One month ago, Citi’s credit team laid out its outlook for the coming year in what – in our opinion – was one of the gloomiest reports for the coming year, and included the following fascinating revelation according to which central bankers appears to have lost control: “That seems to be a growing fear among a number of central bankers that we have spoken to recently. In our experience, they too are somewhat baffled by the lack of volatility and concerned about the lack of response to negative headlines…. Our guess is that sooner or later in the process of retrenchment they will end up going too far – though that will only be obvious with hindsight.” As we said at the time, “frankly, that’s about the scariest admission from one of the world’s biggest banks that we have read in a long time.”
And while Citi’s “base case” was clearly bearish (our summary can be found here) – what was left unsaid was even more interesting, if not troubling. As the bank’s credit team writes “what about the outcomes that didn’t quite make it into our base case? The scenarios that aren’t central, but which aren’t entirely implausible either – both bullish and bearish.
So ‘What if…”:
- idiosyncratic risk is returning to credit?
- European corporates get more aggressive?
- global growth & commodity prices disappoint?
- inflation accelerates as output gaps close?
- the US yield curve inverts?
- central bank tapering really is a non-event?
- the market doesn’t like the choice of ECB successor?”
Here is Citi’s take on each of these possible, if not necessarily probable, scenarios:
1. idiosyncratic risk is returning to € credit?
New Look, Astaldi, and now Steinhoff. If it felt as if idiosyncratic risk was as dead as a dodo in the age of ECB QE, well then, now we know it isn’t. All three credits have at one point seen their bonds drop by more than 25 points in the last few weeks. A mere coincidence or a sign of things to come? With European default rates at 1.5%, strong fundamentals, healthy earnings growth and broad economic growth the best-guess answer has to be “mostly the former”.
But not exclusively so. As we illustrated in our derivatives outlook, the compression of risk premia has been accompanied by a risk-bifurcation in € credit. Well over 90% of credits are historically tight and carry very little of the overall default risk of the market, which is overwhelmingly skewed to a small number of names. Such a bifurcation happens in every bull cycle, but it has been augmented by all the central bank excess liquidity. The government bond buyer, crowded out by the ECB or money market trying to escape negative yields, moved into high-rated corporates. The buyer of high-rated corporates no longer finds sufficient returns and so moves down in quality. The eventual consequence of the excess liquidity is that the whole market ends up priced almost to perfection, except of course the credit that is at high risk of relatively imminent default. It will still trade to its recovery value (Figure 1).
The point we are trying to make is that this inevitably results in a much sharper cliff in pricing between those that are deemed to be going-concern investments and those that are deemed to be over the edge. Hence the precipitous losses (or gains) when a name shifts from one to the other.
But beyond this, blow-ups do tend to come in clusters. It wasn’t just Enron; it was also Worldcom. It wasn’t just Ahold; it was also Tyco. Even if recent blow-ups have no direct bearing on other risky credits, they serve to remind investors of the risk of standing close to the brink. And there is a general tendency for, shall we say, ‘credit homework’ that gets overlooked during bull markets to be rediscovered in a hurry once losses have been incurred. As analysts rummage around there is an increase in the likelihood of something being uncovered in another name. At an absolute minimum, when large quantities of debt have been raised, as the market gets more nervous there is an increased probability of an outsized knee-jerk reaction. This has no doubt contributed to some of the recent underperformance of the HY market versus IG (Figure 2). With curious synchronicity, a similar set of worries seems to have reared its head in other regions (think Windstream and Community Health in the US, or HNA, Noble Group and Kobe Steel in Asia).
Steinhoff is, of course, rather an odd one out in that its investment-grade rating meant it seemingly was nowhere near the fringe. But accounting irregularities, though rare, are, if anything, far worse in that they tend to undermine confidence in the fabric of the system. That it is a bolt out of the blue will be cold comfort for the several funds, which according to public holdings data, held well over 1% of their assets in the 2025 bond and which must have seen a >30bp hit to the aggregate performance as a result of this one blow-up.
It has even been bought by the ECB, though that should be less of a concern. As we discussed last week, the ECB may choose to sell holdings at its discretion but is not forced to do so as a result of a downgrade or even a sudden drop in prices. The Eurosystem can take losses (which will in any case be minute relative to its €2.2tn bond portfolio and ability to absorb them) and there is an agreed mechanism for loss sharing.
We still think the risk of a material rise in idiosyncratic risk in Europe next year is low (our guess would be around ~15%), but each event is progressively more toxic for the broader market, especially if it hits places where there isn’t supposed to be much risk. Solutions? We only have the obvious ones really – favour transparency and diversify your portfolios. Look carefully at groups with acquisitive track records and complex structures, and try to follow the cashflow rather than the earnings! Credit 101 that may be, but all too often it is forgotten in the increasingly desperatehunt for returns.
* * *
2. European corporates get more aggressive?
While we do expect corporates to lavish more cash on their shareholders, material releveraging is not our central scenario. Rising earnings also in 2018 (Figure 3) should facilitate greater payouts without raising net debt to EBITDA. Given the large and persistent divergence between spreads and fundamentals in recent years already (Figure 4), we suspect the rise in net debt would have to be quite substantial to actually trigger a repricing in spreads.
To our minds, the bigger question is whether the market would seamlessly absorb a corresponding increase in issuance. There is almost a 1:1 correlation between M&A volumes and net issuance as illustrated in Figure 5. It is not inconceivable that a few large acquisitions could add €50bn or more to our net supply estimate, which already sees a significant increase in the net private funding requirement (Figure 6) compared to previous years. The CSPP notwithstanding, a reduction in inflows to credit of crowded-out government bond money when spreads are near record lows makes the credit market quite vulnerable to such an increase in our opinion. We think the risk of a sharper increase in M&A (and buybacks for that matter) than in our central scenario it quite high: at a guestimate around 25-35%. The increase in funding requirement would likely lead to spreads to ending modestly wider than in our central scenario.
* * *
3. global growth & commodity prices disappoint?
As our economists have argued, current risks to their global outlook if anything seem skewed to the upside. But we don’t have to go back further than to 2015 to see what a sustained period of negative surprises in the economic data can do to sentiment over time – € IG credit spreads reached 160bp in February 2016 (versus about 45bp now). A repeat of the rout in commodity markets, which triggered – or at least reinforced – that soft patch seems comparatively unlikely right now. Our commodity strategists expect strong, sustained demand for the next couple of years. However, a greater than expected slowdown in China, where both fiscal and credit impulses are weakening considerably, remains a risk to both global growth and commodity prices. In 2016, China’s contribution to global GDP growth was 1 percentage point, or more than 40% of the total.
If the global economy were to experience a similar negative dynamic in 2018, we believe it would once again be a very painful experience for credit. Fears of secular stagnation and deflation still linger and could easily be rekindled. Some of the excess has probably been taken out of the $ HY market since then, but the energy sector remains the third largest in terms of outstanding and is still highly sensitive to fluctuations in prices. Moreover, with € spreads already tight to fundamentals, they would be highly vulnerable to a weakening in the growth and earnings outlook. As in 2015, we suspect that the response from most central banks would lag the market considerably. However, within € credit specifically, it would greatly increase the chance that the ECB ultimately decided to extend the CSPP beyond September 2018. This would not prevent widening in our opinion (despite the CBPP, covered bond spreads widened substantially in H2 2015), but it would probably act as a
We’d informally put the probability of such a scenario in the region of 10-20%, but if it did occur spreads would in all likelihood end substantially wider than our 80bp base case.
* * *
4. inflation accelerates as output gaps close?
The chance of a structural shift in inflation dynamics remains pretty small, in our view, and it is worth highlighting that inflation has undershot Citi’s expectations six years in a row. But in a cyclical sense, it is feasible that ever lower unemployment rates in key economies and greater pricing power do eventually take us far enough out on the tail of the Phillips curve that we start to see upside surprises to inflation. Core inflation is already firming in the US after a succession of weak prints this summer. Across advanced economies, the unemployment rate was 0.4pp below the OECD estimate of the NAIRU in Q3 2017, and wage settlements in places like Japan and Germany will be important to watch. And as noted above, strong demand for commodities or supply disruptions could also pose upside risks to prices.
To the extent that higher inflation reflects higher wage growth, resulting from greater productivity gains and more corporate pricing power, the immediate impact on corporate fundamentals would be rather modest. However, to the extent wage growth exceeds productivity growth and pricing power, it might start to dent earnings growth.
However, a more important impact on credit comes through the central bank response function and market returns. A confirmation of the ECB’s inflation forecasts would greatly increase the chance that asset purchases are terminated in September. It would likely also temper attempts over the coming months from governing councillors to jawbone the market into postponing the implied timing of a future rate hike. In the US, it would cement the path implied by the dot plot and potentially open up scope for a fourth rate hike, which would lift Fed funds to 2.25-2.5% by the end of the year.
Such a scenario would also likely increase yields at the long end, relative to our rates strategists central scenario, taking total returns at least in € credit further into negative territory. While higher yields would likely be a net benefit to flows into credit over time, we believe there is enough money invested in € credit on a total return basis that the process of adjustment in risk-free yields would see outflows from corporate bond funds, in turn leading to wider spreads in the short term.
We would conjecture that the probability of such as scenario is around 20-30%. It would probably result in spreads somewhat above our central scenario.
* * *
5. the US yield curve inverts?
US 1s10s, 2s10s and 2s30s are all the flattest they’ve been since the GFC. While 2- year yields have already risen 50bp since September, a rate hike in December and three rate hikes next year would leave them below the Fed Funds rate, highlighting that there is still upside risk. Conversely, market reluctance to revise terminal rates higher makes an inversion of the 2s10s yield differential, currently at 50bp, a possibility.
The debate about what an inverted yield curve would imply for the future growth outook in the post-QE era has been raging for some time already. We’d subscribe to the view that QE has left long-end real interest rates lower than they otherwise would have been with the implication that outright curve inversion per se might not have quite the same implication for forward recession probabilities that a historical analysis would suggest. However, considering that the flattening is occurring at the very time that the Fed is beginning to reverse its grip on the long end does send an important signal in our view. The market may countenance Fed hawkishness near term, especially in light of tax reforms, but continues to take a rather dim view of the longer-term sustainability of current above-potential growth rates. As we have shown, this curve flattening empirically holds a strong (negative) signal for the future path of credit spreads.
But beyond that, the flattening of the $ curve is a problem for credit because it reduces the opportunity cost of not being invested. At the moment the yield on $ IG corporate credit is 3.2%, while the 1-year risk-free yield is about 1.6%. Another 4 rate hikes over the coming year could easily take the latter to 2.25-2.5%, which then raises the question whether the marginal investor is prepared to take 7½ years of duration risk and BBB credit risk for potentially less than 100bp pickup? If the answer is no, then either spreads will widen or long-end yields need to rise. In either scenario, it would impact total returns, which in turn in the short-term would dent inflows to credit. This might be a dollar rather than a euro story, but given that the historical correlation between corporate spreads in the two markets is more than 85%, it would almost certainly have an impact on € IG as well.
The magnitude of the impact is contingent on the exact driver of the US treasury curve inversion – a bear flattening is probably ultimately less “damaging” than a bull flattening would be, because of the very clear signal about growth that would send. So for the latter, in particular, we would again argue that € IG spreads would end up wider than in our central scenario. Our guesstimated probability of inversion next year is in the region of 15-25%.
* * *
6. central bank tapering really is a non-event?
If you’ve stumbled upon any of our research over the last couple of years, you’ll probably know that we have bees in our bonnets about the impact that QE in a broad sense is having on markets. Crucially, we strongly believe it is the flow, i.e. the pace of net central bank purchases, at a global level, which does the lifting of asset prices. Many market participants and most central banks seem to believe it is the stock. Without going into the whole debate again, it is worth contemplating what our central scenario would look like if we are wrong.
So what happens if reducing the free float of securities and increasing excess liquidity by shifting supply and demand curves has a permanent impact on asset prices and taking away the backstop increases risk appetite among investors, allowing an increase in private flows to seamlessly replace the $1tn drop in central bank net demand in 2018? Obviously, that would, in isolation, be very good news for asset prices. It implies a continuation of excess demand.
Such an outcome is evidently more bullish than our base case, but would it imply actual tightening from current levels? Well, € credit would in all probability still face a drag on spreads from the dwindling opportunity cost discussed above, the direct effect of ending the CSPP, more net supply and a maturing cycle. That said, as illustrated in our most optimistic scenarios in the 2018 Outlook, it might still imply a small net tightening for the year as a whole, which would translate into an excess return (to swaps) of 0.5-1.0%. As in our other scenarios, we think tightening would be front-loaded, with a partial give-back in H2 2018.
And the probability of us being wrong on what is at the crux of our central scenario? We’ll leave it to you to judge.
* * *
7. the market doesn’t like the choice of ECB successor?
We often get asked who will succeed Mario Draghi at the helm of the ECB when his 8-year term expires at the end of October 2019, and what it will do to markets. It is a dark horse for next year (and hence we have included it), but in truth we think the probability that it becomes a major issue already in 2018 is low. The choice of Mario Draghi at the end of Trichet’s tenure was only made public in June 2011 – 4-5 months before the changeover. Indeed, until February 2011 the front-runner was Axel Weber.
We doubt if Draghi’s replacement will be apparent even by the end of 2018. The market might conceivably react as candidates emerge more clearly, and if it does most likely it is due to fears of a shift in a more hawkish direction. At the margin, Portuguese Finance Minister Mario Centeno’s appointment as President of the Eurogroup after Dijsselbloem ought to increase the chance that the next ECB President will come from a core country. But we’d informally put the probability of it becoming a major driver of spreads already in 2018 below 5%.
* * *
We’ll continue with Part II of Citi’s “What If” scenarios tomorrow, in which the bank reveals its final 7 questions for the next year.