Liability-driven investment strategies commonly involve the use of interest rate swaps (IRS). These are financial derivatives used by banks and other financial institutions for hedging market interest-rate risk. A conventional swap comprises two legs, one fixed rate and the other floating rate which is referenced to an external reference such as the London interbank offered rate (Libor). They are commonly used to transform the interest-rate basis of a debt liability, and because the terms of a swap can be set to suit specific user requirements, they are very flexible risk mitigation instruments. As such they are an important asset-liability management (ALM) and risk management tool in financial markets.
The fixed rate payable on a swap represents bank sector credit risk, if we assume that a swap is paying (receiving) the fixed swap rate on one leg and receiving (paying) Libor-flat on the other leg. If one of the counterparties to a swap is not a bank, then either leg of the swap is adjusted to account for the different counterparty risk that results; usually the floating leg will have a spread added to Libor. We can see that this produces a swap curve that lies above the government bond yield curve, if we compare figure one with figure two. Figure one is the dollar swap rates page from Tullett & Tokyo brokers, and figure two is the US Treasury yield curve, both as at 3 July 2006. The higher rate payable on swaps represents the additional risk premium associated with bank credit risk compared to government credit risk. The spread itself is the number of basis points the swap rate lies above the equivalent-maturity government bond yield, quoted on the same interest basis.
In theory, the swap spread represents only the additional credit risk of the interbank market above that of the government debt market. However, as the spread is variable, it becomes apparent that other factors also influence it. A bank asset-liability management desk will want to be aware of these factors, because they influence swap rates. Swaps are an important risk hedging tool, if not the most important, for banks so it is necessary for market practitioners to have an appreciation of what drives the swap spreads.
Historical pattern
If we plot swap spreads over the last ten years, we note that they have tightened in the last five years or so. Figure three shows the spread for dollar and sterling for the period 1997 to the first quarter of 2006.
We see that spreads have reduced in recent years. The highest spreads for both currencies was reached during 2000, when the 10-year sterling swap spread peaked at around 140 basis above the gilt yield. The tightest spreads were reached during 2003, when the 10-year sterling spread reached around 15 basis points towards the end of that year. At the beginning of 2006, sterling spreads were still lower than the 10-year average of 55 basis points. This implies that the perceived risk premium for the capital markets had fallen by that time.
Note how the change in spread levels coincided with macro-level factors and occurrences. For instance, spreads have moved in line with:
- The Asian currency crises of 1997;
- The Russian government bond default and collapse of the Long Term Capital Management hedge fund in1998;
- The technology stocks “dot.com” crash in 2000;
- The loosening of monetary policy after the dot.com crash and the events of 9/11.
Determinants of the spread
We already noted that in theory the swap spread, representing interbank counterparty risk, should reflect only the market’s perception of bank risk over and above government risk. Bank risk is captured in the Libor rate – the rate paid by banks on unsecured deposits to other banks. So in other words, the swap spread is meant to adequately compensate against the risk of bank default. The Libor rate is the floating rate paid against the fixed in the swap transaction, and moves with the perception of bank risk.
As we implied in the previous section though, it would appear that other factors influence the swap spread. We can illustrate this better comparing the swap spread for 10-year quarterly-paying swaps with the spread between 3-month Libor and the 3-month general collateral (GC) government bond repo rate. The GC rate is the risk-free borrowing rate, whereas the Libor rate represents of course bank risk.
In theory, the spread between 3-month Libor and the GC rate should therefore move closely with the swap spread for quarterly-resetting swaps, as both represent bank risk. A look at figure four shows us that this is not the case. Figure four compares the two spreads in the US dollar market, but we hardly need to calculate the correlation or the R2 for the two sets of numbers. Even on cursory observation we can see that the correlation is not high. Therefore we conclude that other factors, in addition to perceived bank default risk, drive one or both spreads.
These other factors influence swap rates and government bond yields, and hence the swap spread, and they include the following:
Level and slope of the yield curve
The magnitude of the swap spread is influenced by the absolute level of base interest rates. If the base rate is 10 per cent so that the government short-term rate is around 10 per cent, with longer-term rates being recorded higher, the spread tends to be greater that that seen if the base rate is 5 per cent. The shape of the yield curve has even greater influence. When the curve is positively sloping, under the expectations hypothesis (one commonly accepted explanation for the shape of the yield curve) investors will expect future rates to be higher, hence floating rates are expected to rise. This would suggest the swap spread will narrow. The opposite happens if the yield curve inverts.
Figure five shows the pound sterling 10-year swap spread compared to the pound sterling gilt yield curve spread (10-year gilt yield minus 2-year yield). We see that the slope of the curve has influenced the swap spread; as the slope is narrowing, swap spreads are increasing and vice-versa.
Supply and demand
The swap spread is influenced greatly by supply and demand for swaps. For example, greater volumes cash market instruments drive up a need for hedging instruments, which will widen swap spreads. The best example of this is corporate bond issuance; as volumes increase the need for underwriters to hedge increases. However, greater bond issuance also has another impact, as issuers seek to swap their fixed-rate liabilities to floating-rate. This also increases demand for swaps.
Market volatility
As suggested by figure three, swap spreads widen during times of market volatility. This may be in times of market uncertainty (for example, the future direction of base rates or possible inversion of the yield curve) or in times of market shock such as 9/11. In some respects widening during periods of volatility reflects the perception of increased bank default risk. It also reflects the “flight to quality” that occurs during times of volatility or market correction: this is the increased demand from fund managers for risk-free assets such as government bonds when credit-risky assets experience excessive price volatility. The impact of this is to drive government bond yields lower and hence swap spreads wider.
Government borrowing
The level of government borrowing influences government bond yields, so perforce will also impact swap spreads. If public sector borrowing is viewed as being in danger of getting out of control, or the government runs persistent large budget deficits, government bond yields will rise. All else being equal, this will lead to narrowing swap spreads.
We can see then that a number of factors influence swap spreads. A bank ALM or treasury desk will be aware of these factors and assess them accordingly, because the swap rate represents a key funding and hedging rate for a bank.
Moorad Choudhry is visiting professor at the Department of Economics, London Metropolitan University. He is the author of The Credit Default Swap Basis, which is published by Bloomberg and distributed by Kogan Page outside the Americas.
