An important positive contributor to the fund’s performance over the past month has been our paid fixed positions in interest rate swaps against three-month Libor, a trade designed to benefit from rising short-dated US dollar yields.
Short-end yields have indeed risen: since 27 June, one-year overnight index swaps (OIS), which most closely track expectations of the US Federal Reserve’s (Fed) policy rates over the next 12 months, have climbed from 0.35% to about 0.49%, as at 10 August. Over the same period, one-year interest rate swap rates, against three-month Libor, surged from 0.65% to 0.90%.
The move has been partly the result of investors seeing an increased probability of Fed rate hikes over the next few years, as can be seen in the 14 basis points of higher policy rates priced in to the one-year OIS rate.
But this only explains about half of the move in one-year swaps against three-month Libor, the other half of which has been the result of a widening in the Libor-OIS basis, or spread between the two rates. This has taken place as market rates on swaps that use three-month Libor as their reference index have risen by more than those in products that more closely track pure policy rate expectations, such as Fed Funds futures, OIS or even short-dated US Treasuries.
FEW SIGNS OF FUNDING STRESS
Why is this happening? Traditionally – or at least since the summer of 2007 – widening in this basis has been symptomatic of a deterioration in funding conditions, as banks are forced to pay up for unsecured term borrowing in the interbank markets. (Or the panel bank submissions, which determine Libor, anticipate that this would be the case, should banks actually attempt to fund in this manner.)
This is not really the case right now. Other metrics that would typically suggest ‘funding stress,’ including cross-currency basis swaps, financial credit default swaps and other G-4 Libor/OIS bases, are not really showing signs of such difficulties.
The explanation lies in imminent reforms, due to kick in on 14 October, to US ‘prime’ money market funds (MMFs), or those that invest primarily in corporate debt securities. Under this regulation (so-called rule 2a-7), prime funds in the US will have to switch to floating net asset values; increase their percentage of very short-dated holdings; and, in some cases, impose redemptions gates and liquidity levies.
As a result of these changes, prime MMFs are expected to see widespread redemptions as investors switch to other products; indeed, total prime MMF assets under management (AUM) has already fallen by 28%, or US$358bn, in the year to date.
In anticipation of a further decrease in AUM – and a general desire to be more nimble amid uncertainty over the impact of the reforms – prime MMFs have begun to cut the weighted average maturity of their holdings significantly. In many cases they have replaced maturing three-month paper with much shorter-dated securities (one-month and less).
THE ROLE OF ROLL-OVER
The important point with regard to three-month Libor is that these prime funds are typically some of the biggest buyers of bank commercial paper (CP); and in turn, the rate at which banks can fund themselves in CP markets is one of the key determinants for where Libor is set (see chart 1 below, showing the relationship between average three-month bank CP yields and three-month Libor).
It is no coincidence that the increase in term funding yields has coincided with the period roughly three months prior to when the MMF reforms take effect. As MMFs have trimmed their plans to roll over longer-dated CP maturities – instead of replacing a maturing three-month security with a new one, they would rather roll their holdings into maturities prior to 14 October – the rates at which banks have to pay to borrow in longer-dated tenors, like three month, have increased materially.
When will it end? Over the medium term, three key questions need to be answered to determine the ultimate clearing level of three-month CP issuance, and hence three-month Libor:
1) To what extent will banks find alternative sources of US dollar funding?
2) What will the actual ultimate reduction in prime MMF AUM be?
3) Where will the cash that has been withdrawn from prime MMFs end up (or relatedly, will other investors step in to lend to banks in CP markets)?
Over the longer term, we are optimistic that in some way or another, banks will stop having to ‘pay up’ for funding. In 6-12 months’ time, the situation is likely to have normalised, with the spreads described earlier between term funding rates and base rates back to tight and predicable levels.
JAPANESE AND FRENCH BANK INTENTIONS
In the short term, however, it is very difficult to predict how disruptive the reform process will be, and to what extent banks are willing to pay higher yields to access funding in the size and maturities to which they have become accustomed.
Given the desired levels of US dollar funding needs, Japanese and French banks represent a disproportionate amount of CP issuance, and in the last month Japanese banks in particular have paid consistently above-average yields to roll over maturing paper. As shown in chart 2, French and Japanese banks have a heavy schedule of US dollar CP maturities between now and mid-September, and the question of where new borrowing is deployed will be a key determinant of the short term evolution of three-month Libor.
For now, the move higher in average bank CP yields shows no signs of reversing. Indeed, on 9 August three-month average bank CP levels maintained a three basis point spread above three-month Libor, when historically they have tended to trade slightly below this rate (see chart 1). This is likely to continue pulling the three-month Libor fixing higher in the short term. The three-month Libor rate on 9 August came in at 0.816%, which puts the ‘spot’ spread between three-month Libor and OIS at 40 basis points – compared to an average of 16 basis points over the past three years.
Based on current market pricing, this spread is expected to increase to about 44 basis points by September, before declining somewhat. Any further upside surprises to the three-month Libor fixing will see market expectations revised higher, which will probably ripple out as higher front-end Libor-based swap yields.
For now, we are holding onto our paid positions in such swaps, but are actively assessing developments in the CP markets as a signal of when to switch this stance into other products.
Chart 1: Fixed CP/CF 3m transaction rates, simple daily averages (%)
Chart 2: Maturity Schedule of Japanese and French bank CP/CD ($bn)