UK and US long bonds are in opposite regimes, for now
The things we tend to trade a lot don’t have a simple economic meaning. Take the current 10y US treasury. From the point of view of understanding the cashflows, its simple enough. If you spend $107.63 to buy it today (11may2020), you’ll receive $100 on the 15th of Feb 2030, and $0.75 every 15th of August and February until that date. That’ll work out to a return of 0.69% if you hold it to maturity.
Understanding why someone might do that is far harder. The economic meaning of the decision to buy a 10y bond is complicated. We might say “investors buy bonds when they expect interest rates to fall” but which interest rate? If the Fed cuts the overnight rate, there’s no guarantee the price of the bond will change. If Libor falls because of resolution of a panic in the financial system, the price of the bond will likely fall not rise as investors no longer prefer the safety of a government guarantee to riskier investments elsewhere.
In any case, an investor might buy the bond without taking interest rate risk by using an interest rate swap to hedge it. She might buy the bond and simultaneously contract to pay a fixed rate and receive the fed funds rate for 10 years. Were she to do that today with the current 10y bond at 0.69% and the 10y OIS at 0.39%, she’d lock in a return of 0.3% with no interest rate risk. Of course, 0.3% is not very much - but it beats lending the cash overnight in the Fed funds market at 0.05%, or buying a 3m Tbill and earning 0.1%. Her decision to invest is based in the opportunity cost of cash. Even better, she doesn’t need to find cash to buy the bonds. She can buy them and simultaneously borrow the money to do so using them as collateral in the repo market. The SOFR rate being at 0.06% indicates that’s around what it will cost her to do so. Whether it’s worth doing for her will then depend on her view on the refinancing risk, and the cost of using her balance sheet.
All this is to say that the decision to buy or sell bonds is only distantly related to the economics. Regimes can and do shift based on who the marginal buyer and seller are, and what is motivating them. The same shock or event can affect the same product differently based on when and where it happens.
The UK and US government bond markets over the crisis period thus far offer a classic case. Here, the white line is the spread between the yield on 30y US treasuries vs 10y - and the yellow is the same for the UK (sourced from Bloomberg as usual):
In the UK, the price of long term vs short term debt has an intensely ambivalent relationship to interest rates because of the way that pension funds are regulated. Pension funds must match the duration of their assets (stocks, bonds etc) and liabilities (payments to pensioners). When interest rates fall, the duration of their liabilities increases and they must substitute short term bonds for long term bonds in their portfolios. Unfortunately, regulators fail to recognise that equities are a long duration asset and so keep forcing pension funds to sell equity at the lows and buy long duration bonds at very low yields. Slower and more anticipated changes in rates don’t have this effect, meaning it’s tough to discern the correlation. The US does not have this problem - and so when interest rates fall unexpectedly, so long as the change is expected to be temporary - short term bonds move more.
Half of the 20bps that US 10s30s has steepened is thanks to the spread between short term and long term rates. The other 10bp is from the spread to OIS - thanks to the widely followed announcement by the US treasury that it’ll be increasing the supply of long bonds. Recall our investor who’s interested only in the opportunity cost of cash. That opportunity cost will be different for different investors. Supply of a particular tenor will fill those with a low opportunity cost, and the next investor won’t buy until the bond offers a better spread. We’ve got a good sense of where the support is for 30y US treasuries thanks to the dramatic selloff in March. Here, the 30y UST yield spread over the 30y OIS:
At just shy of 100bps spread, we run into buyers who appear to have deep enough pockets to keep the spread from moving wider despite highly adverse market conditions. We met them before in 2016.
So, we’re in a regime of find the backstop buyer in long end USTs, vs a scramble for duration in UK Gilts. But change is coming. In UST’s, for the first time in years, it is better than buying domestic bonds for Japanese and European investors to buy UST’s and hedge the FX risk, as pointed out by Stephen Spratt at Bloomberg:
For Gilts, the shock of lower rates is temporary. Having increased the duration of their assets by buying long end gilts, and receiving swaps (pushing the 10y vs 30y swap rate difference into negative territory) the pension funds need do nothing more.
As for supply, the US is ahead of the UK in announcing long term bond issuance but need not stay that way. So far, the fiscal response has been anemic but this doesn’t mean issuance will be low even if that doesn’t change. In a depression, you don’t choose your fiscal deficit - it chooses you as tax revenues fall, and automatic stabilisers kick in. Then in the US, long bonds are competing with mortgage bonds for bids - and the supply here will fall as mortgage origination slows.
None of this is a prediction as to what will happen, rather the aim is to show the possibilities. The current regimes in the US and UK are the result not just of different fundamentals, but different categories of investors facing different constraints and incentives. For the purposes of this note, all I want to do is throw light on how these kind of divergences open up in markets and how they might close.
NB: This post is not investment advice and is not a trade recommendation. The views expressed here are my own and do not reflect those of my employer.