When Chairman Bernanke took the podium on Wednesday to deliver the FOMC's policy initiative, he probably didn't think he would blow up the most liquid and most important market in the world, but that's exactly what happened. I'm not talking about the Treasury market; I'm talking about the Eurodollar market. Eurodollars are USD 3-month LIBOR futures contracts and are used to price dollar-denominated credit across the global banking system.
On Wednesday, aggregate volume totaled 5 million contracts, equaling a notional value of $5 trillion. When prices settled, implied yields in the 2016 and 2017 strip rose by over 30bps. This was a massive move, with big money moving a giant market. The carnage continued on Thursday with volume trading 5.3 million contracts that saw another 13-15bps rise in the strip. On Friday, volume subsided a bit, trading only 3.5 million contracts with yields rising 10bps in the 2016 and 2017 strip. To close the week, the June 2017 Eurodollar contract's implied LIBOR rose by an astounding 88bps to settle at an implied yield of 2.975%.
The initial reaction to the magnitude of the move in the front of the curve was that the market had misinterpreted Bernanke's forecast for tightening. In an article floated late Friday, the Wall Street Journal's Fed reporter Jon Hilsenrath wrote in Fed Toils In Vain to Calm Markets:
Many investors appear to have missed Mr. Bernanke's signals that the Fed might wait longer than expected before raising short-term rates. He said on Wednesday that the 6.5% unemployment rate threshold might be too high and that the Fed might decide to keep rates low for long after the rate drops below that level, especially if inflation remains low.
According to projections released after the meeting, only four Fed officials saw short-term interest rates rising before 2015, while 15 saw rates remaining near zero until 2015 or 2016.
In theory, that should reassure investors that borrowing costs are going to stay relatively low for years.
Then why are Eurodollar futures markets pricing in a 3.0% LIBOR by June 2017?
Last week in The QE Carry Trade is Imploding Right Under Everyone's Noses, my theory was predicated on two assumptions: one, that the tapering was a renege of the Fed's open-ended inflation target after it got the market levered long the trade and two, that it was the bond market's discount for the inflation premium embedded in real interest rates and not the flow of purchases that was responsible for market pricing. Both of these assumptions came to a head on April 10 when the Fed released the minutes from the March FOMC meeting's QE III cost/benefit analysis signaling a reduction in stimulus was forthcoming.
As I said last week:
Cue the April spike in volatility of volatility. In analyzing this development in the context of realized volatility of the equity market for which the VIX is priced vs. the realized volatility of the market that finances carry trade positions, the move looks to be a function of financing costs.
Presumably a market that is short gamma and long inflation premiums was forced to lift volatility because inflation premiums collapsed, raising the cost of financing the trade. I believe this is a shot across the bow that leveraged positions are under pressure....
When Bernanke gave his Joint Economic Committee testimony on May 22, his prepared remarks tried to subdue tapering expectations, but in the Q&A, he admitted that it was forthcoming. The equity market saw a blow-off top and reversed hard to close down on the day. Thus far that is the market top, and each attempt at a backtest has been turned away. This week may be the most important FOMC meeting of Bernanke's tenure. He will no doubt try to calm markets that are in the process of re-pricing the inflation premium. It will be very interesting to see how real interest rates perform based on what he says. If real rates are rising and the equity market fails yet again, it could be the final nail in the QE carry trade coffin.
Brazil 10YR v 10YR Real and 2017 Eurodollars
Positions that are long carry trades financed in USD predicated on the Fed's negative interest rate regime are in effect long Eurodollars (short the floating rate) by virtue of their financing cost. When Eurodollar prices decline, implied LIBOR rates rise and vice versa. This carry trade relationship can be seen in the correlation between real rates, Eurodollars, and a popular carry trade investment such as Brazilian bonds. The correlation is the same with Chinese lending rates.
This week, Izabella Kaminska of the Financial Times' Alphaville blog posted a report titled "An unwind in the great Chinese over-invoicing carry-trade?"
Deutsche Bank's Bilal Hafeez made a strong case for this interpretation last week. We think it's worth revisiting the argument.
There are three main factors that need to be considered, he argued:
1. Adjusted for volatility China now offers the highest FX carry in the world. This, he says, has led to a surge in flows into the CNY (and CNH) by both onshore and offshore entities over the last few years.
2. The performance of the carry was reaching breaking point last week.
3. Any combination of higher US yields, regulatory clampdown or higher FX volatility or CNY appreciation could trigger an unwind.
Due to the CNY/USD peg, it would make sense that carry trades would show up in China, however, this is a pretty wild use of a cross-currency financing mechanism. I don't want to dwell on the specifics of the trade but rather the trigger for the execution.
Kaminska goes on to say, "Hafeez estimates the carry turned positive about the end of December 2011."
Of course it did. The carry trade turned positive in late 2011 when real interest rates turned negative. This is a huge point to understand what is really going on in the bond market.
Real v Nominal 10YR Yield
Back in April in Gold, Interest Rates, and the Great USD Short Unwind, I pointed to a development in USD FX reserve accumulation that in 2007 was the catalyst for famed macro hedge fund Corriente Capital's thesis on shorting peripheral European sovereign debt (before spreads blew up). I thought the 2011 decline in USD reserves held at the Fed was showing a similar carry trade trend that I believed was now present in emerging markets.
USD FX Reserves Held at the Fed
As I said in April:
[Y]ou will notice that since the 2009 peak, the share reserves getting deposited back at the Fed has been in decline with a steep drop beginning in August 2011. This is a major change in trend from what has been in place over the past decade, which is indicative of a capital flow reallocation.
Just as Corriente suspected, the 2007 dip reserves held at the Fed was a carry trade into European sovereigns I believe this larger reduction in custody holdings is indicative of a similar short USD carry trade into emerging markets.
The market's very levered long carry trade was not only predicated on the Fed's negative interest rate regime but also on the specified time period the Fed has committed to keeping interest rates negative.
Consider that at the August 9 2011 FOMC meeting the Fed replaced exceptionally low levels for the federal funds rate for an extended period through mid-2013. A few months later at the January 25 meeting, it extended the rate guidance to at least through late 2014 only to extend it once again in September to at least through mid-2015. This type of calendar guidance for the length of a zero percent interest rate is a dream for a leveraged carry trade. You know your exact cost of funding for a specified time period, and due to this policy, the price of the funding currency will remain depressed. What can go wrong?
However, QE III changed the rules by opting for economic thresholds in lieu of calendar guidance. This introduced a new and uncertain risk factor for carry trades.
In December, only a few months after the Fed extended the guidance to mid-2015, it removed the date and replaced it with a specific 6.5% level in the unemployment rate. This is a very important distinction from the carry trade perspective because now you no longer have a predictable expiration on your position but instead are at the mercy of the unpredictable economic data. At the same time, you have Fed rhetoric discussing tapering purchases in order to reduce the growth of the balance, which is also net dollar positive. What was once a no-brainer dream trade has now become a very complex and volatile nightmare.
This QE carry trade nightmare became reality last week, and the Eurodollar pit was ground zero. As carry trade asset prices come under pressure due to rising US real interest rates, investors are forced to sell Eurodollars to hedge higher financing costs and negative gamma exposure. The magnitude of the selling implies that there is a lot of money exposed, but it's not clear what still needs to unwind.
Last week, there were rumors of bond dealers who were both liquidating MBS inventory and ceasing to bid on these securities until quarter end. There were also accounts of liquidity drying up in the Treasury market. When dealers cease to bid on the assets that collateralize the loans for carry trades, the system is frozen. This is serious.
If you believe the accounts in the media, you would think the Fed believes the move in the front end of the yield curve, including the Eurodollar strip, is a misinterpretation of Fed tightening. The Eurodollar market not only has an interest rate component but also a credit component, and one interpretation of the blow out in the strip is a spike in banking system credit risk.
2017 Eurodollar Spreads v Fed Funds
The steepening in the Eurodollar curve in and of itself is not unprecedented, but in the context of QE and zero interest rates, we need to consider what it may be indicating. In past cycles when the Eurodollar curve steepened, it was a sign either that the Fed needed to ease or that it was about to ease. Curve steepening occurred when rates were falling and was indicative of tight money. This is a very different interpretation of what is happening in the front end of the yield curve -- and if media accounts of the market's reaction are a reflection of the Fed's interpretation, then we should assume it is not aware of these potentially severe tight money conditions.
Is the bond market discounting tapering, or is it a product of the carry trade liquidation? In 2007, when the 10YR yield spiked through 5%, it was initially perceived to be a positive sign that the market was discounting accelerating growth. It turned out to be HF liquidation to meet collateral requirements. It's very possible that we are now seeing a similar liquidation. When you have to sell, you sell what you can first. The fact that this liquid collateral is already tight is a very troubling sign. That great big sucking sound is liquid markets becoming illiquid.
When compared to the US rates' complex and emerging credit markets thus far, the US equity market has remained relatively unscathed. Equity market participants are still very much in the green on the year and appear to be confident that last week's disruption will prove to be a buying opportunity in an ongoing bull market. I believe that the entire global economy and capital market is predicted on the US negative interest rate regime that is no longer negative.
Make no mistake about it: Bernanke is blowing up the QE trade he engineered. The question for markets at this juncture is not what assets are exposed to this trade but rather how much capital is exposed and who will take the other side of the unwind. The move in the most liquid part of the rates curve suggests that the position is very deep; the reluctance of dealers to bid on financing collateral suggests the bid is very shallow. Finding a level where that bid/ask comes together is likely to be a very disruptive process, and if history is any guide, the "collateral" damage will be felt around the world.