key triggers to the 2008 panic. What they found is no surprise to anyone paying attention to the marketplace at that time, but some of the paper's conclusions are eerily relevant to 2012. In finding that US dollars were chronically short (by their estimate of between $1 and $2.2 trillion in 2007), the BIS concluded that:
"What pushed the system to the brink was not cross-currency funding [the dollar shortage] per se, but rather too many large banks employing funding strategies in the same direction , the funding equivalent of a ‘crowded trade'."
Given what we know of the current state of the banking system and all the central bank interventions suffered in the past three years, I think it is safe to say that the banking system is still not only employing funding strategies in the same direction, the scale and pace has actually risen - exhibiting a dangerous degree of self-similarity, meaning a system that remains in a critical state. As ZeroHedge pointed out on Wednesday, the margin calls on collateral posted with the ECB have spiked dramatically recently, suggesting without ambiguity that the 2008 crisis trigger (haircut adjustments on posted collateral) is still very much a concern in the parallel sovereign PIIGS space. Given the substitution of dead money for potential interbank velocity, the possibility for realizing a second round (or third if you view 2011 as such) of the "crowded trade" is not only real, it has, in my opinion, gone up significantly.
The LTRO's may be an improvement on US QE, but that is a low standard of comparison and really doesn't say much about real progress. What has not changed, and what is really important, is that the ECB is replicating the dead money of QE, poisoning the very system they are trying to save at all costs.
