Monday, 19 June 2017

Macro and Credit - Circus Maximus

"I can calculate the motion of heavenly bodies, but not the madness of people." -  Isaac Newton

Watching with interest new records being broken in equities reaching new highs, with credit "carrying on" thanks to uninterrupted inflows into funds ($6.5bn of inflow into European IG credit funds and High grade funds recorded another week of inflows; their 21st in a row according to Bank of America Merrill Lynch), we reminded ourselves for our title analogy of the Circus Maximus, the ancient Roman chariot racing stadium and mass entertainment venue located in Rome, which was the first and largest stadium in ancient Rome and its later Empire. It measured 621 meters (2,037 feet) in length and 118 meters in width and could accommodate over 150,000 spectators. In its fully developed form, it became the model for circuses throughout the Roman Empire. While we have been quite comfortable riding the uninterrupted bullish tide from our short term "Keynesian" perspective as indicated in our earlier musings of 2017, as of late, we indicated that tactically we were reducing our beta exposure credit wise and that we would rather stick to quality given that not only we feel that the credit cycle is slowly but surely turning, but, it feels like when it comes to the abundant generosity from our "Generous Gamblers" aka central bankers, when it comes to the Fed and the latest FOMC, it feels like the tide is turning. We posited in the past the following quote, which was a derivation of Verbal Kint's quote in the Usual Suspects movie:
"The greatest trick central bankers ever pulled was to convince the world that default risk didn't exist" - Macronomics.
We have used in the past as a title for a post a reference to the great text from Charles Baudelaire called the "Generous Gambler". This poem appears to be the 29th poem of Charles Baudelaire masterpiece Spleen de Paris from 1869.

As one knows, the "Devil is in the details" and if indeed the Fed is serious with its hiking plan and balance sheet reduction, then it does indicates that, regardless of the slowly turning credit cycle and the flattening of the yield curve, there is an increasing possibility of the liquidity tide to turn, you have been warned. For now in all markets it's "Circus Maximus" as we move towards the euphoria stage, with additional melt-up in asset prices, in the final innings of this great credit cycle we think.

In this week's conversation, we would like to look at why credit is "carrying on" and why we prefer for the moment playing it safe via Investment Grade in case volatility heats up in the near future.

  • Macro and Credit - If you love low volatility, stick to investment grade credit
  • Final charts - Italy? It's getting complicated

  • Macro and Credit - If you love low volatility, stick to investment grade credit
Reminiscences of a flow credit operator comes to mind in true 2007 fashion these days given the continuous inflows into Investment Grade funds, in conjunction of continuous tightening in credit spreads. Discussing recently with another credit pundit on a macro chat platform, we reminded ourselves of the 2007 spread compression with the current situation:
Him - "I've had banks pretty much telling me I can name my price in itraxx tranches, off the back of structured stuff they've issued to retail."
Me - "Not surprising, this is playing out like 2007, structured products issuance leads to relentless selling in CDS which is compressing even more spreads and market makers can't recycle in the market because the only takers would be loan books buying protection. Lather, rinse, repeat..."
Of course both of us like many others, have seen this movie before. As we pointed out in our recent musing, when it comes to High Yield and in particular European High Yield, given the tightness of the spreads relative to Investment Grade, we have switched to being tactically negative, underweight that is. Sure, some would argue that, it's all about the carry and that given the ECB's supportive stance, it's all about "carrying on" through the summer lull. We have a hard time seeing the relative attractiveness in Europe of € High Yield versus € Investment Grade credit as pointed out by Deutsche Bank in their European Credit Update note from the 13th of June entitled "Carry is King...For Now":
"Credit Valuations and Spread Forecasts
Despite the change in view we still think it will be difficult to see much spread tightening. More specifically, we would argue that valuations appear to be more stretched for EUR HY relative to EUR IG credit. Looking at our often used analysis in Figure 2, showing where spreads currently trade relative to their own histories, we can see that EUR HY spreads are generally at the top of the list.

More specifically BB and B spreads are well into their tightest quartile, in fact for Bs current spread levels are now close to the tightest decile through history. In addition looking at Figure 3 we can see that the HY/IG spread ratio remains around the all time lows having seen HY outperform on a relative basis since September last year.
We should stress here that we think the outperformance would be on a risk-adjusted basis. Given that we think carry will be driving returns from here, the extra yield available (although low in absolute terms) could see absolute outperformance. One other thing to potentially consider is the stark difference in duration of EUR IG vs. HY. The duration to maturity on the IG index is 5.4 years while for the HY index it is 4.1 years. This also does not take into account the fact that many HY bonds are trading to their call date rather than to maturity and therefore the duration of the HY index is likely to be even lower in practice. This could create some further benefit for IG if we do see more spread compression.
The recent trends in ECB QE could also benefit this view. The ECB appears to have trimmed corporate purchases less than government bond purchases as shown in Figure 4.

(For more details, see our report “CSPP Trimmed Less Than PSPP, with a Record Share of Primary Purchases” available at In addition, the latest ECB communication has implied that the overall withdrawal of QE could be even slower. Combined with the Italian election risk seemingly pushed back into Q1 2018, these latest developments are conducive to low volatility and tight spreads. In fact, as the market repriced QE for longer with a relatively increased emphasis on corporate purchases so far, CSPP-eligible bonds have again started to mildly outperform ineligible ones (Figure 5).
With the change in view we have also updated our year-end spread targets and what they translate to in terms of excess returns (Figure 6).

We have also trimmed our view on defaults given the combination of the positive macro data and continued low yield environment. The change in view is focused on the next few months and we will likely review this again as we move towards Q4. We have effectively pushed out our moderate widening view given the events of the last few days. And while we now expect the carry environment to last longer than we originally thought, the moderate widening stage might set in before the year end. Our end-2017 forecasts should therefore be seen with that caveat in mind.
Note that the forecast performance is risk-unadjusted. While higher-beta HY may outperform lower-beta IG in absolute terms, we expect the latter to outperform in risk-adjusted terms as mentioned earlier. On the margin, we expect the following relative performance adjusted for risk, which is in line with our previous published views:
  • IG outperforms HY
  • Financials to outperform non-financials
  • Senior financials outperform subordinated financials
  • EUR credit outperforms GBP credit
The UK elections have only reinforced our view that more macro uncertainty currently hangs over the UK economic outlook than over the eurozone, which makes us relatively more cautious on GBP credit.
The above cash bond excess returns are over government benchmarks. The total returns will additionally depend on the performance of government bonds." - source Deutsche Bank
This ties up nicely with our recent call in favoring style over substance, quality that is, over yield chasing from a tactical perspective. While the Italian elections probability has been postponed and has therefore moved from a "known unknown" towards a "known known", there is still a potential from a geopolitical exogenous factor to reignite in very short order some volatility, which would therefore put a spanner in the summer lull and "carry on" mantra. But as the Circus Maximus goes on, with very supportive inflows into the asset class, the carry trade continues to be the trade du jour for fixed income asset allocators. On this subject we read with interest Société Générale Credit Wrap-up from the 14th of June entitled "Why fixed income allocators are loving credit":
"Market thoughts
Risk-adjusted returns have become an important tool for fixed income allocation in recent years. Volatility and correlations tend to follow regimes, meaning that the past can be a decent approximation of the future under most conditions. Yields give a reasonable sense of what returns could be, unless the market moves sharply in one direction or another. Thus, dividing current yields by historical volatility is a useful tool for knowing which fixed income classes look most attractive at the moment.
Such calculations flatter the credit markets. Chart 1 uses the risk adjusted yield from our quant team (which they define as current yield divided by the standard deviation of yield over the past year) across a range of fixed income asset classes.

The horizontal axis shows the current level, and the vertical axis shows the change in the level since the end of 2016. Credit has suddenly jumped to the top of the pack, ahead of emerging markets, thanks largely to the very low level of volatility in returns since the start of the year.
But past performance is no guarantee of future returns, as every careful reader of financial boilerplates knows. Will corporate bond volatility remain low?
In the short term, yes, quite possibly, and this is one of the reasons we are bullish on the asset class into 4Q (as explained in Why credit markets could test the summer of 2014 tights). In the longer term, however, two factors worry us. First, the level of equity implied volatility is near 30-year trough levels. It could stay there for a while (as we explained in How this period of low volatility will end), but volatility always eventually gets driven higher by credit problems, and rising leverage – mostly notably in China – could be the trigger once again by early next year.
Second, the low level of credit volatility has been driven by a very high negative correlation between government bond yields and credit spreads. In Why government yield/credit spread correlations will likely reverse next year, we noted that there are fundamental reasons for thinking that this correlation could change in 2018. If it does, then the volatility of corporate bonds will necessarily rise because yields will not be a shock-absorber for spreads and vice versa.
Therefore, a word of warning. 2H may bring further inflows into credit from other parts of the fixed income world as asset allocators look at indicators like this one and draw optimistic conclusions. This may however spell the final phase of the credit market rally, for lower spreads and changing correlations could make the risk-adjusted yields look much less enticing come 2018." - source Société Générale.
We agree with the above, namely that from a risk/reward perspective, we have moved at least in European High Yield space from expensive levels to very expensive. While we had some confidence in the rally so far in the first part of the year, no doubt we could see additional melt-up in asset prices, credit included but it remains to be seen how enticing the risk-adjusted yields will look in 2018 in the Circus Maximus. The pain from rising yields and the continuation of the flattening of the yield curve will probably come to bite at a later stage.

When it comes to Deutsche Bank's recommendation in playing Senior financials versus subordinated financials, it makes sense, given the latest Banco Popular bloodbath. On this subject we read with interest Euromoney's article from the 13th of June entitled "Spain’s bank rescue could make tier 2 less Popular":
"Both AT1 and tier-2 investors lost everything when Banco Santander rescued Banco Popular, while senior bondholders were untouched. The rescue has shown that when banks in Europe get into trouble it is liquidity, not capital, that matters and that the fate of subordinated bondholders is anything but predictable.
There is no doubt that the AT1 market took the surprise bail-in of Banco Popular’s subordinated debt and equity in its stride. Despite its tier-1 and tier-2 debt trading at around 50c and above 70c, respectively, the day before, both became worthless when Spain’s Fondo de Reestructuración Ordenada Bancaria placed the bank into resolution on Wednesday, imposing losses of around €3.3 billion on debt and equity investors. The market has been patting itself on the back ever since, calling the sale of Banco Popular to Banco Santander for €1 a textbook outcome. Peripheral bank AT1s barely stirred. According to CreditSights, these bonds traded down over the course of the following week, but many by less than a point. 
Overall, the AT1 market has been trading close to 12-month highs, the memory of the market disruption caused by questions over Deutsche Bank’s ability to meet coupon payments on its AT1 paper in the first quarter of last year seemingly a faint one. Even peripheral names only fell to around 85% to 90% of those 12-month highs after Popular’s wipeout. Sorely needed This was a sorely needed win for the EU’s Bank Recovery and Resolution Directive (BRRD) after its shaky start with Novo Banco at the beginning of 2016 and the protracted horse trading over state support to Italy’s troubled MPS ever since.
The overnight move by the Single Resolution Board (SRB) saw Popular’s AT1s pushed past their CET1 triggers by the extra provisioning that Santander has demanded to take on Popular’s €36.8 billion of non-performing assets – just 45% of which were covered by provisions. So far, so good. However, there are a few things about this bail-in that are not exactly text book as well. The whole point of contingent convertible tier-1 debt is that it has triggers: Popular’s were set at 5.125% and 7%. It has two AT1 deals outstanding – a €500 million 11.5% low trigger deal and a €750 million 8.25% high trigger deal.
In the bank’s Q1 2017 presentation, its CET1 was 10.02% ­– still a long way from both of those, although its fully loaded CET1 ratio was closer to 7.33%. However, Popular had never missed a coupon payment on any of these notes: if this situation had really been played by the book that is what should have happened first. This is what the hoo-ha over Deutsche Bank’s available distributable items was all about last year. Banco Popular's situation has shown that the fate of subordinated bondholders actually has very little to do with the precise structure of the instruments that they are holding. Neither of Popular’s tier-1 notes had breached their triggers before the SRB decided that the bank had become non-viable late on Wednesday. Indeed, European Central Bank (ECB) vice-president Vitor Constâncio has clearly stated that the bank’s solvency was not the issue.
“The reasons that triggered that decision [to deem the bank non-viable] were related to the liquidity problems,” he explains. “There was a bank run. It was not a matter of assessing the developments of solvency as such, but the liquidity issue.”
There certainly was a bank run. €20 billion left Banco Popular’s coffers between the end of March and June 5 – the same day that its chairman Emilio Saracho declared that he did not plan to request emergency assistance from ECB because it was not necessary. The sale to Santander is understood to have been put together in less than 24 hours.
It was thus depositor withdrawals that caused the SRB to deem the bank to be failing or likely to fail under Article 18 (1) of the Single Resolution Mechanism Regulation. This meant that it had hit the point of non-viability (PONV). 
Investors in AT1 instruments should, therefore, be paying much closer attention to the PONV language in their documentation than to trigger language. When a takeover deal that blows through the latter is hammered out overnight there is not a lot that you can do.
The market has long muttered about the death spiral effects in the CoCo market – whereby debt investors that are converted into equity on breach of a trigger are forced to immediately sell that equity and accelerate the demise of the institution. This is the first time that there has been any principal or coupon loss in the AT1 market and there was no sign of a death spiral. 
However, that was only because the bondholders didn’t have time to be converted into anything that could be sold: Popular’s AT1s and tier 2s were converted into shares that were immediately written down to zero. The market has spent too much time fretting over the impact of CET1 triggers that – if Popular is a textbook case – will never get to be hit.
The wipeout of Popular’s equity and AT1 investors is unquestionably the BRRD doing what it was designed to do and investors will not have been surprised by their treatment.
Pimco is understood to have been holding €279 million of the AT1s at the end of March – the giant US money manager also held more than €100 million of Novo Banco senior bonds that were bailed in under its disputed resolution.
However, the fallout from Popular’s demise might be more keenly felt by its tier-2 investors.
The deal with Santander means that Popular’s tier-2 investors were dealt with in exactly the same way as its AT1 investors – they were written down to zero – although the tier-2 investors got the €1. This in effect removes any distinction between the two asset classes in resolution.
The temptation to wipe out anything that you can in resolution is understandable as it makes the bank more attractive to a potential buyer, but if this will always happen then why buy tier 2? You are getting paid more for the same risk with AT1.
All eyes are now on the tier-2 bonds of two other Spanish lenders: Liberbank, whose 6.875% €300 million tier-2 bonds traded below 81c on Friday – while a short selling ban was placed on its shares on Monday.
Another lender, Cajamar, also has a €300 million 7.75% tier-2 bond outstanding that is now trading below 90c. Popular’s tier 2s were trading at between 70c and 80c immediately before they became worthless.
What is key here is that Popular had yet to issue any new style bail-inable senior debt. If banks want investors to buy their tier-2 debt then in future, a resolution such as this might need to involve – very different – haircuts for both tier 2 and senior debt rather than oblivion for one and protection for the other; Popular’s senior bonds were up from 90c to 104c after the deal.
Although this takeover is a neat and straightforward demonstration of what investors can expect under BRRD, there needs to be a closer examination of the loss-absorbing hierarchy of tier 1 and tier 2 in the process. If they are the same then you don’t have a repayment waterfall – you have a lake." - source Euromoney
Of course, we would argue that no matter how much liquidity via LTROs the ECB has injected, liquidity doesn't amount to solvency hence the importance lessons in dealing swiftly with nonperforming loans as done by the Fed, while the ECB has been following a path more akin to Japan (here comes our "japanification" analogy). But, from a technical perspective on the subject discussed in details by Euromoney, there is a technical aspect that needs to be discussed namely protection offered only to some subordinated bondholders via the CDS market:
"Typically, in subordinated CDS single names, the bond reference is a Lower Tier 2 bond (LT2), and not Tier 1 (T1) bonds or Upper Tier 2 bonds (UT2), as coupon payments can be deferred in these structures. For Tier 1 bonds and UT2, missing a coupon does not constitute a credit event, therefore they cannot be used as a reference for a single name financial subordinate CDS, so no CDS on these bonds."
A typical LT2 Structure is as follows:
10 years, non-callable for 5 years (10-NC5)
-Final maturity is hard, meaning no get out clauses.
-No coupon deferral. A coupon deferral would constitute a credit event and therefore trigger a credit event and the relating CDS.
-Ratings: 1 notch below senior debt (Moody's / S&P).
-More standardised structure than Tier 1 or UT2. Given regulatory capital treatment decreases as it moves towards maturity (as it gets closer to 5 years to maturity) so many deals structured have had callable features, meaning the issuer can decide to call the bond.

As a reminder the Dutch bank SNS resolution in 2013 meant that Tier 1 bonds and LT2 losses equated to 100%. At the time the SNS intervention clearly pushed the envelope on how national authorities and now the ECB being the European regulator are willing (and able) to go in the resolution of a failing financial institution. The downside recovery for LT2 (using Irish precedents) was generally assumed to be 20%. This should have been understood to be 0% back in 2013, but yet again investors have been blind. Why the change in recovery rate from 20% to 0% matters in the CDS space?

If the recovery rate for LT2 subordinated bonds is zero (again for Banco Popular and in similar fashion to the SNS case), the significance for the European subordinated CDS market is not neutral given the assumed recovery rate factored in to calculate the value of the CDS spread is assumed to be 20% for single name subordinated CDS and 40% for senior financial CDS. Subordinated debt should in essence trade much wider to Senior! Particularly if liquidity, not capital, matters and if the fate of subordinated bondholders is anything but predictable.

Here is your "known unknown". Yet in Banco Popular's case the short end of the CDS curve was already inverted before the takeover for €1 was announced so while in the above article we read that there was no sign of a death spiral, this is not entirely correct. If indeed AT1 are known unknowns, and are American options like short gamma trades, who really cares about the capital threshold trigger aka the strike price seriously? We don't and continue to dislike these bonds but hey, some might enjoy nonlinearity, we are not big fans.

For those who have been following us, we have been pretty vocal on the evident lack of resolution of the nonperforming loan issues (NPLs) plaguing the Italian banking sector. On a side note we did at the end of last month a podcast on the Futures Radio Show in which we discussed Italy, being for us the biggest risk in Europe.
The on-going issues plaguing the oldest bank in the world namely BMPS aka Banca Monte dei Paschi di Siena can be seen in the CDS market and the NPLs have yet to be meaningfully tackled as indicated in the recent blog post from DataGrapple from the 19th of June entitled "A Tricky Task Just Got More Difficult":
"On Friday, it emerged that Fortress Investment Group and Elliott Capital Management had dropped out of talks to buy bad loans from MONTE ( Banca Monte dei Paschi ) complicating the rescue plan for the lender backed by the Italian government. They were the only international bidders for the riskier tranches of MONTE’s bad loan securitization. That leaves Atlante, the fund set up to help the struggling Italian banking sector, as the only potential buyer and jeopardizes the asset sale, which is a key part of the plan to restructure the bank with a capital injection from the state, after MONTE failed to shore up capital privately. Ultimately, it could also make similar rescue plans for two other northern Italian lenders, Veneto Banca Spa and Popolare Vicenza Spa, much more difficult to pull off. Surprisingly, if MONTE’s 5-year risk premium was marked aggressively wider - insuring senior debt now costs 330bps per year, while insuring subordinated costs 73.5% upfront -, it did not feed through the whole Italian banking sector and most names were actually unchanged to a tad tighter." -source DataGrapple

And guess what 73.5% upfront Subordinated CDS still seems pretty cheap if you ascertain the fact that your recovery value on the subordinated bonds will probably be a big fat zero...Just a thought. So while you might want to continue playing the Circus Maximus credit show, when it comes to favoring Senior Financials over Subordinated bonds, use your credit skills wisely.

If you indeed love low volatility, stick to investment grade credit because as the party continues flow wise in that space, the feeble crowd aka retail is now joining the party with both hands according to Bank of America Merrill Lynch Credit Market Strategist note from the 16th of June entitled "ECB+BOJ&Fed, redux":
"Retail taking over from foreigners
US high grade corporate bonds have returned 4.2% this year driven by 7.3% in the backend. With the timing of the decline in interest rates, not surprisingly about two-thirds of this performance has accrued in 2Q. These stellar returns are now attracting retail money to HG bond funds and ETFs in the usual way (chasing performance). As we have often highlighted, in the first part of the year we had record inflows without supporting preceding performance necessary to attract retail inflows – in fact HG lost almost 3% in 4Q 2016 (Figure 2).

Our view remains that foreign investors were responsible for the big acceleration in HG bond fund/ETF inflows to begin the year, as the timing coincided with a big decline in the cost of dollar hedging (Figure 3).

Furthermore these inflows accelerated more at the end of the Chinese New Year." - source Bank of America Merrill Lynch
It looks to us that indeed Circus Maximus is getting crowded, but for now everyone in the credit "karaoke" is singing "carry on" so you probably got to keep dancing...

In our final charts below and given our take in the podcast, we continue to view Italy as the biggest European risk even though elections have so far been postponed. Growth is not meaningful enough, we think to alleviate the slowing but still growing very large nonperforming loans problem on Italian banks' balance sheets.

  • Final charts - Italy? It's getting complicated
While we won't go through the debt trajectory of Italy and the dismal growth experienced in recent years, for our final charts we would like to point to charts from a recent presentation done by French broker Exane on the 12th of June. They point out to IMF work showing that the NPL ratio drops significantly when GDP growth reaches 1.2% (and particularly when this trend is sustained for a few years). Their charts below also display the relationship with the 2-10 Italian yield curve. This indicates that Italian banks are very sensitive to a surge in yields in the short-end which makes very interesting in the light of the willingness in tapering as of late from the ECB:
Italy: GDP Growth versus NPL ratio

Italy: NPL ratio versus 2-10 yield curve
- source Exane
According to them, YoY growth for Italy is likely to enable a significant decline in NPLs. Yet they indicate that while there is hope for a relaunch of the European project, the situation of the banking sector remains a concern and is still plaguing credit conditions for the corporate sector. They conclude their slide in their French presentation by saying that for the sustainability of the Italian debt, it's getting complicated. We could not agree more: "Troppo complicato!"

"Too much sanity may be madness and the maddest of all, to see life as it is and not as it should be." - Miguel de Cervantes

Stay tuned!

No comments:

Post a Comment

View My Stats