#MytWeekly: Peripheral yields, the remontada
Peripheral yields, the remontada
The yield increase was accompanied by a similar move in peripherals spreads. We propose an approach in terms of risk premium to analyze whether this deviation is insufficient, justified or exaggerated. That’s the investor’s perspective. What about the borrower’s? Sovereigns have benefited greatly from lower rates to improve their debt service. The rebound of the current yields is a worrying news even if the situation is, for the moment, far from being alarming.
Rebound
At the beginning of the year German 10-year rates were still negative, -0.18% at the very beginning of the year. At the time of writing, they tested 1% and even passed over 1.13% during the first week of May.
To put things in perspective, in the chart below, we look at the change in rates over three months. The Bund gained over 100 bps between January 22 and May 22. A variation of this magnitude is extremely rare, and it is even necessary to go back to the last century, in August 1999, to find a similar movement.
Of course, the tide took with it the other countries and the peripheral spreads have increased in sync. It is interesting to note that the movement has been very homogeneous in other countries as shown in the graph below. The increase in spreads was largely homothetic: by multiplying the spread level at the beginning of the year by a factor of 1.5, the current level is obtained with very high reliability.
To be more precise, and with little more econometrics, we find that 95% of spreads movements since the beginning of the year can be explained by the risk premium. The rest is idiosyncratic. This last part, as we can see, is negligible.
The idiosyncratic risk has moved very little (it is nothing to do with French elections or Italian exposure to Russian gas), it is very essentially a common movement of widening spreads, thus an increase in the risk premium.
This may be due to the risk on growth associated with the Ukrainian crisis and other factors. It is also closely linked to the ECB’s announced reduction of QE, which has been already significantly reduced and is likely to end this EQ at the end of June.
Risk premium
The question for investors then is whether the current risk premium is excessive, reasonable, or still below normal.
To do this we look at the chart below which compares the ratings of the 11 main euro zone countries and their spreads. Of course, there is a very close relationship between the two, with investors demanding an increasing premium in compensation for the deterioration of ratings.
At the time of writing, the slope of the curve on the graph is 16 bps per rating unit excluding Greece which tends to bias estimates. In other words, each unit’s rating movement (called a “notch” in market parlance) results, on average, in a change in the country’s risk premium of 16 bps. This figure is therefore a measure of the risk premium.
At the very beginning of the year, spreads were lower, the slope was flatter, close to 11 bps. This is a good illustration of the fact that the risk premium has increased since the beginning of the year. And we find back the magnitude of the adjustment mentioned above.
The graph below shows the evolution of this risk premium since 2015. Why 2015? On the one hand because the Banking Union was introduced at the end of 2014 and has significantly changed the risk profile of European states by helping to break the vicious circle with banks. For more details, see MyStratWeekly from last week. On the other hand, because the period before that date was that of the sovereign crisis where risk premiums have reached galactic levels that have nothing to do, at least we can hope so, with a “normal” world. It therefore seems to us that the period reviewed provides a relevant basis for comparison.
If this is the case, the message of the chart is relatively simple. Rather, it was the level at the beginning of the year, with a risk premium close to 10 bps that was abnormally low. While the current level, around 16 bps, is close to the long-term average. We understand what was happening at the beginning of the year, the ECB’s QE, still very dynamic, artificially compressed the yield spreads. Hence an unusually and artificially low risk premium. As the ECB withdraws, market prices return to a more “normal” level.
In conclusion, there is no alarming sign of oversupply of markets or of panic. Far from it.
领英推荐
So we’re not necessarily done yet.
Bearable?
So far we have only taken the investor’s point of view and analysed the return it requires. What about the borrower? Higher rates are a higher interest burden and therefore a problem for public finances.
In a previous MyStratWeekly we introduced the idea of “apparent interest rate”. In economists' jargon, it’s the observed debt service, divided by the level of debt, so it’s the actual cost of the public debt. Let’s take an example: according to OECD figures, French debt was 2.850 billion in 2021 and debt service was 26 billion over the year. The ratio between the two is 0.92%, which is the average rate paid by the CRF on French debt last year – the “apparent rate”.
To go back to the case of France, the “apparent rate” is therefore lower than the 10 year rate, which is close to 1.5% at the time we write. Market rates have remained almost systematically below the apparent rate since the beginning of the century as shown in the chart below. And so the issuance of new debt is at a lower level than the historical rate, which implies a gradual decline in the “apparent rate”. Unfortunately, in the case of France, the 10 year has just risen above it.
We need to be more specific, however. Of course, countries borrow not just at a 10 year maturity, but over the entire curve. The cost of issue is therefore an average of the rates at different maturities. On the graph below, we calculated the average rate for the five major euro area countries, weighting the maturities according to the size of the existing borrowings. This approach therefore gives a good approximation of the average rate at which new borrowings will be issued.
To come back to the case of France, our calculations show an average rate at 1.08%, which is slightly higher than the apparent rate of 0.92%. This means that, as the French Treasury issued, the average rate paid on loans with a tendency to rise. But we are still at very low levels, and progress is very slow. In fact, it would be more appropriate to talk about stabilization.
It should be noted that Germany and the Netherlands are in the same situation with average rates on their curve which, again, have slightly passed over the apparent rate.
In contrast, Italy and Spain remain in a more favorable situation. This is the consequence of the much higher spreads that existed on these countries in particular before 2015 and which therefore raised the historical cost.
In general, however, the gap is small. For all these countries, the market levels are therefore somewhat similar to the rates paid by the countries on their existing debt. While all these countries have benefited from an impressive reduction in their debt costs, it seems that the trend is coming to an end. To give an order of magnitude, the OECD estimates the French debt service at 26.1 billion in 2021 compared to 50.4 billion in 2010. While over the period the ratio of debt to French GDP has gained almost 30 ppt.
Generally speaking, therefore, there is no real danger. Market levels allow sovereigns to issue debt at a broadly stable cost compared to the existing one.
This view is nevertheless to be nuanced. If the photo is not particularly worrying, the trajectory is more challenging. A rise in rates, or continued spreads, would quickly put levels of borrowing rates on a much more complicated trajectory for some countries. It should also be remembered that the greater the debt stock, the greater the sensitivity of public finances to a rise in interest rates. We’ve had higher rates, and even much higher rates in the past, but the states didn’t have the level of debt that they currently have. This hypersensitivity is also a source of risk.
Conclusion
So far so good. The increase in rates has certainly been unusually rapid, since we have to go back to the 1994 bond market to find a more violent one. Despite this, peripheral spreads levels have finally returned to normal. This is to the extent that the average change since 2015 can be described as “normal”. Rather, it was the ECB’s aggressive QE period during the Covid episode that had been unusually tight on spreads.
We reach a similar conclusion if we take the point of view of the borrower, that is to say of the sovereign. The average issue cost for the five major countries is currently close to the average cost of their existing debt. If the curves stabilize, the apparent interest rate on government debt should therefore remain stationary. Here too the movement looks, to some extent, like a return to normality.
An overreaction of the rates is nevertheless entirely plausible. In this case the very high debt stock creates a high sensitivity to rates and the situation could quickly become much more complicated. So, we have to hope that the adjustment is largely finished and that the markets will not decide to “over-adjust”. This cautious conclusion can also be linked to rumors of a sovereign spreads control mechanism designed to avoid this dynamic and the possible loss of issuer market access, as in 2011 with the SMP.
Stéphane Déo
#ECB #ZoneEuro #Peripheral #Rates #Yields