Are LIBOR swaps still representing Interbank credit risk?
The letters "IB" in LIBOR stand for InterBank. The LIBOR fixings are suppose to represent the rate in the interbank market, with its pure interest rate component and its credit risk component. The spread between the LIBOR of a given tenor and the OIS rate for the same tenor can be viewed as a proxy for the credit quality of banks.
In the August to October period, the spread GBP-LIBOR-6M/OIS-SONIA-6M has been between 17 and 19 basis points. From the end of October to today, the spread has increase from 19 basis points to 30 basis points. On the 3M side, the spread had been between 9 and 11 basis points and increased from 11 to 20 basis point in the same two-month period. In the same period the stock market was tumbling (the FTSE 100 lost more than 10% in the same period) and CDS indices spiking. That make sense as there is a high correlation between low stock prices, high credit spreads and higher default rate.
But what happened to the basis spread for 30Y-tenor swaps LIBOR-3M/SONIA during the same period? Nothing! Or more exactly, the spread that was around 22 basis points did not move in November until the 26 of November (announcement of the ISDA consultation preliminary results) where it went DOWN by 3 basis points in two days and has been decreasing slowly since to 18.30 basis points. This is in line with my prediction of the long term LIBOR market not related to the LIBOR itself anymore but to the new expected fallback rules which is backward looking.
The same phenomenon can be observed in USD, with the USD-LIBOR-3M/OIS-FF-3M spread increasing from 25 to above 40 basis points and in the same time the long term 30Y basis swap moving down slowly from 37.5 basis points on 26 November to 31 basis points now.
To summarise: LIBOR spot credit spread significantly up in the last two months in line with equity and CDS markets, long term credit spread in the basis market steadily down to the historical average.
In the August to October period, the spread GBP-LIBOR-6M/OIS-SONIA-6M has been between 17 and 19 basis points. From the end of October to today, the spread has increase from 19 basis points to 30 basis points. On the 3M side, the spread had been between 9 and 11 basis points and increased from 11 to 20 basis point in the same two-month period. In the same period the stock market was tumbling (the FTSE 100 lost more than 10% in the same period) and CDS indices spiking. That make sense as there is a high correlation between low stock prices, high credit spreads and higher default rate.
But what happened to the basis spread for 30Y-tenor swaps LIBOR-3M/SONIA during the same period? Nothing! Or more exactly, the spread that was around 22 basis points did not move in November until the 26 of November (announcement of the ISDA consultation preliminary results) where it went DOWN by 3 basis points in two days and has been decreasing slowly since to 18.30 basis points. This is in line with my prediction of the long term LIBOR market not related to the LIBOR itself anymore but to the new expected fallback rules which is backward looking.
The same phenomenon can be observed in USD, with the USD-LIBOR-3M/OIS-FF-3M spread increasing from 25 to above 40 basis points and in the same time the long term 30Y basis swap moving down slowly from 37.5 basis points on 26 November to 31 basis points now.
To summarise: LIBOR spot credit spread significantly up in the last two months in line with equity and CDS markets, long term credit spread in the basis market steadily down to the historical average.
Current 3s OIS basis prices consistent with a long term flat forward basis (i.e. fallback spread) of 16.2 bp. The shorter end is less clear though - rising to a peak at 2y then decaying over a number of years. The decay would make sense in terms of uncertainty over when fallback will overtake Libor, but the rise before that not so much. I wonder if some of the price of 3m futures factors in uncertainty over what the future is fixing against, and that uncertainty rises towards 2y - particularly how well the future PV matches the expected value of the IRS it is supposed to hedge. This would slash the historic advantage of owning the future and thus the implied bias in the prices, lowering the longer futures prices (and raising the embedded interest rates).
ReplyDeleteSo perhaps we can see the signs of that in the SEFs SDR data dumps? For the 28th, BGC has a USD 3s1s Mar20 with a size of $2bn, ICAP has a 14x17 (Feb 20) 3s1s on with a size of $1bn, tpSEF has a Mar 20 3s1s 3m at $4bn... So maybe the safer bet is to square the 3s1s fallback risk so the exposure is only to the 3m.