“Even the largest avalanche is triggered by small things.”
– Vernor Vinge, American writer
Looking at the pre-revolutionary mindset of my home country France, with Paris under siege as we warned about in our conversation “Last of the Romans” in mid-November, in conjunction with the very fast fading rally seen on the back of the United States and China trade war truce, when it came to selecting our title analogy given liquidity is continuing to be withdrawn by the Fed’s QT – though as of late it seems they blinked, with softer global macro data including US housing – we decided to go for “The Sorites paradox”.
The “Sorites paradox”, sometimes called the paradox of heap is a paradox that arises from vague predicates. A typical formulation involves a heap of sand, from which grains are individually removed. Under the assumption that removing a single grain does not turn a heap into a non-heap, the paradox is to consider what happens when the process is repeated enough times: is a single remaining grain still a heap? If not, when did it change from a heap to a non-heap? A common first response to the paradox is to call any set of grains that has more than a certain number of grains in it a heap. If one were to set the “fixed boundary” at, say, 10,000 grains, then one would claim that for fewer than 10,000, it is not a heap; for 10,000 or more, it is a heap. A second response attempts to find a fixed boundary that reflects common usage of a term.
So the question is, when is a bubble a bubble? When will QT turn the bubble into not a bubble? We wonder. Is hysteresis being the dependence of the state of a system on its history the answer? Equivalent amounts of sand may be called heaps or not based on how they got there. If a large heap (indisputably described as a heap) is slowly diminished, it preserves its “heap status” to a point, even as the actual amount of sand is reduced to a smaller number of grains. For example, suppose 500 grains is a pile and 1,000 grains is a heap. There will be an overlap for these states. So, if one is reducing it from a heap to a pile, it is a heap going down until, say, 750. At that point, one would stop calling it a heap and start calling it a pile. But if one replaces one grain, it would not instantly turn back into a heap. When going up, it would remain a pile until, say, 900 grains. The numbers picked are arbitrary; the point is, that the same amount can be either a heap or a pile depending on what it was before the change. Also, one can establish the meaning of the word “heap” by appealing to consensus. The consensus approach typically claims that a collection of grains is as much a “heap” (or bubble) as the proportion of people in a group who believe it to be so. In other words, the probability that any collection is considered a heap is the expected value of the distribution of the group’s views aka a “Quasitransitive relation“, but we digress.
In this week’s conversation, we would like to look at why it is increasingly important to play much more defensively as we move towards what could be a jittery 2019.
- Macro and Credit – Don’t be the last one left to pick up the “credit” tab…
- Final chart – Liquidity and credit spreads go hand in hand
Macro and Credit – Don’t be the last one left to pick up the “credit” tab…
Looking at Wednesday’s drop of 3.10% of the Dow Jones, as we pointed out in our last conversation, it wasn’t really that surprising given the weakening decelerating tone in global growth in general and cyclicals such as Autos and Housing in particular:
“Rising dispersion has clearly been the theme in 2018 when it comes to credit. The Fed’s tightening stance in conjunction with QT and the surge in the US dollar have clearly been headwinds for the rest of the world. Yet the US have shown in recent months that it wasn’t immune to gravity and deceleration as the fiscal boost fades in conjunctions in earnings and buybacks. 2018 also marks the return of cash in the allocation tool box and many pundits have started to play defense by parking their cash in the US yield curve front-end.”
– Source: Macronomics, November 2018
As well, we pointed last week that if you wanted to go “short” credit, then US leveraged loans were definitely something to look at, as pointed out by Lisa Abramowicz on a Twitter feed:
“Prices on leveraged loans have dropped to the lowest since 2016 even though their benchmark rate Libor has quickly risen, meaning this debt should pay out higher interest rates. This throws into question the concept of floating-rate debt as a hedge against rising rates.
On one hand, loan investors get more income from their holdings as rates rise. On the other, the market seems to perceive the corporate borrowers as less creditworthy as their cost of financing rises. So if loan investors want to sell their holdings, they’ll get a lower price.” – graph source Bloomberg
– Lisa Abramowicz, Twitter feed
Indeed, liquidity is, as always, a “coward”, and the longer you stay at the “credit” bar, the likelier you are to be shocked by “price discovery” when the credit markets will in earnest turn “South” and you will end up picking an expensive bar tab. Hence our call for reducing your illiquid high beta exposure.
While leverage loans are an evident target pointed out by so many investor pundits, regulators and central bankers in these days and ages, other interesting instruments such as AT1s, aka Contingent Convertibles (CoCos), as well as Corporate Hybrid bonds fit the bill when it comes to being potentially “illiquid” and harmful. Sure, it’s fun on the way up, pretending to generate “alpha” for your clients by playing the “beta” pure carry game, but when the party has been extended, as it has been in credit land, then again, not starting to be a little bit more “cautious” is a good recipe for asking for trouble and finding it eventually through “price discovery”.
On the subject of “illiquidity” and credit, we read with interest J.P. Morgan’s Portfolio Insights note relating to the evolution of market structure. Their paper is entitled “Managing illiquidity risk across public and private markets”:
“In theory, investors are compensated for this through the higher returns available in private assets over the full life cycle of the private investment. In other words, to harvest the illiquidity risk premium in private markets, investors need to be able to stay the course, weathering any variation in the cash flow profile over the full cycle. This means that cash calls would need to be funded from elsewhere in the portfolio.
The ability to accept this type of risk ranges widely across investor types. Those that may be subject to redemptions or fund withdrawals (e.g., mutual fund managers) are less able to bear uncompensated illiquidity risk than those with a long- term pool of capital to deploy (e.g., sovereign wealth investors). Further, during times of market crisis, when investors are already seeking to cut exposure to public markets, threats to liquidity are generally correlated and can compound to become a serious issue for investors. Investors could face liquidity demands arising from redemptions and a prudent desire to hold higher portfolio cash buffers. At the same time, on the private asset side there may be cash calls to finance, calls that are best covered from public assets – and thus, avoiding uncompensated illiquidity traps in public markets becomes a priority. To fully assess the illiquidity risk in a portfolio, all of these factors need to be considered holistically.
Taking high yield (HY) bonds as an example of a potentially illiquid public asset with both market and illiquidity risk, we can ask whether, over a defined time horizon, the probability of being forced to crystallize a loss under adverse liquidity conditions is appropriately compensated (see Addendum, “Modeling the cost of high yield trading under illiquid conditions”). Early in the economic cycle, when credit spreads are wide, the illiquidity premium in an asset such as high yield credit may well offer an additional return compared with a replicating stock-bond portfolio. However, as the cycle matures and credit spreads tighten, there will come a tipping point – some breakeven level of spread – where the return in credit is not sufficient to offset the probability-weighted risk of a loss over a defined time horizon. Effectively, the illiquidity risk has at that point become uncompensated and investors may be better served expressing their desired level of market risk via a replicating stock-bond portfolio.
The scale of the potential illiquidity during times of market stress is demonstrated in Exhibit 8, again using HY credit as an example. The illiquid credit asset will suffer from wider bid-ask spreads and much reduced transaction volumes; large transactions can take considerable time to execute in markets where prices are dropping sequentially over multiple trading sessions.
Turning to private market assets, as investors have increasingly added private assets to portfolios there is commensurately more focus on the risk that they could be forced to liquidate private investments at an inopportune time to meet an additional capital call. Alternately, redemptions and other portfolio-level cash requirements may force them to exit private investments at an undesirable point. Since such events tend to occur during adverse conditions in public markets and the economy at large, the most relevant question is how bad things might really get.”
– Source: J.P. Morgan
Exactly, large positions can take a long time to unwind, particularly when dealer inventories are nowhere near to the level they had prior to the Great Financial Crisis (GFC).
Also as another illustration of what it would take to liquidate a sizable position of $1 billion in US High Yield in the case of a recession and where credit spreads should be to reflect that “illiquidity” premia would be significantly wider, as pointed out in J.P. Morgan’s note:
“For an investor that may need to liquidate $1 billion of high yield and anticipates any crisis to be average in its severity, credit spreads above around 270bps compensate for illiquidity risk. But if the investor’s subjective view of the probability of recession over the next year were to increase to 33%, then the breakeven credit spread required to compensate fully for illiquidity risk would jump to 320bps and as high as 398bps in a worst-case drawdown scenario. As portfolio size increases – and the potential illiquid asset trade size grows – the ex-ante breakeven spread required to compensate for illiquidity risk increases. Crucially, there is no economy of scale for illiquidity risks and, indeed, there are very apparent diseconomies of scale.”
– Source: J.P. Morgan
Given the significant rise in corporate credit issuance in recent years, one could conclude that current credit spreads do not reflect this “illiquidity” premium.
But given that the Fed seems to have recently “blinked” recently following a more dovish tone at the Economics Club of NY by Jerome Powell, one could indeed think that there is still value for “credit” pickers even in US High Yield. We have long argued that as dispersion is rising in this late-cycle environment, proven credit specialists in the issuer selection process would outperform the passive investment crowd.
This effectively could dampen slightly the widening moves seen recently. On this subject, we read Wells Fargo’s take from their Credit Connections note from the 30th of November, entitled “Honey B’s”:
“Credit markets continue to groan under the pressure of policy uncertainty (both monetary and fiscal), trade wars, a recent upsurge of idiosyncratic events and heavy bond supply. That said, secondary market credit spreads appear to be stabilizing after Fed Chairman Jay Powell struck a more dovish tone at his recent presentation to the Economic Club of NY. The shift was subtle and nuanced, but enough to convince markets that the Fed may slow the pace of policy tightening in the coming months. With credit spreads at year-to-date and multi-year wides, we recommend that investors start to set up for 2019 with select longs in credits poised to deleverage next year and beyond. Triple-B and double-B credits look particularly attractive to us in this environment.
The build-up of debt, increased borrowing costs and tighter monetary policy suggests that a growing number of companies will need to focus on deleveraging and balance sheet repair next year to preserve credit ratings and ensure ongoing access to capital markets. While this might not be great news for the economy it should certainly help the overall credit worthiness of the corporate sector and allow credit spreads to compress (somewhat) after a year of sustained widening. Conversely, those companies that cannot or will not address balance sheet issues will find a much less forgiving investor base and considerably higher borrowing costs. In this environment credit selection is paramount.
When considering which companies have the greatest urgency to deleverage it is helpful to look at the distribution of debt maturities. The term structure of maturities is a simple analysis to look at when companies are faced with refinancing pressures. At a high level, the amount of refinancing risk faced by IG companies is considerably greater than the risk faced by HY companies over the next three years. As the charts below show, about $2.2 trillion of investment grade debt is scheduled to mature 2019-2021.
This represents about 30% of all outstanding IG debt, with roughly 53% owed by non-financial companies and 47% owed by banks, insurers, financial companies and REITs. Maturities steadily climb over the next three years and peak in 2021 with about $800 billion of debt scheduled to mature. This stands in sharp contrast to HY. Over the same period, about $250 billion of HY debt scheduled to mature represents about 15% of all outstanding HY debt. In fact, next year HY maturities total just $40 billion.
Considering HY companies have issued about $175 billion of debt this year, refinancing pressures look particularly light next year. As a result, in the aggregate, HY companies look better positioned than IG companies to deal with tighter monetary conditions and higher interest rates over the next few years.”
– Source: Wells Fargo
Though, as pointed out by Morgan Stanley in our recent conversation, the effect of widening credit spreads on the unemployment rate is significant and positive:
“every 10bp sustained widening of BBB/Baa corporate credit spreads is associated with a 0.15pp rise in the unemployment rate after two quarters, all else equal.”
– Source: Morgan Stanley
The most important chart going forward? Jobless Claims vs. Job Cut Announcements (trend forming?):
– Graph Source: Bloomberg
Leading indicators of initial unemployment claims rose again last week to 234K (and 220K consensus) from 224K, with 4-week average up to 223K from 219K (up from 211K average in Q3). Take notice of this!
So, from a “Sorites paradox” perspective, if wider spreads impact unemployment going forward due to balance sheet deleveraging, earnings and profitability matter as well when it comes to predicting a surge in defaults rates, as highlighted by our friend Edward J Casey in his November credit commentary:
“Probability of nonfinancial corporates is a key driver of the default rates
The US default rate declined to 3.2% in October and is expected to improve to 2% in next year.
One driver behind profit growth has been the massive expansion in corporate debt which has grown by +4.3%, nearly twice the pace of real GDP.
Cumulatively nonfinancial profits gained +115% compared to GDP of +23% and corporate debt of +46%.”
– Source: Edward J Casey
In our “credit” heap of sand, grains are individually removed, thanks to the Fed’s QT. The one question that remains is which grain will trigger the avalanche?
So, if illiquidity is not “priced” correctly in credit and valuations are considered “lofty” from a consensus approach perspective, then again one may rightly ask when it will break. As pointed out by Edward J Casey, it is all depending on corporate profits rolling over. Watching earnings in 2019 will be paramount:
On prior occasions spikes in the cost of debt have been coincident with annual gains in corporate profits rolling over”
– Source: Edward J Casey
If wages growth continues to accelerate in conjunction with earnings coming under pressure, then equities valuations in 2019 could be “repriced” much lower, just saying.
And when it comes to earnings, we are closely watching the space. We read with interest Bank of America Merrill Lynch’s Revision Ratios report from the 30th of November, entitled “More cutting, less raising”:
“Earnings Revision Ratio
US joins the rest of the world in negative revisions
In November, the 3-month earnings estimate revision ratio (ERR) fell to 0.87 from 1.11 – the biggest decline since April and the lowest level in nearly two years. A ratio of below 1 means more cuts than raises to earnings estimates, and this is the first time in over a year-and-a-half that analysts have taken down estimates. The US has caught up with the rest of the world with more cuts than raises. Trends have decelerated since early 2018 following tax reform as we expected, but the ratio now sits just a hair above its long-term average. We use the 3m ERR as one of five inputs in our market outlook: an above-average ratio has generally preceded strong near-term returns, whereas a ratio below average has signaled more muted near-term returns (Chart 2).
- In November, the three-month (3m) earnings estimate revision ratio (ERR) fell to 0.87 from 1.11 – the biggest monthly decline since April and the lowest level in nearly two years.
- With the ratio now below 1.0, this suggests more cuts than raises to earnings estimates for the first time in over a year-and-a-half.
- The ERR sits just slightly above its long-term average of 0.87.
- The more volatile one-month (1m) ERR similarly fell to a two-year low of 0.73 from 0.77.
- The ratio is below 1.0, suggesting more cuts than raises to earnings estimates for the second consecutive month.
- In November, all sectors (except Staples) saw their 3m ERR moderate (Chart 1). Energy, Real Estate, and Technology saw the biggest deterioration.
- Most sectors are now seeing more cuts than raises to earnings estimates amid widespread deterioration in revision trends. Energy, Utilities, and Tech are seeing slightly more positive than negative revisions to earnings estimates.
- Utilities is the only sector with an above-average ERR, while the ratio is in line with the historical average for Financials, Health Care, Industrials, and Technology.
- Similarly, throughout November, most sectors except Utilities and Technology saw more negative than positive revisions to estimates.
- Energy, Materials, and Comm Svcs had the weakest one month revision trends.
– Source: Bank of America Merrill Lynch
On top of the deteriorating picture for “earnings”, Bank of America Merrill Lynch also added the following in their report:
“Bad breadth in credit? Distress ratio ticks up
We watch revision ratios closely because “breadth” measures can sometimes be early indications of broader issues within markets. Our High Yield team’s distress ratio – the percentage of US high yield bonds with an option-adjusted spread above 1000bp – has similarly proven to be a good leading indicator of defaults. This ratio has recently worsened, and now sits at a 10-month high.”
– Source: Bank of America Merrill Lynch
Because we do not like this picture, we think from a “Sorites paradox” perspective you should continue to reduce your high beta exposure and rotate from growth to consumer staples equities-wise, also reduce your financials’ subordinated pocket (until and if a new LTRO is announced by ECB), and continue to raise cash levels and park it in the US front-end of the curve. We think as well that recent moves in the long end of the US yield curve look enticing. We are looking at the 30-year bucket and long-dated zero coupons given that the “deflationista” camp seems to make a comeback with global growth clearly decelerating.
As we pointed out earlier in our conversation, QT is indeed reducing the size of the credit heap (bubble). Trade accordingly, as per our final chart below.
Final chart – Liquidity and credit spreads go hand in hand
The tone in credit spreads has had a weaker tone in recent weeks on the back of significant outflows from mutual funds, as the Fed has been continuing to withdraw liquidity in the system with QT. Our final chart comes from CITI European Portfolio Strategist note from the 22nd of November, entitled “Post QE World – Bear Market, Bull Market or Kangaroo Market”, and displays the relationship between central banks liquidity and Investment Grade credit spreads:
Liquidity and financial markets have been tied closely together
Citi credit strategists have shown a powerful relationship between net central bank purchases and moves in credit spreads and equity markets. By extension, reducing/reversing QE should drive credit spreads sharply higher and equity prices sharply lower. From our (equity) side, we argue that it depends very much on the nominal growth backdrop and the pace of tightening. Progressive tightening and an extending economic cycle are still likely to see credit spreads widen, but also are likely to support an extending profit cycle. Historically, there is a phase in markets where credit spreads widen but equity markets make fresh highs driven by rising EPS. This remains our base case, but we acknowledge the end cycle debate.”
– Source: CITI
Sure, it was a fun and exciting long credit party, but from our perspective and in respect to the “The Sorites paradox” and the risk of an avalanche, we would rather start in earnest to play “defense”, given 2019 might prove to be even more problematic for risky asset prices than 2018. Just saying…
“The one certainly for anyone in the path of an avalanche is this: standing still is not an option.”
– Norman Davies, British historian