view, this autumn’s Riksbank trade with best risk/reward is the yield spread compression trade, buying the 5Y covered bond and selling the 5Y government bond (note that this can be done in bond futures, although SGB1047 is not the underlying bond in the 5Y government bond future). We especially like the current timing of this trade. Let us explain. The pending QE purchase programme is having greater and greater impact on the bond market, which is what you would expect given that the Riksbank now holds some 35% of issued stock and roughly 45% in certain ISIN codes. So, we are getting closer to the point when, if the Riksbank needs to extend QE, other assets have to be considered (if our inflation forecast for the rest of the year proves correct the pressure on Riksbank will mount again – see chart 3). In such a case, covered bonds are the only asset that offers enough market liquidity and size. So, in our view, another extension of QE would mean the market would start to price in a premium for this possible scenario in the spread covered bonds/govies. In our opinion, no such premium is discounted in the current traded yield spread. Another possible scenario unfolding in the autumn (although we assign the above scenario a much greater probability) is that the Riksbank inflation path materialises and it decides to end QE purchases when the current programme ends at the turn of the year. In our view, this would lead to a clear repricing of the government bond curve with a clear steepening. Also, in this scenario, we would expect covered bonds to perform relative to govies. In short, the current path of QE, purchases only in government bonds, is soon to end. Either the Riksbank needs to expand purchases to other assets or the QE programme is getting closer to an end. Hence, we see good risk/reward for the performance in covered bonds relative to government bonds.
Nominal bonds are being squeezed more in the repo market and trade closer to the ‘floor’ in the repo market. Market makers can get bonds from the Debt Office at the Riksbank repo rate of -40bp. This sets more or less a limit on how much the (squeezed) repo market can affect bond yields. The current situation in the repo market makes it expensive to finance the short position in the government bond and does ‘cannibalise’ on the yield spread between the assets. However, the difference in finance cost relative to yield spread is currently not particularly stretched but rather at a two-year average (see Chart 1). The spread in the repo market is between 30bp and 40bp depending on which two bonds are compared. The repo market for govies is close to the ‘floor rate’ and any change in QE purchases, as in the two scenarios presented above, could change the situation in repos. In addition, amid the hunt for yield, we see more new foreign real money investors becoming more interested in the Swedish covered bond market attracted by the yield spread to government bonds and the pickup (see chart 2) (for these investors, the repo market do not really matter). In all, it is hard to see the situation in the repo market deteriorating further and becoming a big obstacle for the yield compression trade.
To read the entire report Please click on the pdf File Below