October Treasury 'Crash' Shines a Light On a Witless Fed
It was nice to see that the FOMC considered US treasury bond trading on October 15 at its October FOMC meeting. More than a month later, there still isn't much by way of consensus on what caused the "crash" nor how to interpret the events of that day. Even the "crash" itself was odd in that what occurred was not a precipitous drop in price, but the exact opposite. In terms of yields on treasury bonds, it was there where the usual convention about such things related.
To say it was an extraordinary event is to be severely understated, though attention is limited outside the immediate rungs of the financial credit world. In a little over an hour, the 10-year treasury yield plunged 30 basis points. Like with other financial indications in such times, that doesn't sound like much but it represented a massive and volatile action that has left the "largest and most liquid" market in the world searching for answers.
At 8:29am, the 10-year was yielding 2.16% and was down only slightly in futures trading ahead of that day's retail sales report (which would be "unexpectedly" weak). By 9am, the yield was threatening to fall under 2%, which was not just psychologically important but a level where a massive number of technical and preset triggers were located. As it was, the initial breach of 2% did not unleash a "buying panic" but one followed not long thereafter as the low yield was hit at 9:39am at just 1.86%.
The selling then came forward at that point and by the end of the day the 10-year yield was about back almost to where it started, closing down just 2 bps at 2.14%. Such a V-shaped occurrence immediately brought back memories of US stocks on May 6, 2010, the so-called "flash crash." And so commentary, especially mainstream and orthodox, has settled on computer trading as the proximate cause of such "illiquidity."
It may sound strange, but a "buying panic" in US treasury bonds is a form of illiquidity. The modern shadow system has in many ways inverted or at least distorted what could fairly be called "currency" or a "dollar." As I have said on many occasions, repo collateral is in many ways "moneylike" in its behavior and properties. At times, collateral becomes far dearer than actual currency, or even ledger balances of currency, and thus behaves more "moneylike." Such derivative function makes straightforward analysis extremely difficult, especially to the outside.
That is why I find the electronic trading explanation so lacking in its comprehensiveness, but at the same time I understand why it has persisted. There is a mechanism in UST trading which is supposed to mitigate any such supply and demand imbalances, as the role of primary dealers is to see such "arbitrage" and use their immense balance sheet power to restrain it and turn it back toward "equilibrium" (if such a thing exists more than brief and fleeting moments). That clearly did not happen, so the presence of electronic order systems and even high frequency trading does not explain, at all, the absence of dealers.
As it was, the 10-year bond was not the only security undergoing rapid ascent, as the entire treasury complex (aside from t-bills) was unusually bid against offerless conditions. But the unusual exhibition in UST that day was actually tame in comparison to other places.
"Corporate-bond investors have struggled this week to find trading partners for some large orders, causing unusual price drops and raising concerns that trading could freeze in future market turmoil.
"'Buyers just disappeared' early Thursday for many low-grade bonds and even some higher-grade ones, said Jason Graybill, senior managing director at Carret Asset Management LLC, which oversees $2 billion."
That Thursday was actually October 16, the morning after the affair in the treasury market (and to which the Wall Street Journal ironically attributes, to all things, the stock "swoon"). So the day after there is an enormous scramble for treasury bonds, which are now the primary form of collateral, there are no bids in corporate credit, even investment grade. Further, as we know from other anecdotes, the inability to trade in anything other than massive size continued for some days thereafter. All of which means that for some period around and after October 15, dealer and even financial ability to maintain flow and leverage was severely impaired - none of which has anything to do, again, with electronic trading.
If we look beyond the settling convention, the events start to make sense. A huge rush of buying right at the open indicates, to me, collateral calls either from the previous day's positions to be settled or, more likely, counterparties thinking ahead toward settling that day's collateral imbalances before 2pm. The reason there was such a rush at that moment had nothing to do with the retail sales report or anything that was "news", but rather the crush of reverse pricing in illiquid corporate credit (among other "risky" places) that had been building for some time prior to October 15.
We have very little direct knowledge of pricing in the more illiquid parts of these credit markets, something that has been little changed from pre-crisis to post-crisis. In 2007, pricing irregularities were drawn toward structured mortgage finance while in 2014 it is very much a creature of junk corporate debt. That includes not just high yield bonds that trade sporadically on the secondary market (which is even more distorted by the rise of bond funds as the primary vehicle for attaining junk bonds), but something called leveraged loans. Leveraged loans are a form of junk debt, or "high yield", that are loans rather than bonds, typically syndicated across numerous financial participants.
By the close of trading on October 14, the S&P/LSTA Leveraged Loan Index had already seen some heightened negativity. The "market value" portion of the index, which separates par prices from interest receipts, was down almost 55 bps in two days (again, these are not "normal" movements, especially to the downside). Going back to September 19, that proxy for the market value of leveraged loans was down just about 1.3% in a streak that featured very little upside. To give you a sense of scale and comparison, the market value portion of the Leveraged Loan Index was down a total of 1.8% for the entire 2013 selloff!
Since that index includes only the 100 most liquid leverage loan tranches, we can fairly infer that pricing difficulties were probably much more volatile where visible trades simply do not exist. If selling and fewer bids are evident in the most liquid segments, then you can be sure that there is an even greater bid/ask spread elsewhere and more volatile movement elsewhere. In short, the most leveraged parts of the credit markets were seeing regular and significant selling right into October 15.
That is the important point to be made, in that these credit positions are leveraged, and highly so. As prices, even of proxies, fall at a steady pace, the margin declines until it reaches a breaking point. At that moment, which looks to be the days just before October 15, counterparties must settle these price imbalances by either selling their positions in size or posting "good" collateral. As it is, these kinds of events are self-feeding, gaining steam and strength from the very start unless there is a counterforce to arrest the trend.
From that we can infer that corporate declines, as well as leverage being applied in the treasury market itself, including short positions, were taking collateral calls all through the early part of October (tracing back to somewhere not just in September, though that may be where a critical mass was reached).
If you go back and look at various financial markets in recent months, the date July 3 and the first week of July sticks out almost everywhere. The Russell 2000 small cap index, for example, reached almost an all-time high that day, failing by less than index point, representing a near-term peak. It was also the week that the German DAX index peaked, as the Russell and DAX have "somehow" correlated in replicating each other very closely. We also saw an inflection in inflation breakevens in the treasury market, which have turned steadily lower since then in direct opposition to the Fed's proclamation that the economy is moving toward full recovery. July 1 was also the peak in the market value for the Leverage Loan Index.
The only factor that can tie all these disparate and seemingly unrelated financial markets together so neatly and tightly is the "dollar." Sure enough, that week the "dollar" began its most recent "rise", to which so much commentary is devoted to its assumed "strengthening." This relates back to Bernanke's ludicrous assumption of a "global savings glut" that he first suggested in 2005. His belief in that goes a long way toward explaining the FOMC's complete and total failure only two years later, as a policymaker dedicated to the idea of a savings glut is going to be completely surprised by the size, speed, composition and scale of what is really a global dollar short.
The global dollar short in this context is nothing more than a system of leverage and liquidity that exists outside the domestic "dollar" perception, but that is also in many ways closely tied to it. In more recent years, such as 2013, that is why the threat of taper was so disastrous globally, as the eurodollar market (the epicenter of both the dollar short's existence and thus the 2008 panic) essentially transmitted a liquidity shock all over the world. Inside the US, that meant a treasury curve that both steepened and rose in terms of nominal yields (which is when I pondered if the Fed had lost control over the UST curve, an ominous note).
"Something" has drastically changed in the eighteen months since the curve actually acted as it was "supposed to." I have described on numerous occasions the still-unexplained activities and anomalies of November 20, 2013, which, in this context, stands out all the more. That is especially true as the UST curve has acted exactly opposite its 2013 behavior - steadily and bearishly flattening while nominal yields decline (particularly at the long end). And when Janet Yellen, in her first official press conference after the March 2014 FOMC meeting, threatened to raise interest rates and end ZIRP, the UST curve burst not upward and steeper like it had after Bernanke's threats in May 2013, but exactly opposite with sharply lower yields and flatter, as if the FOMC had indeed lost control over bonds in the opposite direction.
In that respect, bond markets do not seem to be very much enamored with the uniform and ubiquitous central bank narrative that is replicated worldwide. In the US, the Fed keeps saying it sees recovery when, since at least November 20 of last year, credit markets do not. Since European banks are still very large participants in the global dollar short, their financial presence offers a window and gateway into what might otherwise be solely European "problems."
That offered a tantalizing clue about when the ECB succumbed to desperation on June 10 of this year. The "narrative" for recovery was well-established in Europe too, but by June such confidence had already waned as reality away from monetary models left little room for anything but the third recession since this all started in 2007. On June 10, the ECB, undoubtedly trying anything to spark "normalcy" in loan creation, pushed its deposit rate floor to a negative nominal rate. It was the first time in history of such a major step, and undoubtedly credit markets took note.
That same week, fails in US$ repo markets suddenly and sharply surged upward. Repo fails the week of June 4, 2014, totaled (both fails to receive and fails to deliver) about $130 billion, which was completely in-line with repo mechanics since the last surge in repo fails in 2013. The fails events in 2013 were tied to the massive selloff in credit markets surrounding, again, taper threats but also the global dollar short growing too "short." The week Mario Draghi and the ECB went nuclear with the deposit rate, suddenly US$ repo fails ignited to $388 billion, or the highest (worst) level since the weeks leading up to the euro crisis in 2011 (which was as much another "dollar" crisis).
Worse than that, repo fails remained elevated the following week, that of June 18, nearly touching half a trillion, confounding not just the 3% fails penalty but also conclusively demonstrating the Fed's reverse repo program as totally inadequate. During this elevation in repo fails, total repo volume especially in UST trading and liquidity also rose sharply after declining precipitously since the 2013 credit "event." The combination of the rise in volume and the surge in fails more than suggests that systemic liquidity, as a feature of shadow finance surrounding the global dollar short, was unable to cope with what looks to be a paradigm shift.
The timing here suggests an amplification of the trend that began on November 20, 2013, in the US treasury flattening. If the ECB's June 10 action to a negative deposit rate was received systemically in Europe negatively (either as an interpretation of heightened economic risk or even just as a negative commentary on the state of the euro; negative rates being unfavorable in pure pricing terms), it stands to reason that US$ repo might be used to absorb that "liquidity" function however it shook out. The net result of tight collateral conditions (which are unquestionably related to the QE's), however, undoubtedly had negative implications downstream in the global dollar short.
So we had, in succession, a growing bearishness in US credit followed by a confirmation about Europe on June 10 that then stressed "dollar" liquidity in repo, followed closely by one of the most intense and sustained "dollar" surges outside of the panic in 2008. A surge in the dollar, as the position in 2008 suggests, is not "strength" but "short." In other words, US$ leverage was receding throughout the entire time, and thus the lack of US$ liquidity caused all those global markets to correlate to the downside when global liquidity fell off. Selling beget selling, which was really "dollar" participants reigning in their balance sheet capacities toward offering "dollars" in the most "risky" markets. The wholesale funding model, or what Bernanke believed as a "global savings glut", turned inward and began to erode until it finally broke loose on October 15.
The trend of higher repo market fails that took place in June never abated, as the gyration of fails continued regularly all the way until, of course, the week of October 15. Thus we have good liquidity indication of exactly that type of problem, especially in operation following October 15 as repo fails have, for the first time since August, been free from such angst for consecutive weeks.
That only adds to the suggestion of what went wrong that day, since the steady problem of repo fails is an unmistakable instance of collateral shortage. Since that was met, on the morning of October 15, by what looks just as unmistakable as a collateral call, you can see how it would turn so drastic so quickly - pre-existing shortage, sudden surge in demand, netting into "panic buying." The confluence of all these events, and the timing of them, is highly suggestive of a systemic liquidity problem running globally until it finally broke out of prior constraint in an action that was simply breathtaking in size and scale for its unappreciated disruption. Since repo markets have been calm since then, we can infer that collateral positioning was at least satisfied, which can only mean a lower aggregate state of liquidity and leverage.
While US stocks only marginally were aware, markets all over the world have yet to see anything like a rebound (and that includes the Russell 2000). There are still indications of corporate credit irregularities, though not nearly as severe, which indicates exactly that altered and reduced state of financialism.
As I mentioned at the outset, the FOMC in October discussed this at their meeting, though we have no idea what was actually said or presented apart from the highly sanitized minutes just released Wednesday. For their part, they were far from re-assuring:
"Many participants commented on the turbulence in financial markets that occurred in mid-October. Some participants pointed out that, despite the market volatility, financial conditions remained highly accommodative and that further pockets of turbulence were likely to arise as the start of policy normalization approached. That said, more work to better understand the recent market dynamics was seen as desirable. In addition, a couple of participants noted the potential usefulness of collecting additional data on wholesale funding markets in order to better understand how changes in interest rates could influence those markets."
In other words, nothing has changed since 2005 as the Fed still has no idea about wholesale funding, somehow still stuck in that "global savings glut" mindset that misled them so wrongly these past seven years. Apparently, now in 2014, they find it "useful" to tell some staff economist to run some regressions on wholesale funding while they stick to the same "narrative" that credit markets globally no longer believe.
And that is really the major point to be made. If liquidity and leverage have declined since June, and done so so much as to finally break even the UST market in a (inverted) quasi-panic, that is a huge caveat about perceptions and expectations in opposition to the overly sunny disposition that dominates orthodox discussion and policy settings. If eurodollar and global dollar short participants are cutting back liquidity and leverage on offer, that is not because of electronic trading or even Dodd-Frank's increased cost of capital for dealers to carry needed inventory (in order to act as dealers), though those factors are still important.
In short, systemic leverage is as much about willingness toward risk, and thus if leverage is falling backward then so is risk. To some that may not be a big deal, or is at least expected as supposedly the Fed "exits", but this is something that has not been seen on this scale since 2011, if not 2007. Further, the manner in which it manifested more than suggests not purely financial factors. This is a warning, so much so that the Fed had to actually acknowledge it in empirical opposition to its previous proclamations about market "resilience" and confirm, once more, they don't know what they are doing.