The inauguration of the new 20Y bond is due to take place next week. The bond, named SGB1056, will mature on 1 June 2023 and will be introduced through a syndication in switch auction. The SNDO buys the 8Y and 10Y bonds SGB1047 and SGB1054, respectively, and sells the new bond. The switches will be made in notional terms, i.e. buy-backs of the same notional amount of SGB 1047 and/or SGB 1054 as issued in SGB 1056. Investors may choose any combination of the two bonds. The coupon will be set at the time of launch and the lead managers will announce this information.
Valuation
A fair valuation of the new 20Y bond is a question of how much one should take the German bond and EUR swap curves into consideration. The hump-shaped EUR swap and German bond curves are a result of, and got that shape, when Danish and Dutch L&P companies started to receive ultra-long EUR swaps a few years ago. The peak of the hump has moved to the 15Y segment. We see little reason why the Swedish bond curve should have the same extreme shape as the European ones even if it is influenced somewhat. EUR swaps are less appropriate hedging rates today, following the debt crisis, as the correlation with Swedish rates has diminished. The question is at what yield Swedish L&Ps want to switch some of the holdings in 10Y bonds into the new 20Y bond and achieve a better cash flow match between assets and liabilities in their base currency. The alternative is to stay with the current mismatch or to use European rates with a poorer correlation with how Swedish rates move. Pricing in line with Europe should result in a yield above SGB1053, a challenging pricing perhaps.
So, we put most weight on the Swedish bond and swap curves when carrying out the valuation but spice it up with some influence from the European term structures. Moreover, in our view, liquidity will probably be limited and a small new issuance premium should be taken into consideration. The valuation is done on the assumption that all curves are priced correctly at 14:30 CET on Wednesday 7 March. Moreover, we assume that the coupon will be set to 2.25%.
Our valuation suggests a spread 40-45bp above SGB1054 (1 June 2022). Given the considerations above, we are in favour of the higher end of this range. The spread of 45bp above SGB0154 is equal to a yield some 5bp below the 30Y bond (SGB1053) and provides a small hump shape of the term structure. In relation to the swap curve, it implies a spread at -28.5. So, our valuation is not a simple linear interpolation along the bond and the ASW curves but introduces a small premium. However, in our mind, an even higher spread above SGB1054 would give a too-accentuated hump shape and a too-elevated forward rate.
Moreover, comparing the ASW curves in Sweden and Germany supports our stance that the Swedish government bond curve should not have the same hump as the German government bond curve (see Chart 1). The 5-10Y segment is much flatter in the Swedish bond curve versus SEK swaps than in the German ASW curve. So, why should a much more pronounced hump along the Swedish bond curve suddenly be introduced beyond the 10Y segment?
The 45bp spread above the SGB1054 implies a BEI rate for SGBi3104 at 1.65%, resulting in a forward BEI rate between 3102/3104 at 1.87%, which also seems relatively fair.
Our valuation of the new 20Y bond SGB1056 suggests a spread at +40-45bp above the 10Y bond SGB1054. However, given the extreme hump shape of the European curves, the limited liquidity and a small issuance premium, we believe a spread at the higher end of the interval is a decent pricing at issuance. The charts on the next page show how the pricing of the bond affects curves.