Spread 'em

If something seems to good to be true, it usually is.   So it was with the set up in Treasury bonds yesterday, as they delivered a tasty squeeze- but only after early Gilt market weakness sucked in more sellers of UK bonds before reversing and reminding punters that selling intraday weakness in fixed income has been a sucker's game for a long, long time.

Today of course sees the release of UK CPI for September, a month which saw sterling start near its post-Brexit vote highs and end near its (pre flash crash) lows.  It seems virtually axiomatic that inflation is going higher in Britain, beyond the symbolism of he lowly tub of Marmite.  The forecast is for a mere 0.1% monthly rise, which seems modest; that being said, if it comes in as expected a further bout of short covering in Gilts may well ensue.  That will be the price action you want to sell into.

Back in the US, per yesterday's post curve steepeners are all the rage.   It's not hard to see why; beyond the steepening rhetoric from Rosengren and the softly-softly comments from Yellen and Fischer, the chart has not only shown signs of life in the last couple of months...

...but also remains at very attractive levels from a more strategic perspective.

Then again, if we like steepeners in bonds at 127 bps, we must love them in swaps at a mere 69 bps!  Moreover, unlike the Treasury curve, the 5/30 curve in swaps is still flatter than its lowest point in 2015.   What's going on here?

The answer, of course, is swap spreads, most notably at the back end of the curve.   30 year swap spreads are plumbing uncharted depths, reaching -56 bps yesterday, an all time low.

Here's the chart of the last year...

...and here's what it looks like from a longer term perspective.

It is a measure of the perversity inflicted upon financial market pricing by regulatory and monetary policies that there are a number of valid reasons why spreads are negative:

* Many holders of Treasuries have historically lent them out in the repo market as collateral for short term monetary loans- which, in the case of leveraged buyers of Treasuries, are used to pay for the bonds in the first place.  However, the repo business is balance sheet intensive for banks and generates a fairly paltry return in compensation.   In the new regime, banks are simply not interested in using their balance sheet to help clients buy bonds, which has decreased the potential buying power of Treasuries.

* Historically, the spread between LIBOR and the GC repo rate has been a good proxy for swap spreads.   Although it's declined a bit recently, as you can see it's nowhere near the post-crisis lows like swap spreads are.   In this case, regulation is also having an impact, via the much-discussed change to money fund regulation (which went into effect last week) that has pushed LIBOR higher.  Last year, however, there was briefly a time when LIBOR (uncollateralized lending between banks) was lower than GC repo (lending with Treasury collateral.)  Taken at face value, this implies that a bank could get a lower rate by calling a competitor and asking for a 3 month loan with no strings attached than if he offered to post Treasury bonds as collateral.  To paraphrase Inigo Montoya, I do not think these rates mean what you think they mean.

* Reserve managers are selling Treasuries.  This remains an ongoing consequence of dollar strength, and while it may explain some of the recent decline in swap spreads it certainly doesn't explain the level; 30 year spreads were negative even when CBs remained enthusiastic buyers of Treasury bonds.

*  Swaps are less risky than was historically the case.   Until a couple of years ago, swaps were a bilateral agreement between two consenting parties, which created counterparty risk.  This was obviously a major feature of the spread widening in 1998 (LTCM) and 2008 (Lehman) when notable counterparties in the swap market went bust.  With the move to central clearing and an exchange as a counterparty, this bilateral credit risk has been largely mitigated for vanilla swaps, so it does kind of make sense that they should be seen as less risky relative to Treasuries than was historically the case.   It still doesn't mean they should be less risky in an absolute sense of course.

* Corporate issuance.   It's been well documented that corporate issuance volumes in the stratosphere.  When a corporate issues a bond, they undertake to pay a fixed coupon.  With an upward sloping yield curve, it makes sense to swap this into floating rate obligations to reduce interest cost.   Thus, when corporates issue debt (which they have done in record volumes), they hedge by receiving swaps.  On top of the other factors mentioned above, this has created something of a one-way street in terms of the demand to receive swaps relative to owning Treasuries...hence the decline in swap spreads.

While some of these factors are structural, others are more cyclical, and thus more apt to reverse in a timely fashion.  Who knows what the Platonic ideal of the "correct" value for long dated swap spreads might be, but at all time lows it seems a reasonable proposition that it is somewhere north of here.