Big Problem: The Fed's Guidance Is Short Volatility, Based on Data That's Volatile Itself
Each data release will be an excuse for the market to recalibrate forward guidance of highly leveraged positions, meaning the slightest change in economic data could elicit a massive reaction.
This past week, Fed Chair nominee Janet Yellen submitted answers to questions from two Senators clarifying her views on the current state of monetary policy, and seemed to cement her objective for the new regime. The message was loud and clear: Rates will stay lower for longer. I believe last week was the passing of the baton. For all intents and purposes, this is now Yellen's Fed. Of course the markets have been hearing this accommodative rhetoric ever since the Fed decided to forego tapering at its September meeting, opting for continued stimulus in the face of record equity market gains.
The bond market's reaction to Ms. Yellen's comments has been what you might have expected. The curve traded with a steepening bias as the front rallied, anchored by the "lower for longer" rate policy, while the back sold off, discounting the tapering of QE and eventual exit.
The Fed is going to try to maintain extreme accommodation while backing out of its asset-purchase program. It remains to be seen if this is possible. Accommodation is about expectations, and participants' belief in future inflation. Forward guidance (FG) is really an attempt to tell the market that policy will be accommodative as long as inflation remains below target, therefore future inflation expectations should remain elevated; this keeps real interest rates low, stimulating the economy. Make no mistake about it: Monetary policy is targeting real interest rates by keeping nominal rates low and inflation expectations high.
Last week's CPI release showed a year-over-year growth rate of 1.0% which, with the exception of the 2008-2009 deflation scare, was the lowest rate of headline inflation over the past 50 years. This is significant because longer-term Treasury yields have been rising due to taper expectations, thus the 1% CPI put the actual real 10-year yield near the highs of the cycle at 1.78%, up from negative 2.0% in 2011.
This pricing of actual real yields at 1.78% is over 125bps relative to where the TIPS market is pricing real yields. TIPS are indexed to the CPI, so this 1% pace is what investors will be paid over the 52bps implied real yield despite the implied inflation break-even rate of 2.20%. In other words TIPS investors are paying a huge premium for future inflation risk relative to where they are being compensated for actual inflation.
Real 10-Year TIPS Breakeven Vs. CPI
The Fed thinks it will remain accommodative absent QE as long as it pledges to keep Fed funds (FF) at zero for the foreseeable future. But at the margin, stimulus is being withdrawn, and investors' inflation expectations are priced at the margin. As the Fed pulls back, I would expect inflation expectations to fall and the spread between TIPS yields and actual real yields to narrow.
If real yields rise because inflation premiums fall, this could anchor the long end of the curve at a time when the market believes it is going to rise due to tapering. The real interest rate at 2.0% is not expensive relative to the last 20 years and is cheaper than during the credit-boom years when nominal yields were closer to 5.0%.
Contrary to consensus, the long end of the curve looks relatively cheap. At 3.0% the 10-year is in line with the growth rate of nominal GDP, personal consumption, and S&P 500 (INDEXSP:.INX) revenues. It's higher than 2.2% wage growth and well above 1.0% inflation. The front of the curve is another story. The front of the curve is anticipating zero percent Fed funds for the foreseeable future and is priced accordingly, regardless of fundamentals.
The Fed is moving toward forward guidance of the Fed funds rate as its primary policy tool, as "lower for longer" replaces asset purchases. Since September the market has been positioning for this reality in a big way. The optimal way to make money in a negative interest/zero interest rate environment has been to leverage spread carry trades. For QE, this meant leveraging up duration and risk premiums, which took place in mortgages and credit, such as high-yield and EM bonds. For FG this means leveraging up the front of Treasury and eurodollar curves for 90-day LIBOR futures.
The eurodollar market, a.k.a. the money curve, is the largest and most important market in the world and ground zero for forward guidance positioning. The so-called converge trade, whereby investors get long the eurodollar futures strip in order to roll down the curve towards spot LIBOR, was on big earlier this year and subsequently blown to pieces when taper talk brought on rising real rates and carry-trade unwinds.
December 2015 Eurodollars (EDZ5) Vs. Open Interest
The most recent example of this convergence trade has been in the massive buying of December 2015 eurodollars (EDZ5). The position has gotten so large that the buyer has been given the nickname of Z5 Capital. His position can be seen by the exploding open interest that, at the end of August, was 618K contracts but today has almost doubled at 1.163 million. Keep in mind that the notional size of a eurodollar contract is $1 million, so this open interest represents a notional value of over $1 trillion. In aggregate, the open interest in the contracts in 2015 represents positions of $3.2 trillion, which is comparable to the total size of the Fed's balance sheet.
On Friday I was paying special attention to what a savvy eurodollar trader in Chicago, who goes by the name Shinebox (must follow), was reporting on Twitter (NYSE:TWTR) about the positioning in the front of the eurodollar curve. He noted that there was big size trading in December 2014, and that it probably had to do with "Z5 Capital" rolling up down the curve to cover some calls he was short. He noted that there had been big RV (relative value) players in the curve over the past few days. I responded that they were "comfortably" long the convergence trade and asked him if this positioning could be setting up long gamma trades around big economic data releases. He seemed to agree.
Gamma is the second derivative of an option's price. It is the rate of change in implied volatility premiums for a given change in delta. Gamma is a unique characteristic of market positioning because of its exponential impact on prices. When you are long gamma, your position gets longer the more price goes up and shorter the more price goes down. Conversely, when you are short gamma, your position gets shorter the more price goes up and longer the more price goes down. Short gamma "blowouts" are typically responsible for outsized market moves and often play a role in crashes.
The convergence trade I spoke about can be seen in in the collapse of various eurodollar curve butterfly spreads. A butterfly trade is a spread trade where you take a position in the belly of the curve and spread this position with offsetting positions in the "wings" weighted 1x2x1. If you buy the butterfly, you own the belly and sell the wings. An example is to buy the 5-year note and sell the 2-year and 10-year (short one 2-year, long two 5-years, short one 10-year). In the eurodollar curve this spread collapse can be seen in the December 2014/2015/2016 and December 2015/2016/2017 butterflies.
Eurodollar Butterfly Spreads
The move in butterfly spreads is notable. Intrigued by the significance of this move, and more importantly, the positioning, I asked another eurodollar trader privately about his opinion on the magnitude of the move.
I asked the following: Would you consider the collapse in flys in Dec 14/15/16 and 15/16/17 to be a big move? And more importantly, does it present systemic risk on hiccup?
He responded with the following: Extraordinary move, and they are implicitly a gamma sale as they are taking term premia (implied carry/roll) out of the market. At some point the 2015s will be tested
Gamma sale? 2015s tested? This was in-line with my idea about long gamma setups in front of big economic data releases.
I gave him my thesis: My concern is that the data is itself so volatile. How can the Fed keep it together when there is no trend? We could get an NFP print of 250K then the following month get 80K; that's the trade. The market is short FG volatility predicated on economic data that is long volatility.
He responded, "Good point."
The Fed's forward guidance has once again levered up the ZIRP convergence trade in eurodollars. The same one it blew up a few months ago with taper, only this time around it appears the trade is much bigger, and judging by butterfly spreads, it is a much tighter curve. This is where the systemic risk is buried. Everyone is looking for long bonds or stocks to crash, but they need to be looking at the front end of the yield curve. The bond market is short forward guidance volatility predicated on economic data that is extremely volatile.
Forward guidance sounds simple enough. Tell the market where you want it to go and for how long. In practice, though, it's not quite that simple. Fed communication has been erratic and contradictable. Positioning for forward guidance is simple. Get long the convergence trade and ride the curve to zero. Maintaining your position is not quite that simple. Each data release will be an excuse for the market to recalibrate forward guidance of highly leveraged positions. The market is not going to wait around for the Fed to tell it what to do. Positions are short volatility, but more importantly, short gamma. The slightest change in economic data could elicit a massive market reaction.
There is a policy shift underway. The Fed is prepping markets for a transition from QE to FG. The no-brainer trade going into this policy shift is to be long the front end and short the long end. However I think you should be looking to fade this sentiment. The accident risk is in the front, and this risk is going to be put to the test with each and every important data release. That is risk I think you want to sell.