Macro and Credit - Under the Volcano

"A democracy is a volcano which conceals the fiery materials of its own destruction. These will produce an eruption and carry desolation in their way." -  Fisher Ames, American statesman

Looking at the large wobbles experienced in financial markets this week, leaving many pundits wondering if we had attained "The Amstrong limit" and trying to figure out if it was the start of something much larger at play, when it came to selecting this week's title analogy, we decided to steer back towards literature this time around. "Under the Volcano" is a famous1947 novel by English writer Malcom Lowry. The novel tells the story of Geoffrey Firmin, an alcoholic British consul in the small Mexican town of Quauhnahuac, on the Day of the Dead, 2 November 1938. When it comes to QE and alcoholism, we reminded ourselves our September conversation "The Korsakoff syndrome" being an amnestic disorder caused by thiamine deficiency (Vitamin B) associated with prolonged ingestion of alcohol (or QE...some might argue). But what is of interest to us in our chosen analogy, is that this great novel of the 20th century has 12 chapters and the following 11 chapters beside the first introductory chapter happen in a single day. In similar fashion one could posit that the credit clock has 12 hours. In his novel Lowry alludes to Goethe's Faust as well as references to Charles Baudelaire's les Fleurs du Mal. We also used similar reference to Baudelaire's Les Fleurs du Mal back in December 2011 in our conversation "The Generous Gambler" and in 2014 in our conversation "Sympathy for the Devil":

"The greatest trick European central bankers ever pulled was to convince the world that default risk didn't exist" - Macronomics.

Throughout Malcom Lowry's novel the number 7 appears, in similar fashion, we are seeing many signs reminiscent in the current credit cycle of the year 2007 or even with 1987 (the DJIA topped in '87 at 2700) given today we have both dividends and buybacks paid out in excess of operating cash flow.  Both are being funded with debt accumulation exactly as it was the case in the year 2007. Also, one may argue that somewhat, European bond investors made a "Faustian bargain" with Mario Draghi aka our "generous gambler" but we ramble again. 

In this week's conversation, we would like to look at once again where we stand in the credit cycle and ask ourselves how long until we see it definitely running.

Synopsis:

• Macro and Credit -  What's under the "credit" volcano?
• Final chart - Large standard deviation moves, the "market" volcano is becoming more "active"

• Macro and Credit -  What's under the "credit" volcano?

As we pointed out in our most recent conversation, the latest quarterly Fed Senior Loan Officers Opinion Survey (SLOOs) continues to indicate overall support for credit markets yet the market feels more and more complacent à la 2007 we think:

- graph source Macrobond

The most predictive variable for default rates remains credit availability and if credit availability in US dollar terms vanishes, it could portend surging defaults down the line. The Fed quarterly SLOO survey reflects the ability of medium sized enterprises (annual sales greater than $50mn) to get funding from regional banks. Since HY issuers fit this criterion, this survey is also well correlated with their ability to tap the bank lending market. The SLOOS report does a much better job of estimating defaults when they are being driven by a systemic factor, such as a turn in business cycle or an all-encompassing macro event. Credit always leads equities in our book.

Credit investors look at the CDS roll. The most recent roll into the new contracts was in September, the new “on-the-run” benchmark series. The current steepness of CDS curves is a headwind for anyone “bearish” on credit and wanting to express it through CDS products. Too costly right now:

- Graph source Edward Casey - Bloomberg

Yet, there is no doubt rising “dispersion” which in effect means that credit investors are becoming more discerning when it comes to their selection process of various issuers’ profile. This we think is another sign of a late credit cycle.

To illustrate further the deterioration in the credit cycle overall picture, one could look at European High Yield and in particular Consumers and Cyclicals as shown in the below chart by DataGrapple on the 9^th of October:

“It was a mixed session with BTPs, stocks and rates sending contradicting signals throughout the day. In credit, there was a constant theme though, as investors sold risk on higher beta auto and autopart related names. The sector has been heavy for a couple of days, a phenomenon that was pinned down to the upcoming EU environment ministers meeting to discuss emission caps, which is widely expected to result in a push for a more ambitious set of rules. It culminated this morning in a proper battering of TTMTIN (Jaguar Land Rover Automotive Plc) which saw its 5-year risk premium marked 45bps wider at the open. This aggressive move followed the September sales numbers reported by the company. The year-on-year decline amounts to 12.3%, as strong sales for new models were offset by weakness in China where demand dropped 46.2% on the back of import duty changes and continued trade tensions. This came exactly a month after Ralf Speth, the CEO, warned that a hard Brexit would cost the company £1.2Bln a year and would wipe out its profits. The company also confirmed the two-week temporary closure of its Solihull factory, which employs almost a quarter of the group’s workforce in the UK. Some profit taking on short risk positions eventually emerged at the end of the session and limited the widening of TTMTN’s 5-year risk premium to 28bps at 485bps, but the negative trend of the past nine months which is obvious on the above grapple shows no sign of abating and is in fact gathering momentum since the roll.” – source DataGrapple

With rising dispersion, and global trade deceleration and the effects of the trade war narrative, we are already seeing cyclicals underperforming. 

In similar fashion to 2007, when default rates are low, credit investors believe that stability is the norm, and start piling up on leverage or CLOs with lack of covenants such as Cov-lite loans as per the below chart from Bank of America Merrill Lynch, indicative of aggressive issuance:

- graph source Bank of America Merrill Lynch

What is of interest to us, regardless of the "liquidity" issues many pundits have been talking to about in relation to mutual funds and the strong growth in passive management through ETFs in recent years has been the rapid growth of the private debt market in this very long credit cycle.

On this subject we read with interest Bank of America Merrill Lynch High Yield Strategy note from the 12th of October entitled "The Next Credit Cycle - Scenarios for HY, Loans, and Private Debt":

"Private debt, the fastest growing segment in US credit

By its very nature, the private debt market is more difficult to analyze as most deals never get included in any widely followed indexes or make it into otherwise publicly reported portfolios. Even estimating the size of this market is a challenge, and we had to go about it backwards, by starting with known overall corporate debt stack and removing otherwise known and attributable pieces. We think, the market is somewhere between $400-$700bn in size, and it was the fastest growing segment of US credit, including bonds, bank and non-bank loans, over the past five years.

This report outline our understanding of structure, major investor types, growth, sector composition, leverage and covenant trends, key risks and mitigating factors of the private debt space. We find this asset to feature many hallmarks of a classic new hot market, which often results in unsustainable growth trajectories leading to eventual corrections, required to stabilize the market at longer-term sustainable levels. This report is also part of our broader take on US lending landscape that we published in collaboration with our banks and asset managers equity research and economics teams.

Loan covenants are the defining feature of this cycle

The syndicated leveraged loans continued to attract investor interest since the GFC, as their investment thesis (significant yield pickup coupled with no interest-rate sensitivity) remains appealing to many. As a result, the leveraged loan market has grown by 19% in the last two years, 44% in the last five, and doubled in the last ten. Strong demand forces asset managers have to compete for new deal allocations on both pricing and structures. Coupons are getting squeezed, leverage pushed up, and covenants dropped. And while tight pricing and elevated leverage are expected side-effects at this stage of the cycle, the degree of covenant deterioration has reached new levels in recent years, well beyond the outdated “cov-lite” label.

The next credit cycle: modeling potential credit losses

We bring all our knowledge of the three segments of leveraged finance – HY, loans, and private debt – in one place by running side-by-side credit loss models for three distinct scenarios: consensus middle-of-the-road, mild recession and a full-scale recession. Our interest primarily focuses on the last one as it helps us better understand the downside scenario and help us make more informed risk management decisions.

Key takeaways

We estimate the next credit cycle, when it happens, could bring credit losses to the extent of 2x of expected annual yield income in high yield and leveraged loans, and 1.3x in private debt. Investors could also experience temporary mark-to-market losses of up to 5x of their annual income. To put this downside risk into perspective, it would take a 325bps increase in yield to wipe out 2 years of yield income in HY, given the 4yr duration of this asset class. In other words, a 150bps increase in Treasury yields coupled with a 150bps widening in spreads is less damaging than a cyclical turn. While we do not believe the next credit cycle turn is imminent, this evidence improves our confidence in the existing positioning recommendation to begin underweighting lowest quality segments of the market in favor of higher quality segments." - source Bank of America Merrill Lynch

As per the above executive summary from their very interesting report, we do agree that the next credit cycle downturn is not imminent, yet we see rising M&A activity and rising dispersion as additional indicators of how late the credit cycle is. The summer drift for Emerging Markets has created some additional dispersion this time between EM High Yield and US High Yield as per the below chart from Bank of America Merrill Lynch:

- source Bank of America Merrill Lynch

Both rising oil prices and strong earnings have been very supportive of US High Yield so far.

But, returning to the subject of loose covenants aka Cov-lite loans we read with interest Bank of America Merrill Lynch's take:

"Loan covenants are an epitome of this cycle

The syndicated leveraged loans continued to attract investor interest in the last few years, as their investment thesis (significant yield pickup coupled with no interest-rate sensitivity) remains appealing to many. As a result, the leveraged loan market has grown by 19% in the last two years, 44% in the last five, and doubled in the last ten. Both syndicated loan and CLO issuance is hitting new records (Figure 8).

With strong demand for loans in recent years, asset managers have to compete for new deal allocations primarily on two scales: pricing and structures. Coupons are getting squeezed, leverage pushed up, and covenants dropped.

CLOs are in a particularly sensitive spot, where their ability to compete on pricing and leverage is limited as they have to make math work over the cost of funding and adhere to minimum rating constraints. As a result, some managers could be more inclined to compete by accepting looser investor protections for the same price and leverage.

A typical CLO ramp-up period includes a warehousing stage that could last for about six months. During this stage, new loans are being acquired as a collateral for the future CLO deal, and an equity investor in a warehouse facility carries the risk of market conditions moving against them during this ramp-up period. Therefore, equity investors are incentivized to close the ramp-up period as soon as possible.

This pressure is counterbalanced by established time windows on CLO warehousing facilities, which arguably allow managers some flexibility to bypass on deals they view as particularly unattractive. The choice of a CLO manager could depend on how quickly such manager is expected to complete this stage. There is a premium associated with well-established managers. In some cases, CLO manager and equity investor are the same entity.

Pressure to ramp up a portfolio for future CLO at the time of record CLO issuance volumes puts some managers in a position where they are forced to compete on the strength of investor protections for a given level of credit risk/coupon.

Retail funds also contribute to excess demand for loan product as they continue to see inflows. YTD 2018, loan funds are seeing a 10% inflow, compared to a 9% outflow from HY funds. Loan funds have higher tolerance towards lower quality (B2/B3) paper compared to CLOs.

While there are some natural limits on how aggressive they can be on pricing (via pricing floor on their liabilities), there are no immediate consequences to accepting looser covenants. During the period when default rates are low (like today), the impact from looser covenants through lower loan recoveries is negligible. This would likely change, once default rate increase in the next credit cycle.

Key risks in continued deterioration of investor protections

Strong competition in the new CLO/loan asset management space in the last few years led to deterioration in key investor protections, such as restricted payments, asset sales, EBITDA add-backs, and incremental debt capacity.

These covenants are critically important to recovery in case of default, as they are capable of directly affecting the pool of assets available to creditors in bankruptcy, and the extent of creditors’ ability to establish claim over it vs. unsecured and equity investors.

Loan recoveries, defined as post-default trading prices, averaged a relatively high 65% since 2007 as a function a large proportion of loans recover near-par in restructuring. Tight covenant packages helped them achieve stronger controls over asset pools in bankruptcy or other distressed resolution.

This may change in the next cycle as key covenants have been eroded in recent years. Assuming the proportion of near-par recoveries is cut in half, average first lien loan recovery rate could drop to low-50s%. For example, on a $1.1tn loan market size with 15% peak default rate and 15% undershoot in recovery (50% vs 65% historical) this is an equivalent of $25bn of capital being permanently wiped out purely as a function of poor covenants. The next credit cycle is likely to bring some very poor recovery prints in certain most aggressive capital structures. We discuss various scenarios for defaults/recoveries later in this report.

Covenants are particularly weak in the broadly syndicated loan market, where the competition for new deal allocations is high. The private/direct lending space has also seen some deterioration in investor protections, but to a lesser extent than what we have seen in the syndicated transactions.

Key mitigating factors

Not all loans lacking covenants carry the same risk of low recovery. “Cov-lite” is not a new term, as it was coined at the end of last credit cycle, in 2006-2007, when a growing number of new loans were coming in without a maintenance covenant. In such cases,  issuers were not required to adhere to leverage tests once the loan was issued. We have long found this particular covenant mundane, as the experience of multiple breached maintenance covenants has demonstrated that lenders generally reserved their right to declare technical default, and instead chose to provide waivers for a fee.

Post-Global Financial Crisis, the number of such “cov-lites” has grown to the vast majority of new leveraged loans by around 2013, so again, not a new development. In a sense, the “cov-lite” misnomer is an unfortunate label that muddies the waters of a real problem for the next credit cycle, which is epitomized by the new structures lacking other key covenants.

(Definitions of certain key covenants: Structural subordination: Protection against lien dilution or structural subordination for existing lenders. Restricted payments: Protection against cash leakage and value transfers that depletes value of associated collateral. Debt Incurrence: Protection against issuers leveraging up or paying other debt holders at the detriment of existing lenders. Investments: Protection against issuer taking on risky investments through carve-outs and builder baskets. Asset sales: Protection for lenders to enable them to benefit from asset sale proceeds and excess cash flows.) - Source Bank of America Merrill Lynch

As far as aggressive indicators are concerned we have yet to see an equivalent surge into LBOs we did in the previous credit cycle as a percentage of market size as per the below chart from Bank of America Merrill Lynch:

- source Bank of America Merrill Lynch

Given the deterioration in credit quality overall, as we have stated in numerous of our previous conversations we expect lower recovery values during the next downturn.

On the subject of what is "Under the Volcano" credit wise and what could possibly happen in terms of credit losses during the next downturn, in their report Bank of America Merrill Lynch does an interesting analysis:

"The next credit cycle: sensitivity analysis

In this section, we take three major asset classes under our coverage: HY, syndicated loans and private debt, describe their current pricing in fundamentals, and run three scenarios for the future (Figure 10).

The first scenario is base-case, consensus, middle of the road: the current economic trajectory persists, the Fed delivers on its dot plot estimates, and credit losses stay relatively modest, even for the floating rate instruments.

The second scenario, is the stressed case, which resembles a full scale recessionary environment, with earnings dropping 30% and the Fed being forced to cut rates back to zero. This is a scenario we pay most attention to in an attempt to properly manage a risk budget in coming years. The third scenario (shown in greyed-out columns next to stressed, is designed to represent a modest recession with better outcomes. Think of an energy experience in 2014-2015, perhaps a touch heavier or lasting a few months longer.

Note that while we show HY and syndicated loan spaces in two separate columns, the reality of the situation is that these spaces are not mutually exclusive as some issuers are present in both markets. With this limitation in mind, we think of this attribution as being defined by issuers that are predominantly HY or predominantly loans. We believe that such representation, while imperfect, allows us to more properly model the capital structure behavior of these otherwise distinct asset classes.

 Scenario #1: +100bp move in LIBOR, “average” loss ratesThis section is the base-case for the next couple of years, implies the macro environment remains broadly supportive and the Fed achieves its longer-term dot plot forecast. We note the following dynamics in our analysis:

• The impact on issuer fundamentals here is visible in changing coupons to the extent they are floating, and interest coverage ratios (ICRs) change in response to coupons.
• ICRs get somewhat problematic in syndicated loans and private debt space, but they remain generally manageable and comfortably above 2x.
• Leverage here is assumed to be unchanged, even though one could reasonably expect both earnings and debt to grow, somewhat out of sync with each other, over the next few years in a scenario where the Fed is able to deliver four more rate hikes. We did not aim to make this exercise about our judgment on those two imperfectly synchronized growth rates, and decided to leave leverage assumption unchanged in pursuit of simplicity and clarity of more consequential arguments that follow.
• We think some moderate credit losses could come out of this scenario, but unlikely to mark a turn in credit cycle more broadly. Such incremental moderate credit losses are more likely to surface in the syndicated loan and private debt spaces, where capital structures are predominantly floating rate.
• Importantly, we do not view this scenario as being directly linked to the next substantial pickup in credit losses. This is not how the cycle ends.

Scenario #2: a full-scale recession

The key component of our sensitivity analysis is designed to define a full-scale recessionary experience.

• We assume earnings decline 30% (normal recessionary range 30-40%), Fed cuts rate down to zero and Libor bottoms out at 0.50%, leverage/ICR ratios respond accordingly as functions of unchanged debt levels, lower earnings and somewhat lower interest expenses, to the extent of their floating nature.
• Given these changes in issuer fundamentals, leverage would be likely to increase to 6.7xin HY, 8.6x in syndicated loans, and 7.5x in private debt.

• Under these prevailing leverage conditions, we argue the default rates could hit 10% in HY (normal recessionary range 8-12%), meaningfully higher level of 14% in loans, and 12% in private debt.

• The HY bond market has an established track record of peak default rates over three independent credit cycles, with a normal recessionary peak level of 8-12%. We thus argue for a middle-of-the-road type of default experience here in the next credit cycle.

• Such track record is materially less reliable in syndicated loans, where the 2001-2002 cycle arrived when the asset class was in its infancy, and the 2008-2009 was arguably softened by the extraordinary policy response  aimed specifically at banks and structured finance products, although not directly CLOs.

• Our argument for a 14% default rate rests on our understanding of substantial growth rates that were witnessed here in recent years, coupled with the higher leverage measures relative to other related asset classes. Leverage in the syndicate loan market could hit 8.6x under a moderate assumption of a 30% drop in EBITDAs.

• Private debt space has no meaningful track record in previous credit cycles as the asset class has grown to its present size only in the past few years, although its early origins are traceable to the previous decade. We thus rely our 12% default rate assumption here primarily on its leverage measures, which are assumed to be (but not always directly observable) around 5x-5.5x, in between HY and syndicated loans.

• We also assume recovery rates of 35% in HY, 60% in loans and private debt. Recovery rates here are defined as trading prices shortly after the event of default. This measure differs from ultimate recovery, which is the payout on the other side of a restructuring process.

• Syndicate loan recoveries are penalized as a function of three factors: poor investor protections/covenants and poor tangible asset coverage in sectors most exposed to syndicated loans (technology, services, and retail). We do give the loan market a benefit for the fact that its structure is now materially less exposed to mark-to-market instruments, thus limiting the extent of fire sales that took place in 2008-2009.

• A 60% recovery assumption in private debt, is a very rough estimate, given absence of verifiable historical track records and extremely low liquidity. Paradoxically, the latter could be viewed as a benefit, as absence of any practical ability to trade out of a position could arguably prevent many private loans from ever being “marked-to-market” in a restructuring process. We aim to approach this question more holistically however, essentially making an argument that if an independent expert were to make a bona-fide assessment of such loan’s true market value in a distressed situation, he/she must have applied an additional discount for illiquidity.

• While we heard a wide range of opinions on this particular aspect of our scenario analysis from various experts in this subject matter we felt that at the end of the day, inability to trade cannot be reasonably argued to increase intrinsic value, even if it does make its determination less transparent.

• Permanent credit losses are defined as the peak default rate times expected duration of the cycle (we assume 2 years) times (1 minus recovery rate).

• We also calculate temporary mark-to-market losses based on assumed low print in secondary market prices of 65c in HY, 70c loans, and 60c in private debt. Naturally the confidence in these assumptions must be taken in consideration with expected depth of liquidity.
• We separate between permanent and temporary loss here in an effort to highlight the fact that the latter is not crystalized unless an investor sells at that low print, although everyone is taken for a ride to that level. The permanent loss is unavoidable if a portfolio is exposed to an instrument in question.
• We estimate permanent losses to be roughly 2x the current annual income generated in HY and syndicated loans and 1.3x in private debt. Temporary losses are estimated at 4-5x the annual income level.
• To put it another way, investors stand to wipe out 4-5 years of their income if a recessionary scenario described above were to materialize in this exact form, although a material portion of that is likely to be recaptured in a subsequent upswing. They are also likely to never recover 2 years of their current income, assuming a passive benchmark exposure to HY/loans and 1.3 years to private debt.

 Scenarios #3: a mild/short recession

• Highlighted in grey next to each scenario, we are also showing less stressed scenarios, to give readers a better sense of the range of likely outcomes. We think of these more- and less-stressed scenarios as equally likely to materialize over the next few years, dependent on currently unknown circumstances of the next downturn.
• We also give the private debt a greater benefit of the doubt that recoveries there could be materially better in such less stressed scenario, function of lower leverage and better covenant protections in that space.
• The key takeaway here is that temporary losses could be limited to 3 years of income in HY/loans and 2 years in private debt. Permanent losses could claim 1.5yrs, 1yrs, and 0.6x yrs respectively.
• In a more optimistic scenario, we assume somewhat lower credit losses in loans and private debt. Default rates are assumed at 10% in this less stressed scenario, while recoveries are at 70% in syndicated loans and 75% in private debt (credit given for patient institutional capital, and better structured deals vs syndicated loans)." - source Bank of America Merrill Lynch

We do expect on our side, to repeat ourselves, lower recoveries into the next downturn given "Under the Volcano", there is we think the "liquidity illusion" which is an important factor to take into account in such a scenario analysis and exercise. Anyone who has been through the credit market turmoil of 2007/2008 will tell you that liquidity is a coward and often "bids" are "by appointment only" in such instances.

This is of course a concern which is as well highlighted in Bank of America Merrill Lynch's long interesting report:

"Constrained liquidity as a factor in our analysis

Liquidity has generally been a constraining factor throughout the history of leveraged finance markets. HY bond and leveraged loans have rarely provided investors with particularly deep secondary trading markets – at least, if one’s point of reference is determined by experienced in large cap equities, higher-quality bonds, FX, or commodities.

In the past, there were episodes when lev fin liquidity was relatively good, as was the case in 2006-2007. Additionally, throughout history, there were selected large capital structures that often had deep two-sided markets. Rarely do experienced leveraged finance investors expect deep liquidity to last over considerable time or encompass a considerable number of issuers in this market.

The topic of liquidity in the leveraged finance space has emerged as an issue of particular concern to credit investors, particularly after the Global Financial Crisis. After all, dealers curtailed their market-making activities as a result of both new regulations (capital requirements and the Volcker Rule, the latter which we detail later this section), as well as changes to dealer risk appetite and policies. The days of multi-billion dollar inventories of HY bonds on bank balance sheets came to an end shortly after 2008.

In recent years, aggregate dealer inventories in HY rarely exceed $5bn. This $5bn stand against a $1.3tn market by size and against $6-8bn of average trading volume it generates in a given day.

These facts lead to concerns that while the liquidity situation appears sustainable in times of inflows into the asset class, it may be easily disrupted in times of market stress and significant investor withdrawals. Additionally, if liquidity can be described as limited in HY bonds, and perhaps even more constrained in broadly syndicated loans, it is may be nonexistent in smaller middle-market and private debt spaces, where the whole tranches are often held in only a handful of accounts.

We generally share these concerns and agree with the argument that the next credit cycle will present an important test to the stability of leveraged finance market’s trading infrastructure. The key point here is to remember that while the AUM (assets under management) in funds promising investors daily liquidity gas grown by hundreds of billions of dollars in recent years, the dealer balance sheets went the other way and compressed to a significant extent. With all these reservations in mind, we do not count ourselves among doomsayers that predict a severe dislocation in corporate credit as a result of liquidity constraints.

As we introduced this topic above, we started with a description of the secondary market that has been perennially illiquid with exceptions due to unusually lax risk management episodes or unusually well traded cap structures. Seasoned investors who have participated in this market over several credit cycles understand its liquidity constrains on the DNA level.

The fact that dealers have stepped back has been balanced with the fact of new trading venues, counterparties, and instruments emerging to fill the void.

There are several competing electronic trading platforms in credit space today that did not exist prior to the financial crisis. Hedge funds and other opportunistic investor types are counting themselves among active market makers and they have stepped in during the recent episodes of market volatility with firm bids. Portfolio instruments such as ETFs, total return swaps, and options now complement CDX (credit default swap) indexes in allowing investors to transfer risk more efficiently.

Will the bid-ask spreads widen meaningfully in the next stress episode? Of course they will. Will the market necessarily malfunction in that scenario? Not necessarily. Recent deep stress volatility events such as Dec 2015 (a small distressed fund failing), Jun 2016 (Brexit), Nov 2016 (Trump election), and Jan 2017 (VIX fund failures) have proven that the leveraged finance markets continued to operate. In fact one could argue that all these episodes rewarded those who had the discipline, the risk budget, and the market sense to step in and take advantage of those temporary dislocations. We count ourselves among those who believe in this argument." - source Bank of America Merrill Lynch

We are no perma bears or doomsayers per se but, for us, liquidity in credit markets is a concern, particularly given record issuance levels in recent years also in private credit markets. The GAM fund meltdown during the summer is illustrative of our concerns. 

Growth in issuance is a problem also highlighted by Morgan Stanley in their Corporate Credit Research note from the 5th of October entitled "The Nature of the BBBeast":

"BBB IG debt outstanding has grown significantly in this cycle, a story most IG credit investors know quite well. For example, at ~$2.5 trillion outstanding, BBB par has increased 227% since the beginning of 2009.

The majority of the increase in BBB debt stems from net issuance ($1.2 trillion), followed by downgraded debt ($745 billion). Notably, the growth in BBB debt outstanding is not being skewed by a single sector or a small part of the market. Yes, large issuers have grown significantly. For example, the top 25 non-financial BBB names have a total of $685 billion in index debt (up from $257 billion in 1Q09). But the number of BBB issuers has also increased by 60% since 2009, while all sectors have increased BBB debt, large and small companies alike. In other words, the increase has been broad-based across the market.

So what does this mean big picture? Credit cycles are always different from one to the next. But a consistent rule of thumb over time that we live by when looking for problems down the line: Follow the debt growth. Very simply, applying to the current cycle, we think BBBs will be one (of a few) stress points when the cycle does turn. Downgrade activity will likely be meaningful. And when thinking about other markets that could feel the effect, remember the BBB part of the IG index is now ~2.5x as large as the entire HY index.

The good news is that this is not a story for today, in that ratings downgrades tend to lag the market. In other words, the big wave of downgrades will likely not come until credit spreads are much wider than they are right now, which will take time to play out. But more importantly, valuations are pricing in very few fundamental risks, in our view, with the BBB/A spread basis still near cycle tights. Hence we remain up-in-quality." - source Morgan Stanley

As central banks are pulling back, “macro” driven markets are no doubt making a return and credit indices such as Itraxx Main Europe and CDX IG and High Yield in the US are useful tool to hedge rising “liquidity” risk coming from credit markets when next downturn will show up.

Finally, for our final chart, as we pointed out during in previous conversations, 2018 displayed larger and larger standard deviations move, typical as well of late cycle behavior in conjunction with rising dispersion. 

• Final chart - Large standard deviation moves, the "market" volcano is becoming more "active"

The latest bout of volatility wasn't that much of a surprise, it was a conjunction of several factors such as fast rising real rates, a more aggressive tone from the Fed in general and Powell in particular. Whereas the February event was the equivalent for the house of straw of the short-vol pigs of the eruption of Mount Vesuvius in 79 AD, vaporizing in an instant large players of the short-volatility complex, the latest event was mostly a tremor, geopolitical risks aside. We do not yet see credit spreads turning decisively, nonetheless the deteriorating trend for cyclicals in conjunction with trade deceleration outside the United States warrants close attention we think. Our final chart comes from Morgan Stanley's Cross-Asset Dispatches note from the 11th of October entitled " FAQ After a Large Decline":

"Large moves are still happening more often: 

This remains true; 2018 is still on pace for some of the highest frequency of 3-sigma moves post-crisis. Liquidity remains poor.

What happened?

We think that recent moves are about several factors colliding around the 3.20% level for 10-year Treasuries, rather than a simple case of 'higher rates are bad for risk'. Those factors? A break of a 5-year+ real yield range, compression of the US equity ERP above 3.25% and very stretched performance of value versus growth (see Cross-Asset Dispatches: Are Rising Rates a Problem? October 7, 2018).

How unusual was this move?

The overall move for S&P 500 wasn't that extreme versus what we saw over the last several years but yesterday was the worst day for the NASDAQ in almost seven years. More broadly, this was also one of the worst days for growth globally. The value outperformance was even more pronounced outside the US as European value posted the best one-day performance versus growth post-crisis.

Positioning – it is light, but in pain: 

2018 has been a hard year (see Easier Financial Conditions, Still a Tough Year, September 23, 2018). The last five days have only confirmed this, bringing losses to one of the last bastions of strong performance and concentrated positioning – growth/tech. Investors have been hit hard by recent price action, which makes us less optimistic than we'd otherwise be about overall positioning indicators looking quite light.

2018, unfortunately, seems to be a year where every asset class has a turn in the barrel.

The Fed 'put' remains out-of-the-money: 

We also do not expect much help for policymakers, at least not yet. US inflation and unemployment remain in a very different place than under Chairs Yellen and Bernanke. As of late September, the Chicago Fed's Adjusted Financial Conditions Index was still easier year on year (in a tightening cycle, we think that the Fed would want this tightening). And we think that the Federal Reserve strongly values its independence; comments by the administration are unlikely to have an impact." - source Morgan Stanley

Sure, things are brewing "under the volcano à la 2007", one might opine, and of course geopolitical events continues to be known unknowns, yet the US still appear for the time being as much stronger magnet for global capital than Europe for instances as per the significant amount of outflows seen in recent weeks. It is again a case of "Dissymmetry of lift" we think, yet, the latest signs of global liquidity withdrawal are showing again dispersion such as rotation from growth to value, and investors turning more defensive in some instances given we are entering the latest innings of this long credit cycle but, we are repeating ourselves again...

"Hope of ill gain is the beginning of loss." - Democritus

Stay tuned !