When the European Central Bank signalled that its next likely rate move would be upwards, it triggered a sharp shift in interest rate expectations. The euro swaps yield curve dramatically inverted between the two-year and 10-year maturities shortly thereafter the first time for benchmark European yields since the early 1990s.
"A lot of this interest rate move is a fundamental reversal, an unwinding of the markets mispricing of where policy rates would go," says Mark Schofield, head of interest rate strategy at Citi.
Schofield argues that this shift started in March after the bailout of Bear Stearns and points to the fact that the US Fed funds futures strip has swung 200 basis points higher a dramatic move for the front end of the curve.
With inflation clearly on all monetary policy makers agenda, sterling and US swaps yield curves have also inverted/flattened between the two-year and 10-year maturities.
|
2-10 yield curves invert/flatten |
|
Historically, Inversion in 2s10s curve has lasted a long period of time |
 |
|
Source: Credit Suisse |
It is a well-understood canon of financial markets that banks dislike inverted yield curves. This is largely explained by perhaps the most fundamental aspect of their business model: borrowing short and lending long. This so-called carry trade works fine as long as yield curves are upward sloping a normal occurrence in finance. But normalcy in the rates business was thin on the ground shortly after June 9 following ECB president Jean-Claude Trichets hawkish sounds which is why the swap curve inverted shortly thereafter to minus 75bp between the two-year and 10-year maturities.
The inversion happened dramatically, at a speed that cannot be explained by the shift in interest rate expectations. The key factor was dealer hedging of positions created by structured note transactions.
"The last part of the shift, which has affected the two to 10s swaps curve is related to the hedging of range accruals," says Schofield.
"This was caused by the risk management required from the selling of digital floors on a variety of euro CMS spread pairs; one significant example is the CMS spread range accrual. The way these transactions are structured means that the hedge providers risk changes from a long to a short gamma position as the underlying moves through the strike of the digital option, eg, the implicit coupon floor of zero if the lower spread barrier on a range accrual is struck at the spread being equal to zero," says Chris Jones, global head of EMTNs and structured notes at HSBC.
Gamma measures the sensitivity of an options delta, while delta is the sensitivity of that option to changes in the price of the underlying asset.
"When the curve goes through the floor they have to put on extra curve flatteners to hedge their risk," says Jones.
The key structures are non-inversion notes, which as the name suggests are sold to investors on the premise that the yield curve will not invert but will slope upwards. The payout structure of these notes is that investors are paid an enhanced coupon over the floor which is normally set at zero but sometimes at minus 20 or 30bp. They are also known as range accruals; they are frequently created using constant maturity swaps.
|
CMS-linked range accrual activity |
|
Billions of structured notes need hedging |
 |
|
Source: MTNi |
For such notes, there is an embedded option for which the dealer is both long a digital floor and short a linear floor. When the CMS curve goes past inversion the digital option is no longer positive but becomes negative. This results in the dealer being very short gamma. Suddenly the dealers hedging changes and could require paying two-year and receiving 10-year swaps (curve flattening), instead of trying to steepen the curve by receiving twos and paying 10s, when the curve was not inverted. The exotic options desks at banks dynamically hedge these notes. But once the floor is breached the market dislocates and instead of stabilizing the yield curves moves, risk management of these structures exacerbates it.
The two-year swap was at 3.55% and in a matter of weeks has jumped by 200bp. And at the most severe time of the dislocation, normalized volatility in the curve doubled in two days.
According to HSBC, this delta and vega hedging via dealers putting on curve flattener trades contributed to the inversion, which then required more hedging and further inversion.
For the dealers that structured these range accrual notes, the faster they hedge the better it is. But because everyone is heading for the same exit, the price moves against dealers. By all accounts, several dealers have lost sizeable sums on the dislocation.