You Can't Understand 2008 Without Understanding the Eurodollar Market
If you want to understand the last decade (or three), and I mean really understand, there are very few places you can turn to. Since the panic in 2008, an enormous amount of time and effort has gone into identifying causes and trying to preview possible future solutions. All that did was to prove that the amount of attention paid to something is not proportional to the level of understanding it attains.
The reasons for that and for similarly aborted voyages of discovery are most often blindness, whether ideology or some other biases that very often cause us to dismiss what might be the key piece of evidence unraveling what at first might have seemed an impenetrable mystery. The scientific process of examining data and observation is supposed to be an open-minded affair, but human beings are human and often need some help getting outside the box. Complexity after finding this “key” no longer seems so dense and impassable.
In the case of economic and financial history, few are available that can explain both the panic as well as its aftermath. One of those that can is the eurodollar futures market. Part of the reason for that is what seems only at first a contradiction; it is, arguably, the largest and most liquid market in the world yet remains one that practically no one has ever heard of. Futures as a whole are something of a fuzzy mystery to begin with, but I’d wager that less than one in a hundred has ever come across the word eurodollar. That is true not just for laypeople going about their daily lives wondering what happened to jobs and economic growth, but also for almost the whole of the financial services industry that often purports to understand these things and their consequences. I won’t bother to describe Economists.
The CME Group (formerly known as the Chicago Mercantile Exchange, which at one time absorbed the CBOE) tells us that the current front month eurodollar futures contract, February 2017, had open interest as of Wednesday of 90,207 contracts. Since each contract gives the holder the right to a $1 million eurodollar deposit at maturity paying 3-month LIBOR for the cash, that translates into a possible $90.2 billion changing hands in a few weeks’ time. That obviously doesn’t happen to that degree, as most contracts are settled by other means than the creation of an actual deposit balance.
And it’s a good thing, too, because a February contract is an-off color month. Most of the business is done in the quarter-ending contracts of March, June, September, and December. The open interest for the March 2017 eurodollar futures contract is 1,449,252, suggesting $1.45 trillion in eurodollar deposits that could be demanded in just over a month from now. In fact, that one and the next four quarterly tenors are all over 1 million contracts each, meaning over $1 trillion at each interval, $6.4 trillion in combined total among just those five.
In short, there is enormous resource, depth, and liquidity being traded here. As such, it is one of the few markets that Federal Reserve officials have paid attention to, if at times reluctantly. I would write that they have paid close attention to eurodollar futures, but that wouldn’t be quite right. The FOMC deliberations have often considered them, but since 2007 it has been more of a strained relationship, where Fed members see eurodollar futures but do not appreciate what they might suggest. It was commonly believed that futures prices reflected expectations over nothing more than monetary policy, and therefore eurodollar futures, as deep as they are and for the enormous space that they cover, are likewise a reflection of only that direction of causation.
The irony of the story they tell is that is exactly how it all started out. Eurodollar futures were a place where the biggest and most powerful financial players bought whole-heartedly the Fed story. Even in early 2007, when things started to really diverge from the Fed’s point of view, the eurodollar futures market was not positioned to be against the Fed but more so to recognize risks that it felt the Fed would eventually have to, too.
In the first few months of that year, futures prices were bid up suggesting the market expected lower interest rates into 2008, which roiled the FOMC to no end because it truly did not at the time see any systemic risks arising from the small corner related to subprime (and therefore the Committee was sure they would not move from a neutral stance, so the eurodollar futures market was acting in sort of hyperbolic fantasy). This wholesale market, by contrast, was uniquely positioned to appreciate that the problem was not actually subprime, nor even, really, mortgages and housing.
Even throughout the whole crisis where the Fed failed time and again, toward the end of it the futures market remained faithful to core monetary policy philosophy. On December 17, 2008, the day that the FOMC voted for ZIRP and QE1, the eurodollar futures curve was alarmingly flat, suggesting that money rates would be low and remain low for the foreseeable future. From December 17 forward all throughout 2009, the curve turned right around and instead steepened very sharply. Though the front month contract was bid higher in December 2009 than December 2008, the back end had sold off remarkably. The June 2013 contract was trading at 97.35 in late 2008, but a year later the equivalent (in terms of time) June 2014 contract was trading at 95.06.
It was an enthusiastic endorsement of QE1 and ZIRP; the lower front month comparison simply a reflection of ZIRP’s introduction, where the lower price in the back of the curve signifying the assumed success of the operation. Higher expected money rates in the future speak to normalcy and economic opportunity, therefore the steeper curve was saying that this hugely important market believed those dramatic, “emergency” measures were going to be effective at getting the economy out of the Great “Recession.”
What was seemingly at odds before then, for why the curve flattened so much in the first place, can be attributed to the same doubts expressed all the way back in 2007. In other words, eurodollar futures traded more and more as if the Fed was only behind the curve, literally, in this case, and it wasn’t until the prospects for “money printing” in introducing quantitative easing that it was believed monetary policy had finally, after enormous damage had been done, caught up. It was thought our central bank lacked the will, not the ability.
At the end of December 2009, however, a new crisis emerged, or at least what seemed to be a new one. Though it started with whispers out of the Middle East, this next one was thoroughly European in nature. That was how it was treated and reported, as a European crisis. Once again, however, the eurodollar futures market was distinctively situated to appreciate that it really wasn’t; indeed, it wasn’t even truly a separate event from 2008, instead an ongoing evolution in the form of it.
Predictably, the eurodollar futures curve flattened enormously all over again, correctly anticipating the events of 2010. Later that year, the FOMC responded with a second QE, and even though the futures curve had shriveled while QE1 was still running, the introduction of QE2 in November 2010 had the same effect as the introduction of QE1. By late February 2011, the curve was nearly back to where it had been in December 2009 – suggesting that “lower for longer” in ZIRP and QE would still work, if just delayed in delivering the full recovery to normalcy.
What we can interpret, then, from this period and the behavior of eurodollar futures was truly in keeping with the spirit of QE. In other words, because the futures market rebounded upon announcement of the second, it seems very clear what had caused prior consternation was not, at that time, QE itself but questions about the size of it, a quite logical first place to look. The additional balance sheet expansion in late 2010 was received favorably as fixing that possible deficiency, an experiment in one isolated parameter.
That all changed in 2011. When a year after that in 2012 the FOMC voted once more for a third QE, this time the eurodollar futures curve didn’t budge. It wasn’t until later in December 2012 and the expansion of QE3 as both quantity of purchases and restarting them in US Treasuries (which was truly a QE4) that the eurodollar futures curve began to move steeper again. And even then, it was nothing like what it had been in 2009 or 2010-11.
The biggest move came not at its start but more so contemplating its end in the summer of 2013, during the so-called taper tantrum. It therefore provides us with a perfect benchmark for evolution in thinking and trading. If in 2010 eurodollar futures investors were to have believed that quantity was the defect in terms of QE, by 2012 they were no longer so sure that was the case. It wasn’t until the FOMC threatened to taper QE3 (and 4) because they viewed the economy as having strengthened sufficiently that eurodollar futures began to react as they had those prior times. If the curve had jumped on the announcement and thus by mere faith before, by 2013 it clearly wanted evidence first before reacting favorably (to the normalcy scenario).
Of course, it didn’t last long (again) and never reached those former levels of faith. On September 5, 2013, eurodollar futures broke forever from QE. Though QE3 and 4 were still running throughout later 2013 and almost all of 2014, it didn’t matter any longer. The total amount of flattening that took place in the nearly three years after that inflection was simply astounding, a level of pessimism and bleakness that I struggle to describe adequately. The curve even after this latest brush with “reflation” in the second half of 2016 remains a shell of its former self, a withered husk as compared to either December 2009 or February 2011, a final judgement that the global economy will not see normalcy in the extended future. And in the end, the Federal Reserve and its policymakers finally agreed with it, quietly surrendering to the logic of the curve last year.
In trying to piece together a meaningful review, there is no struggle to pinpoint where it all started to change – late July 2011. From that point forward, it was never the same, like the air being let out of a balloon. And the pin that pricked the balloon was a monetary one.
In accordance with its policy for a five year wait period, the Federal Reserve released last month the transcripts for FOMC discussions during the calendar year 2011. To say that I have been waiting for these particular records would be an understatement, perhaps to a degree even more than for the 2008 versions. I still urge, as often as I can, everyone to read those from 2008 because they clearly show that the Fed had no idea what was going on or what it was doing, a determination the eurodollar futures market traded contemporarily to those discussions. But both the Fed and eurodollar futures erred in how that was supposedly resolved, with “money printing”, as one of the two of them found out in 2011. Reading the transcripts, however, you wonder in stunned amazement how it wasn’t both.
On August 1, 2011, the Fed held an emergency conference call ostensibly to game plan how it might go should the debt ceiling votes not result in the expected outcomes. But there was more than that on their minds, as two of the twelve options presented to the FOMC that day involved, essentially, bailing out the repo market (both cash and collateral, both repos and reverse repos), which had acted alarmingly in the few days before that meeting. Though the media focused on the politics of the US debt, of course, especially after S&P downgraded it for the first time in history, the FOMC knew different
“MR. SACK. I would actually like to argue that the declines on Monday were not entirely driven by the downgrade. I think they were a continuation of the decline in sentiment and the concerns about economic growth that we had already seen develop very intensely over the previous week.”
Sharply negative economic sentiment had developed even though QE2 had finished only a few weeks before. The timing of these developments suggested that QE wasn’t having as much impact on the real economy as they had hoped. In fact, much of the discussion turned on that point, where optimism the FOMC members had expressed when QE2 was voted on just nine months before had almost entirely vanished.
It wasn’t just the economic side that troubled the discussions, of course, as severe funding strains had also re-emerged. The two, as the FOMC correctly surmised, were related. Rising risk in economic terms will always get translated in liquidity. Typically, a central bank responds to that situation by raising it, but that is what the Fed had just done (or so it thought) in the form of a second enormous QE. As Brian Sack, Manager of the Open Market Desk, remarked on several occasions, “And I think it’s worth pointing out that this is all happening with $1.6 trillion of reserves in the system.”
When the FOMC met again on September 20-21, 2011, a lot had happened over the intervening weeks. They never did bail out repo, and by mid-September repo markets were a problem as was FX, a fact the Committee had to recognize in the September policy discussion.
“MR. SACK. A few signs of pressure have even emerged for secured funding markets. It has become more difficult to borrow against less liquid collateral, with investors requiring over-collateralization and higher rates for some transactions. Moreover, there are some instances of investors cutting off secured funding against all types of collateral for particular European counterparties…
“In response to the intensifying dollar funding strains for European institutions, the ECB, the Bank of England, and the Swiss National Bank announced that they would begin offering 84-day dollar funding operations in mid-October, using the liquidity swap lines that are in place with the Federal Reserve.”
The FOMC could not help but notice a dramatic shift in the distribution of bank reserves created by the QE’s (the level of bank reserves are determined solely by monetary policy, but it is the market that defines how, or if, they are redistributed). European banks had before the crisis borrowed heavily in domestic money markets and shifted those dollar balances to their foreign headquarters where dollar assets were housed. After, the process was reversing and accelerating. The amount just from August 2010 to September 2011 was nearly $500 billion, an astonishing shift. The members and the Fed staff struggled to find an answer.
“MR. CARPENTER. I guess I would offer the following. The transmission channel that you described about the switch in the funding position of the branches relative to their headquarters overseas is consistent with the anecdotal reports that we are hearing. And you can think about the reasons for this working independently and then reinforcing each other, with reserves being an attractive asset, paying more than other perfectly safe things, and the foreign supervisors wanting there to be more dollar-denominated liquidity on their books for liquidity reasons. We’re definitely hearing that story.” [emphasis added]
And:
"MR. ENGLISH. Just one other thought. If you literally look at holdings of Treasury securities and straight agency debt at the foreign banks, they’ve actually been growing over the past year or so, right through August. I think that’s consistent with what Seth said. Some of these institutions are being urged by their supervisors to beef up their U.S. dollar liquidity, and so they may be holding both more reserves and more Treasury securities." [emphasis added]
And:
"MR. ROSENGREN. As money market funds and other investors have shrunk as a source of short-term dollar funding, the maturity of funding has shortened significantly, and in some cases, peripheral banks have been entirely shut out of short-term wholesale funding. The wholesale funding of dollar assets with increasingly short-term funds is all too reminiscent of the SIV problem experienced in 2007." [emphasis added]
The answer was right there in front of them, but they never made the complete connection because ideology prevented it. Again, as Mr. Sack exclaimed several times for emphasis, “this is all happening with $1.6 trillion of reserves in the system.”
The eurodollar futures market thought exactly the same way, that at first it was the quantity of bank reserves that would be fixed by the additions of QE2. Therefore, we can more easily understand why the events of summer 2011 were so damaging to faith in monetary policy, because, as Mr. Sack repeated, bank reserves of enormous quantity were not having the intended effects even in money markets where “money printing” surely would matter. From that point forward, eurodollar futures began to suspect more and more that the Fed wasn’t, in fact, printing money via QE’s.
The pieces of discussion I highlighted above (and many, many others) suggest the mechanism by which this evolved thinking was made. Banks primarily in Europe, but not limited to just those, were increasing their own dollar liquidity at a time when liquidity was supposed to be in systemic capacity beyond plentiful, an irreconcilable difference. We know that both repo and (perhaps more importantly) FX markets didn’t agree with the official determination (and then some), so there was real-time refutation to monetary theory in worldwide dollar practice (that reached all the way to the FOMC deliberations also in real-time as they debated a repo intervention). The banks increasing their dollar liquidity margins, which used to be called hoarding, would very likely have hedged these measures by among other things eurodollar futures.
The combination of the history of the eurodollar futures market plus the 2011 transcripts provides us with the whole narrative, or at least enough of it to be meaningfully comprehensive. Thus, the scale of reversal has been total – first if the Fed would act, then whether it could act, and now, a decade later, if it ever can. And to that evolution the FOMC has decided that the problem was never money but Baby Boomers and a labor market filled with drug addicts and lazy Americans (I’m not kidding). Of course they would come to that conclusion because to admit the problem was money all along would be to admit they failed in every way possible. In one 2011 breath, they would talk about all those bank reserves as if the world was filled with dollars, but then in the very next describe all the ways in which the world clearly wasn’t. Thus, they had their eyes all the time on the problem but could never see it, certainly not in the way the eurodollar futures market finally did, and as a matter of experimentally established fact.
The way in which the Committee reconciled these incompatible positions was in what Mr. Rosengren suggested with his discouraging allusion to the SIV’s of 2007. In other words, the Fed would treat wholesale funding as almost completely non-monetary in nature, as if it was something else altogether beyond the scope of their monetary policy. There was and remains some truth to the idea, as the wholesale system is surely different and in many ways separate, but that doesn’t make it any less money simply because they say so. Bank reserves are treated as the monetary base, but they are not it, and the economy of the last ten years has proved it without a doubt.
In 2007, the eurodollar market believed the Fed ill-informed but no less capable should it ever overcome that situation. By 2011, the market started to suspect that it was incapable as well as uninformed, as it turned out to be. Officials that year had no legitimate reason for not arriving at the same conclusion. It was all right there. The events of 2008 did what they did, and put us all on this course, but it was so-called European emergency that locked out any other possibilities. It was never a second crisis, but a disastrous extension of the first; a “dollar” problem that central banks hadn’t solved with QE, and one that by the end of 2011 made it clear they never could.
