Choudhry on Long-Term Swap and Bond Yields Puzzle
This column has pleaded for cold, unemotional logic when analyzing the financial markets, but it appears this is actually quite rare.
This week, we hope to ensure a dispassionate debate in our consideration of the yield curve, and ask whether the return from a government bond should always sit below that of a bank asset.
The almost universal answer to this question is yes of course, and in a perfect economic environment of complete information and no transaction costs this would be the right answer.
After all, a government bond is, if not risk-free, at least the lowest credit risk of all borrowers in the domestic market, and so it should pay the lowest interest rate.
Readers may have guessed by now however that this is not always the case.
Currently we observe an interesting conundrum in that for US dollars and pound sterling, the long-dated swap rate actually sits below the government bond yield.
The spread is even more pronounced when the comparison is made with “overnight-index swaps” (OIS), which are derivative contracts that exchange a fixed interest rate against the average of the overnight interest rate for the same period.
These lie even further below the government bond yield.
On a purely academic basis we would want to know why this is the case.
However, from a policy angle we would also want to know as central banks use these curves in their macroeconomic analysis.
For example, they infer what the expected future base rate is from the forward swap or OIS curve.
If these are trading at spurious levels, it could result in misinformed policy making.
Under what circumstances should the swap curve lie below the government curve? (Theoretically, the scenario implies that banks are a better credit risk than the UK or US government over a 30-year horizon. Does any market participant actually believe that?).
Essentially it’s a combination of thin markets, supply and demand and the dealing preferences of swap market makers.
Much of the demand for long-dated swaps is from bank customers such as corporates and governments wishing to switch fixed-rate liabilities into floating.
Banks have to hedge this with the market, with the result that banks are paying a lower fixed rate when quoting the swap price.
At the same time, this is not that liquid a market – there is only a select group of banks in the world that will quote two-way 30-year swap prices, and so demand will have a significant downward impact on the swap rate when the banks are facing a big requirement for pay-fixed/receive-floating swaps.
The combination of customer demand and thin markets leads to the anomaly we observe above.
The main conclusion we draw is that the prices are not quite “correct”.
By that we don’t mean we have departed from our beloved Chicago-school view of the world, rather that the prices do not reflect the real-world credit risk of banks vis-à-vis the UK and US government.
At the same time, the above prices do present some interesting arbitrage opportunities for those whose blood is rich enough.
This column has been written ahead of the 2 August deadline for the US government and Congress to agree on raising the public debt ceiling.
If that hasn’t been achieved, we’ll all be faced with more serious issues than an inverted swap curve…and un-emotional analysis will again prove difficult to achieve!
The author is Dr Moorad Choudhry, Head of Business Treasury, Global Banking & Markets, Royal Bank of Scotland, and Visiting Professor at London Metropolitan University.