The FOMC Begins To 'Lose Control' Of the Curve
Modern finance is perhaps the most complicated jigsaw puzzle ever devised. We like to think monetary applications are simple relationships, as in the relationship of monetary policy to the economy. The FOMC decides to reduce its rate target, the amount of borrowing increases and leads to better economic fortunes. Deep inside those assumed correlations, however, hides an unwieldy evolution of both function and form.
In the middle of 1981, for example, the World Bank had a problem. The institution's main function at the time, though we know it today in the middle of the European mess, was to borrow funds in developed markets and lend to developing nations. It was a true intermediary, performing both risk and rate transformations on behalf of the emerging world. But in 1981, the bank had maxed out its operations in the Swiss and German markets.
That left it with the unappealing alternative of US dollars. The World Bank could source US dollars at about 17%, but it wanted the 8% rate it might receive on Swiss borrowings had it not already exceeded its borrowing cap in Switzerland. It had also done so in West German marks, even at the relatively unappetizing 12% rate there. Enter Salomon Brothers.
Jon Rotenstreich, a partner in the fixed income division, and David Swenson, a banker in the corporate finance division, approached IBM with the World Bank's conundrum. IBM had similar problems in the opposite direction - it owed hundreds of millions on debt priced in both marks and Swiss francs. Concerned about appreciation in both currencies, the company was eager to hedge its currency risk, but did not want the cost to retire the debt outright.
What Rotenstreich and Swenson proposed was essentially the first derivative swap of note. These kinds of swaps had been around in nascent form as finance grew global in the 1960's and 1970's, but the derivative transaction proposed here was something new and revolutionary.
The World Bank would borrow $290 million and swap that dollar obligation with IBM. The World Bank would pay IBM's franc and mark-denominated interest payments and amortizations, while IBM would pay the World Bank dollars for its obligations. IBM removed its currency concerns while the World Bank sourced better terms in Swiss francs and marks without upsetting capital limitations.
The fact that two entities of this size and stature had embarked upon a novel financial process gave the swap credibility. Salomon Brothers would develop the "product" further, beginning to issue and market the first interest rate swaps of any scale in 1985 (a seminal year for debt saturation).
Before swap markets could really take off, there needed to be more clarity on financial terms. In a plain vanilla interest rate swap you need to price both the fixed and floating legs. The fixed leg is easy; it's fixed, so it prices much like a bond. Estimating floating payments is something altogether more difficult.
In December 1981, the Chicago Mercantile Exchange offered its first eurodollar futures contract. According to CME history, eurodollar futures were not readily accepted. It quotes Institutional Investor magazine scoffing at the notion, "Eurodollar contracts do not appear to have much of a future...Five months after their noisy launch on the Chicago Mercantile Exchange, Eurodollar contracts still haven't caught on." That sentiment proved very wrong as trading in eurodollar futures took off around 1985.
By 1990, interest rate swap markets, almost exclusively conducted over-the-counter (a private contract between two parties, not using a formal exchange or market), also began to gain traction. The two markets developed in close proximity, largely because they were complimentary. Eurodollar futures, by their very nature, attempted to predict future interest rates. That proved extremely useful in pricing and developing expectations for the floating leg of an interest rate swap.
As complex system capability evolved with shadow banking, the depth of these markets improved. Eventually, investment banks found them to be highly profitable and thus deployed resources to both service and develop them (read: sell).
For more than two centuries prior to 2003, JP Morgan was largely a classical or traditional bank. It had "earned" profits from lending or arranging financing, funding those credit production operations through various means that also evolved - from deposits to wholesale money. As the 21st century dawned, the bank, and its big Wall Street cousins, began to pull in a new direction. For the first time in 2003, JP Morgan derived more revenue from its derivative dealer book (mostly swaps) than traditional lending.
At the same moment, the FOMC was debating taking its federal funds target to 1% - a low never before seen, provoking all manner of inner monetary reflection. As Alan Greenspan himself worried, they had very little experience of policy interactions in financial climates at such extremes (it may seem quaint in a ZIRP-world, but there was serious debate and consternation, rightfully as it turns out). More than anything, however, Greenspan, and thus the FOMC, held fast to their orthodox world of ordered financial foundations. They believed that would be enough.
The starting point for that order began with the concept of "risk-free." The financial world is supposed to derive all its meanings, expectations and even values from the concept of interest rate components. In a Fisherian model, dating to Irving Fisher, financial markets are believed to create interest rate expectations from discrete modules - a risk-free rate plus spreads for inflation and credit risks. That would place the risk-free rate at the center of the financial universe (giving government borrowers exorbitant privilege).
There was a bit of a challenge to that order in the middle of 2004, as the FOMC began to reverse that historic policy "stimulation." Because swaps are extremely sensitive to changes in interest rate expectations there was great deal to be gained from predicting policy changes. If the Federal Reserve were expected to "tighten", for example, investors that were short swaps (receiving fixed) would stand to lose out (since you are paying floating, higher rates mean higher swap payments relative to a fixed receipt). As a matter of financial expedience, that would increase the need to hedge against possible cash losses in that eventuality.
Here, too, eurodollar futures were embedded as a companion. In the case where interest rates were expected to rise, those short swaps could hedge by selling eurodollar futures. Subject to concerns over convexity, where rising rates could cause hedges to accelerate beyond value losses on the original swap (and thus opens the door for dealers to advance from hedging into speculation, but that is another topic), the links between interest rate swaps and eurodollars should be durable. Swap shorts (floating rate payers) express expectations for rising interest rates by selling eurodollar futures as a hedge.
In early 2004, that is essentially what took place. The two-year swap spread, that is the two-year swap rate minus the two-year treasury yield, or the premium swap investors pay over the "risk-free" rate, entered that March at about 25 basis points. As economic data of the growing housing bubble began to allay fears after the dot-com bust and jobless recovery, swap market investors anticipated the eventual change in FOMC stance off historic lows. It did not matter that monetary policy, when actually needed, was/is conducted in the federal funds market, as history has shown there was a very tight relationship between federal funds rates and LIBOR (the eurodollar rate). That meant rates and dollar conditions across all wholesale markets were effectively controlled via the FOMC target.
The FOMC did raise the federal funds target beginning in June 2004. In the three months prior to that decision, the two-year swap spread rose from that relatively low 25 basis points all the way up to 43 basis points. And, as it turned out, there was a significant selloff in eurodollar futures largely as a result of swap shorts looking to hedge their rate exposure. The September 2006 eurodollar futures contract, the two-year equivalent at that time, dropped sharply from 96.45 on March 31, 2004, to 95.06 in mid-June.
Since eurodollar futures are contracts to deliver US dollars at future dates, that meant the cost of future dollars had increased. Another way of saying that is that dollar funding conditions had "tightened." That was exactly what the FOMC wanted to achieve at that time, but did not have to do anything to actually achieve its policy aims. By raising its interest rate target in June 2004, the Federal Reserve only promised it might conduct open market operations to change the federal funds availability should market conditions stray. The action in swaps and eurodollar futures projected those intentions, before the Fed actually committed to them, into real dollar conditions further out in time.
Again, going back to Greenspan's expectations, risk-free rates were supposed to set the stage to transmit policy into a broader spectrum of financial conditions and interest rates. However, almost immediately after the June 2004 FOMC decision, swap spreads began leaking back lower. In fact, the five and ten-year swap spreads began to fall backward even before the policy meeting. The two-year spread began to compress by July 2004.
In eurodollar futures markets, dollars conditions reversed toward "loosening." From June 14, 2004, until October 14, 2004, the September 2006 eurodollar futures contract retraced the entire selloff from April. Despite the beginning of a "tightening" cycle in Fed policy, markets were unsure as to its ultimate disposition across all markets. That meant for dollar funding markets as a whole, and not just ultra-short markets, there was more than a little divergence. In fact, with some minor volatility along the way, that September 2006 contract would not trade and stay significantly below 96 until September 2005; nearly a year and a quarter past the opening of the Fed's tightening cycle.
Of course, these sorts of market divergences only grew to extremes in 2007-08. All of a sudden, in early May 2007, the March 2010 eurodollar futures contract, for example, dropped precipitously. By the end of that June, it had fallen from above 95 to below 94.50. That was unrelated to policy stance, but it was a warning that dollar funding conditions were less than ideal at that moment. Only a month later, overnight funding would freeze in eurodollars.
Worse than that, however, that March 2010 contract had been rising, with some minor bumps, from October 2007 until March 17, 2008 - reflecting better reception in funding markets. Then, after Bear Stearns effectively failed, eurodollars tightened despite all intentions and policy efforts directed otherwise. From March 17, that contract fell from a price of 97.25 all the way to 95.25 by June 2008. While Bernanke and the FOMC were proclaiming victory and that the "worst was behind", eurodollar funding desperately proclaimed the opposite. That tightening in London dollar markets opened a spread between London and New York that had never appeared before (at least in any lasting fashion), belying the interpretations and expectations of policymakers and "experts." Dollar markets had become totally and durably fragmented.
By December 2008, even Greenspan's notion of comfort in orthodoxy translating into finance was shattered. For the first time in history (in the United States, they should have been paying attention to Japan where this happened before) there was a negative swap spread. It began in the 30-year tenor, but by March 2010 there were negative swap spreads down to the 7-year. The risk-free rate was no longer the elemental piece of the financial world after all.
What was most curious about this outbreak of negative swap spreads in 2010 was their relation to QE. Between March and November 2010, the 10-year swap spread continued to be low or negative, averaging just 0.01%. Over the same period, eurodollar futures prices rose, suggesting dollar loosening and conforming to swap-driven expectations. Then suddenly, just as QE2 came on line in early November 2010, they suddenly reversed. The June 2014 eurodollar futures contract traded at 95.20 in early April 2010, rising all the way to a peak on 97.88 on November 5 - the day after Chairman Bernanke's op-ed in the Washington Post welcoming the world to QE2. Spreads and futures prices then indicated tightening throughout the bulk of the QE2 operation.
In September 2012, just as QE3 was announced, swap spreads dropped significantly again. The 10-year spread was 7 basis points the day QE3 was declared, and over the next week it would drop to -3 basis points. Eurodollar futures would continue to rise in price all the way to May 2013.
But that was a strange occurrence. Between QE3's announcement and the end of January 2013, the 10-year swap spread averaged just 3 basis points; with many negative days in that time. Between February 14 and May 1, 2013, the spread averaged 11 basis points; there were no negative spreads. These might not seem like significant numbers or changes, but it was a full inflection in swap market conditions that was conspicuously unmatched by eurodollar futures.
So where swaps began anticipating higher rates, fully contradicting QE3&4, eurodollar futures did not until the full-blown taper talk in May. And when eurodollar futures tightened dramatically in the selloff of May/June, swap spreads remained largely unmoved.
The implications here are not minor trivia. When eurodollar funding markets, indicated by lower futures prices, "tightened" against the QE directive, as taper was priced by eurodollar rates but not swap spreads, markets all over the globe swooned (of course stock prices largely ignored all of this). The emerging market currency crisis that nearly led to panic in India and Brazil was forged by restrictive dollar funding conditions, as was the dramatic reversal in mortgage finance in the United States.
I can only speculate at this point why this divergence between swaps and eurodollar futures has appeared and persisted. Any interpretation of this kind, though, has to begin back with Greenspan's observation/warning. We do not know how markets will react or change in the face of extreme conditions for prolonged periods.
I have shared my contention before that swap dealers have been taking hedged positions to speculative proportions in the face of the latest QE's; meaning they were overly short swaps. As that crowded trade grew too crowded, it may well have set the market on a course where dealer capacity to absorb risk has been somewhat compromised. I only have partial and circumstantial indications of this (and it would require another lengthy discourse to describe it all), but I am increasingly convinced that there has been a major change in market behavior due to derivatives positioning and policy last year.
That would suggest, further, that the market relationship between policy intentions and market actions have become even more remote. It has always been tenuous to begin with, but if I am right about the analysis of current conditions, it would suggest that the FOMC is beginning to "lose control" of the curve and funding markets. Now, in reality they never had control, as demonstrated by the anecdotes above, but in this case it might be something new. It remains to be seen exactly how this all gets resolved, and the scenarios run through the intriguingly unquantifiable, so it all bears very close observation. We already know that QE is deleterious to collateralized lending systems, and this is a parallel function, so it may just be the inevitable end of the inevitable limitations of any system in extremis.