"Always speak the truth, think before you speak, and write it down afterwards." - Lewis Carroll
"I don't know what you mean by 'glory,' " Alice said. Humpty Dumpty smiled contemptuously. "Of course you don't—till I tell you. I meant 'there's a nice knock-down argument for you!' " "But 'glory' doesn't mean 'a nice knock-down argument'," Alice objected. "When I use a word," Humpty Dumpty said, in rather a scornful tone, "it means just what I choose it to mean—neither more nor less." "The question is," said Alice, "whether you can make words mean so many different things." "The question is," said Humpty Dumpty, "which is to be master—that's all." - source, Lewis Carroll's Through the Looking-Glass (1872)
"I don't know what you mean by 'incoming data consistent with economic growth picking up in the second quarter, labor market conditions continuing to strengthen, and inflation making progress toward the Committee’s 2 percent objective' " Alice said.
Janet Yellen smiled contemptuously. "Of course you don't—till I tell you. I meant 'there's a nice knock-down argument for you!' "
"But 'incoming data consistent with economic growth picking up' doesn't mean 'a nice knock-down argument'," Alice objected.
"When I use a word," Janet Yellen said, in rather a scornful tone, "it means just what I choose it to mean—neither more nor less."
- Macro and Credit - Regardless of QE, ZIRP and now NIRP, the ECB is failing in spurring credit growth
- Macro and Credit - Lending growth and the credit cycle and why the Fed is in a bind
- Final chart: US bond market - The warning sign from the long end
- Macro and Credit - Regardless of QE, ZIRP and now NIRP, the ECB is failing in spurring credit growth
"As we have stated on numerous occasions, when it comes to European banks, you are better off sticking to credit (for now) than with equities given the amount of "deleveraging" that still needs to happen in Europe." - source Macronomics, February 2015We also quoted at the time Berenberg's take on the "Japanification" of Europe "Through the Looking-Glass" of its banking sector:
"In short, until there is true clarity in the value of European banks’ assets, then the value of the equity is highly uncertain, making European banks uninvestable. In our view, what Europe needs to do, and what happened in Japan, is to force banks to dispose of a material proportion of their non-performing loans." - source Berenberg as per Macronomics note from February 2015
"No matter how low interest rates on corporate loans have fallen and has been much vaunted by the ECB and many pundits as a "great success", lack of "Aggregate Demand" (AD) and loans flowing to SMEs thanks to insufficient demand, this will not, rest assured, resolve the on-going woes, which have been much increased by the recent implementation of Negative Interest Rate Policy (NIRP), of the Italian banking sector. Either you remove the NPLs from the bloated Italian Banks' balance sheets and the ECB monetizes the lot, or they don't. Anything in between is an exercice of dubious intellectual utility, hence our chosen title. Also as per our analogy, we wonder if, at some point, in similar fashion to Ayn Rand's book, investors will not go on "strike" when it comes to helping out the Italian banking sector as a whole." - source Macronomics, April 2016
"From bel canto to bank analysis
It would be fair, we think, to typify the first trimester of this year for the Italian banks as a torrid one. It’s been tough for all global financials but Italian banks have particularly suffered. YTD Unicredit’s stock has fallen -43%, Monte dei Paschi -53%, Banco Popolare -65% and Intesa -25%. There are a number of reasons for this underperformance but at the base we see the systemic asset quality, credibility and capital problems in Italy as the drivers of investor concerns.
From bel canto to bank bond analysis: we measure the empirical riskiness of bonds by looking at the standard deviation of daily excess returns. We are not surprised that the Top 5 riskiest bonds in Europe YTD are all Italian banks, specifically Monte dei Paschi Tier 2, Veneto Banca T2 and Unicredit USD AT1. Excluding DB, Italian bank bonds occupy all 9 places of the Top 10 most volatile bonds. Following the 1Q reporting season and the recent events in the Italian banking sector, perhaps it’s a good time to reassess whether the market’s assessment is really a robust one.
We’re mindful that Italian financials are a significant part of the HY Index. There is €27.5bn of Italian paper in the HY Fins Index which is 45% of the total. It’s one thing to be negatively positioned in these when they are in decline but what if there is a turnaround in the assessment of the fortunes of these banks?" - source Bank of America Merrill LynchWe do not think there will be a turnaround through the looking glass of their "better earnings" thanks to "lower provisioning levels". On this subject we read with interest Bank of America Merrill Lynch's take from the same note:
In 2014, some banks briefly circulated the idea that they were more conservative in their classification of NPLs than other European peers as the reason for the high quantum – this got quashed when the AQR clearly demonstrated the opposite. There have also variously been tax reasons and legacy lending issues, amongst others. We think the reasons may encompass many of these, but at root the answer is probably simpler: Italian banks did not seem to be very good at underwriting, in our view. We think high levels of NPLs are, in some respects, a management choice. To be fair, up to 2016, no one seemed to care and we could get little traction with investors when we talked of asset risks in Italy. This indifference possibly explains the relatively high level of complacency on the part of the banks on the NPL front. Remember this is a jurisdiction where a bank with a 16% NPA ratio is considered best in class.
In contrast to what we might describe as the pusillanimity of the banks in face of this significant challenge, we think the Italian Government has moved relatively proactively to try and address systemic concerns. It orchestrated a fund to help recapitalize some of the banks and avoid failed IPOs which would likely have caused further systemic issues, we think. Importantly, there have also been a series of measures which have been designed to reform insolvency practice in Italy and address tax issues around provisioning. There has even been an attempt to create a kind of bad bank, or more precisely to kick start a more liquid market for NPLs, through the provision of a Government-guarantee to NPL securitisations. We consider the efficacy of these operations in turn.
Initially, we assessed the Atlante fund as a positive development for the Italian banks as we saw it as an attempt to break the cycle of bad news around the banks. We were slightly disappointed that the fund raised only €4.25bn compared to the €5-6bn that was originally mooted. Originally designed to ensure the success of the BP Vicenza IPO, and in our view, also help rescue Unicredit from its ill-fated decision to be sole underwriter for said transaction, the IPO of BP Vicenza has now passed, with Atlante having to take up the entirety of the €1.5bn in stock that was offered, because outside investors did not take up any allocation in sufficient quantity. According to Borsa Italiana, 10 insitutions offered to buy 5.1% of the shares which was insufficient free float. Atlante now has €2.75bn in resources left. The IPO of Veneto Banca is upcoming (pre-marketing was to start at the end of this week) and has been widely considered e.g. in the Italian press to have a greater chance of success when compared with Vicenza, not least because the bank is trying to place a smaller amount (€1bn). Market conditions remains hazardous though, we think, as recent comments by CONSOB underline. We have seen some discussion in the Italian press that the fund could be scaled up but we haven’t seen anything concrete on this – Bloomberg reported on Wednesday that the Italian Finance Minister was suggesting in an interview that it could be enlarged. Atlante is in any case a closed fund now but 66% of holders could vote to expand it, so we can’t exclude that the fund will grow, especially if it needs to.
In spite of being ostensibly private to avoid the state aid rules in Europe, Atlante seems to us to be serving a specifically public policy role, on our reading of its presentation, ‘by eliminating excessive supply with respect to the demand for shares’ though it is supposed to have the ‘interest of investors as its sole objective’. It seems to be an unusual set-up even by European standards.
Atlante has already served one of its primary purposes, we think, in subscribing to the Vicenza increase and sub-underwriting the IPO thereby reducing the risk to Unicredit ofbeing left with a significant overhang of shares. 70% of the funds resources are designed to be available for the support of capital raises, with the balance, or just under €1.3bn currently, for the purchase of NPLs. On NPLs, the idea is not that Atlante replaces the NPL market, but that it promotes ‘the creation and development of an efficient market of distressed assets in Italy’. In other words, Atlante may facilitate the sales of NPLs e.g. by buying mezzanine or equity tranches of NPL securitisations – especially those that take advantage of the Government guarantee (the so-called GACS) for the senior/investment grade tranche of the structure. The first bad loan securitization, with a GACS guarantee, is already happening in Italy with a €500m 3 year transaction for BP Bari. However, as the chart shows, loan sales in Italy have been relatively few. The expectation is that Atlante, plus GACS (plus the legal reforms, below) might provide the conditions to accelerate the creation of a more active market for bank bad loans.
We keep an open mind on the Atlante structure and its benefits – the market has been skeptical hitherto. We recall Fitch’s warnings that Atlante potentially drags healthy banks down in their rescuing less healthy institutions. But we still believe perceptions can quickly change as e.g. NPL transactions materialize using GACS, whether or not Atlante participates. Currently, in our view market sentiment around the Italian banks is still overwhelmingly bearish, especially after the last reporting season which was, in summary, rather underwhelming, in our view. We think it’s rather soon to assess the potential benefit from Atlante. We will need to see successful transactions concluded, however. Positive conclusions for the upcoming IPOs of other Italian banks would also be helpful, we think." - source Bank of America Merrill Lynch
"Perhaps the banks are anticipating the benefits of the patto marciano and so believe there is less need to keep provisions high (we understand that many European banks are also eager to anticipate the -0.40% rate at which they may be able to borrow from the ECB, even though that rate strictly speaking can’t be calculated ex ante). Lower provisions was a driver of many of the beats to consensus expectations in Italy, though, it seems. " - source Bank of America Merrill Lynch
"Net interest income is nothing more complicated than the revenue spread on the balance sheet. It depends on three pretty straightforward variables: the size of the balance sheet, the yield on assets and the cost of liabilities. It is the first two of these variables that have been under consistent, sustained pressure across the sector. Liabilities have not got cheap enough, quickly enough to compensate.We hate sounding like a broken record but, no credit, no loan growth, no loan growth, no economic growth and no reduction of budget deficits and NPLs.
Across the sector, NII contributes c. 60% of the revenue base, and so is the major part of the dynamic. Non-interest income is more volatile, and spread between a multitude of different business lines. Essentially, fee income is a flow on the franchise value of the bank: the branches, the product range, the staff, etc. The outlook is less clear, and the drivers can change quickly. Strong markets would tend to be the biggest single force." - source Société Générale
"Revenue wipe-out European banks are suffering. Every bank we cover has reported a year-on-year drop in Q1 16 revenue. A heady mix of zero rates, tough markets, weak CIB and stagnant lending volumes is taking a heavy toll. This matters for the sector. While nearterm earnings have been underpinned by better credit quality, this is not a theme that can continue forever. Sector earnings are stuck in a downgrade cycle, and pressure on the revenue line is at the heart of this." - source Société GénéraleAs we have argued before QE will not be sufficient enough on its own in Europe to offset the lack of Aggregate Demand (AD) we think. Although the ECB has been trumping the convergence in Europe in interest rates charged on new corporate loans, as we stated before, the "fun" is uphill, in the bond market, not downhill, in the "real economy". If there is one chart displaying the failure of the ECB we believe it is the below chart from the same Société Générale note that illustrates the ECB's failure in spurring credit growth:
"The weak revenue environment helps to explain a major anomaly with the Q1 16 results season: the sector generally ‘beat’ consensus on earnings, but consensus EPS downgrades have continued unabated." - source Société Générale
- Macro and Credit - Lending growth and the credit cycle and why the Fed is in a bind
"The above grapple depicts the weekly change of the risk premia of the constituents of the US Corporate cluster identified by DataGrapple. Retailers are easy to spot. In an otherwise resilient market as shown by the greenish shades of most boxes, all of them are red (and some bright red). Most of them reported first quarter numbers, and all of them managed to disappoint. Today, JCP (J C Penney Company, Inc) posted revenues that trailed analysts’ estimates and joined fellow discount-oriented KSS (Kohl’s Corporation) which missed estimates yesterday. Higher end rivals did not fare any better. M (Macy’s, Inc) reported lacklustre results and lowered EPS guidance for the year by 57cts (to $3.15-$3.40 from $3.80-$3.90), while JWN (Nordstrom, Inc) also added to evidence that the department store industry is mired in a deep slump when it cut its annual earning forecast. Shoppers across the income spectrum appear to pull back on purchases of apparel and other discretionary goods." - source Datagrapple, 13th of May 2016For those with a "short bias mentality", please note that CMBX.6 has the highest percentage of retail exposure. CMBX.6 has considerably more exposure to B/C quality malls...just saying.
And when we say the Fed is in a bind because of this relationship between "retail" and CRE we are not the only one meaning it. For instance Bank of America Merrill Lynch make the following interesting points in their Weekly Securitization note:
"While limited issuance might otherwise provide a powerful tailwind for spreads, myriad uncertainties remain on the horizon lead us to adopt a more cautious near term view.
First, as we mentioned above, is the increasing probability that the Fed may raise rates over the next two months. Although a rate hike in and of itself won’t be overwhelmingly negative, it is likely, if not probable, that any near-term hike will be negatively viewed by many investors and could possibly be interpreted to signify that the Fed will be overly zealous as they seek to normalize interest rates against what they believe is an improving economic environment. Second, to the extent higher rates exacerbate the recent credit tightening, it is reasonable to fear that CRE price growth, which is already showing signs of slowing, could be exacerbated to the downside. While the recent Federal Reserve Senior Loan Officers Survey showed that bank lending standards have tightened since the end of 2015, rising recent conduit debt yields (Chart 52) and stabilizing Moody’s stressed LTV (Chart 53) metrics indicate the same is true within CMBS. Risk retention, which dealers are actively planning for, is likely to tighten lending standards further.
The combination of better underwriting, subdued new issuance activity and shrinking dealer balance sheets (Chart 54), which make it difficult for investors to add bonds in size away from the new issue market, have likely contributed to the cash bond spread rally.
Recent CMBX spread/price movements, however, tell a different story and may provide insight into the macro-related nervousness investors are feeling.
On the week, CMBX prices, especially for tranches at or near the bottom of the capital structure, fell by as much as three points (Chart 55) and have fallen by as much as five points since the beginning of the month (Chart 56).
Again, to the extent that oil prices remain rangebound or increase and no major economic disruptions occur over the next month, we anticipate that CMBX spreads will trade directionally with broader markets." - source Bank of America Merrill Lynch
"In corporate (non-household) lending markets we believe the proverbial plot has thickened considerably. The two key lending segments are commercial and industrial (C&I) and commercial real estate loans (CRE). For C&I lending, banks comprise less than 20% of total lending; the bond markets are the new marginal provider of liquidity (Figure 2).
Our non-bank proxy, incorporating bond and trade finance measures of credit conditions, has been indicating more tightening than bank proxies (i.e., the Fed's SLOOS survey) for several quarters. And our proxy is still suggesting further tightening ahead, primarily given that new issuance in US high yield and leveraged loan markets remains sluggish despite recent outperformance (US HY and institutional LL issuance are down 47% and 33%, respectively, in 2016; CLO issuance is off 68% YTD). That said, the rate of tightening has slowed, as HY issuance and trade-credit standards improved in April (Figure 3).
In short, the trend in lending conditions is still tighter – but there has been some easing in funding conditions.
Conversely, lending conditions in commercial real estate have deteriorated. Banks comprise about 55% of total lending, so the Fed's SLOS survey is more telling. And, in the last quarter, banks tightened lending conditions in CRE, specifically in multifamily and construction/land development vs nonfarm non-residential (36% and 24% net tightening versus 12%, respectively, Figure 4; CMBS issuance is also down 38% through April).
Why are banks tightening?
The most important factor cited was not CRE fundamentals, cap rates, competition nor funding, but 'other' factors – and by a large margin. What is our interpretation? Increased regulation was the underlying cause. We have previously flagged concerns around the pervasive easing of underwriting standards for corporate lending broadly 5 . Since 2007 FDIC-insured banks have increased nonfarm, nonresidential loans (ex-owner occupied) approximately 82% to $730bn; multifamily loans rose by roughly 142% to $344bn6 . In December, the OCC released its Statement on Prudent Risk Management for CRE Lending in response to "significant growth in CRE lending, increased competitive pressures, historically low cap rates and rising property values". This guidance was originally issued back in 2006, requiring banks with higher CRE concentrations to tighten risk and managerial controls.
We estimate approximately 8% and 16% of FDIC-insured banks by count and CRE debt outstanding, respectively, were above at least one of the concentration levels at YE 2015 (i.e., >100% CLD vs total capital, or >300% CRE vs total capital and >50% growth prior 3 years). For comparison, in 2006 about 31% and 40% of banks, respectively, exceeded one of the thresholds. Last month an American Bankers Association survey suggested similar, but modestly higher, figures in terms of bank concentration levels among respondents. Further, 40% indicated they expected a measurable reduction in credit availability to certain CRE sectors, while another 25% suggested a measurable reduction in credit availability across all sectors from the guidance.
Why is this important? First, regulation can have a significant impact on the supply of credit, with US leveraged loans a recent case in point8. Back in 2006, the CRE guidance for banks also coincided with a significant tightening in CRE lending conditions. Second, changes in lending conditions for C&I and CRE loans have been quite highly correlated in the past (see Figure 4 above ).
One study finds banks that were above CRE guidance concentrations not only slowed CRE loan growth, but also tended to reduce C&I loan growth. Simply put, macroprudential regulation can have a more broad-based and severe effect on commercial bank lending. One could argue that constraints on lending as the credit cycle matures may guard against excess losses; conversely, an exogenous shock to the supply of credit at a time of lacklustre global growth and upcoming risk events could exacerbate
And the linkages between CRE and C&I lending are multi-faceted. The lenders – in particular regional and community banks – tend to have higher concentrations of commercial and C&I loans, and some smaller business loans are collateralized by commercial property. Second, borrowers can also be concentrated in certain sectors across C&I and CRE. According to the Fed's Shared National Credit Review the largest industries in the leveraged C&I portfolio included Healthcare (14%), Media/Telecom (13%), Finance/Insurance (12%), Materials/Commodities (6%) and Retail (5%)10. In CRE portfolios, larger concentrations lie in multifamily (28%), office (18%), retail (16%), industrial (12%), hospitality (8%) and healthcare (7%) according to the ABA. Much of the CRE growth has occurred in coastal cities. This implies, while not directly comparable, C&I and CRE depend more heavily on similar industries – media/telecom, finance (non-bank), retail and healthcare.
The conclusion is that the plot is thickening with respect to the signals of corporate lending conditions, and we believe clients should increasingly view them in a holistic framework. While there are signs of moderation in the rate of tightening for C&I loans, we are already seeing rising delinquencies and defaults weigh on corporate profits and growth (stages 3 and 4 of our rudimentary credit cycle model outlined earlier). But for CRE loans we think there is greater cause for concern as potential harbingers of greater credit constriction emerge.
Importantly, we have yet to see much rise in delinquency and default rates – although examiners noted rising concerns over CRE credit risk for 75% of banks and expected credit risk to rise in 50% of all commercial loans at year-end. The ESRB provides a simple but useful framework for the CRE cycle (Figure 5); while the CRE cycle appears less advanced than the C&I cycle, we believe investors should closely watch for evidence of potential negative effects on credit availability in CRE and potential broader spillover to C&I lending.
In particular, investors should closely watch the media/telecom, finance (non-bank), retail and healthcare industries given the linkages across both markets. In terms of investments, our view of recent trends in corporate lending conditions does not support 'reach for yield' or down-in-quality trades. In corporate credit, we continue to prefer longer duration US high grade bonds versus high yield." - source UBS
- Final chart: US bond market - The warning sign from the long end
"Bond market vs. Fed speak: look at the long end, not EDs
Recent Fed speak and a slew of important speakers lined up to talk before the June meeting (Dudley, Yellen, Fischer) has shifted attention back to the front end of the rates curve. Eurodollar bears that were forced into hibernation since March have woken up, revisiting similar arguments of dots vs. the Fed, rates vs. US data surprises etc. To us, there is one worrying sign in the reaction of the bond market to the better data and hawkish Fed speak unlike 2013: instead of the optimistic signal the yield curve sent during the taper tantrum, long end rates now suggests a re-ignition of the policy mistake trade. Raising June/July probabilities is a small victory that is coming at the cost of dealing with higher probability of inverted yield curves 2-years forward, in our view.
This is not 2013: reigniting the policy mistake trade
A critical difference between 2013 and today is the inability of Fed optimism to filter through into higher long end yields. Chart 1 plots OIS forwards in mid and end 2013 (pre and post taper tantrum).
The combination of better data and shift in messaging from the Fed in 2013 was viewed to be a sign of an optimistic longer term growth picture – the resulting rise in yields was a healthy combination of 1) higher terminal rates 2) higher term premiums 3) and thereby a license for the Fed to hike sooner and faster than was originally thought. Recent communication however struggled in this regard: hawkish Fed talk and better US data has only helped 1) strengthen the dollar and weaken inflation expectations 2) lower terminal rates priced in 3) increase hike probabilities for the near meetings without shifting medium term expectations.
Lack of credibility constrains effectiveness
The policy mistake angle assigned to the Fed is visible in more areas than the yield curve. Chart 3 plots long term real rates in five major countries vs. the miss to the respective central bank’s inflation target. The market continues to believe that the Fed will deliver real rates that are far too high and miss on its long term inflation target by at least 50bp. To us, this inability of the Fed to improve longer term expectations priced in to the market tows the line of igniting a bigger concern: getting dangerously close to the market pricing in inverted yield curves, 2 to 3 years forward." - source Bank of America Merrill Lynch
"If you stop being scared, that's when entropy sets in, and you may as well go home." - Tamsin Greig, English actress