Wednesday 28 February 2018

Macro and Credit - Buckling

"There exists everywhere a medium in things, determined by equilibrium." -  Dmitri Mendeleev, Russian scientist

Watching with interest the slowly grind higher in US interest rates with some weakening signs coming from US economic data such as the US trade deficit in goods getting spanked with orders for larger domestic appliances and other durable goods falling by a cool 3.7% from the month before, led by a hard drop in vehicle demand, when it came to choosing our title analogy for this week's conversation we reminded ourselves of "buckling" being a mathematical instability that leads to a failure mode. When a structure is subjected to compressive stress, buckling may occur. Buckling is characterized by a sudden sideways deflection of a structural member. This may occur even though the stresses that develop in the structure are well below those needed to cause failure of the material of which the structure is composed. As an applied load is increased (US interest rate hikes) on a member, such as a column, it will ultimately become large enough to cause the member to become unstable and it is said to have buckled. Further loading will cause significant and somewhat unpredictable deformations, possibly leading to complete loss of the member's load-carrying capacity. If the deformations that occur after buckling do not cause the complete collapse of that member, the member will continue to support the load that caused it to buckle. If the buckled member is part of a larger assemblage of components such as a building, any load applied to the buckled part of the structure beyond that which caused the member to buckle will be redistributed within the structure. In a mathematical sense, buckling is a bifurcation in the solution to the equations of static equilibrium. At a certain point, under an increasing load, any further load is able to be sustained in one of two states of equilibrium: a purely compressed state (with no lateral deviation) or a laterally-deformed state. Obviously we thing that financial markets have reached a bifurcation point and we have yet to see how the buckle of rising interest rates will be redistributed within the complex structures without leading to some renewed avalanche in some parts of the markets.

In this week's conversation, we would like to look at the vulnerability of equities and credit markets to a more hawkish tone of the Fed which would lead to more aggressive rate hikes should the "Big Bad Wolf" aka inflation continue to rear its ugly head.  

Synopsis:
  • Macro and Credit - Hike it till you break it
  • Final chart - Afraid of buckling? Watch credit availability


In our March 2017 conversation entitled "The Endless Summer" we concluded our missive at the time asking ourselves how many hikes it would take before the Fed finally breaks something. Given the arrival of a new Fed "sheriff" in town one might wonder if the pace will be as gradual as it seems should the Fed feels it is falling behind the curves when it comes the "Big Bad Wolf" aka inflation and current loose financial conditions. As we pointed out in our recent conversations the recent uptick in inflation coincided with a sharp sell-off in equities. Sure, one would point out to us that correlation doesn't mean causation, but, it certainly felt like the very crowded short-volatility complex was looking for a match that triggered the explosion and for some their ultimate demise.  The U.S. Average Hourly Earnings triggered the "buckling" as it brought back the fear in the markets of the return of the Big Bad Wolf aka "inflation".  For some it seems like us, it seems the "Big Bad Wolf" has already blown apart the "short vol" pig's house which was made of straw. If indeed the short-vol house was made of straw we wonder if the pig's equities markets house is made of sticks or and if the pig's credit markets house is made of bricks. The difficulty for the Fed in the current environment is the velocity of both the rates rise and inflation, because if indeed the Fed hike rates too quickly then it will trigger some other avalanches down the capital structure (short-vol complex being the equity tranche or first loss piece of the capital structure we think). If inflation and growth rise well above trend, then obviously the Fed will be under tremendous pressure to accelerate its normalization process. It is a very difficult balancing act.

When it comes to the bounce back for equities following the short-vol avalanche, which could have been possibly triggered by the recent uptick of inflation, we read with interest Deutsche Bank's Asset Allocation note from the 23rd of February entitled "Inflation and Equities" with the long summary below:
"The recent uptick in inflation coincided with a sharp correction in equities
Whether this was cause and effect is debatable for a variety of reasons and around half the correction reversed quickly (Stretched Consensus Positioning, Jan 31 2018; An Update On The Unwind, Feb 12 2018). Nonetheless, late in the business cycle with a tight labor market, strong growth, a lower dollar, higher oil prices and a fading of one off factors, all point to inflation moving up. What does higher inflation mean for equities? We discuss five key questions.
Is inflation bad for margins and earnings? Historically, higher inflation has been associated with higher margins and strong earnings growth
■ Conceptually, higher inflation is ambiguous. From a pricing vs cost perspective, whether higher inflation leads to higher or lower margins depends on the relative strengths of price vs wage and other input cost inflation. It depends on the relative importance of variable vs fixed costs. And on the extent to which corporates can increase productivity in response to cost pressures. It is notable that while markets seem to have been surprised by the recent uptick in wage inflation, corporates have been noting it for at least a year. Finally, inflation does not occur in a vacuum. The drivers of higher inflation matter and when it reflects strong growth, it implies not only higher sales but operating leverage from fixed costs can raise margins and amplify the impact on earnings.
Historically, the empirical evidence is unambiguous. Higher inflation was clearly associated with higher margins and strong earnings growth.
Does higher inflation mean lower equity multiples? By how much? A 1 pp rise in inflation compresses equity multiples by 1 point or a decline in prices of around 5% from recent pre-correction levels
■ The correlation between bond yields and equities depends on the driver: inflation (-) or real rates (+). Contrary to popular notions that higher bond yields mean lower equities, the historical relationship between bond yields and equities has been ambiguous (Long Cycles In The Bond-Equity Correlation, May 2014). Instead, the impact of higher yields on equities depends on whether they reflect higher inflation (-) which has always been negative for equities; or whether higher yields reflect higher real rates (+) which have always been positive for equities until real rates reached very high levels (greater than 4%--seen only once during the Volcker disinflation) (Do Higher Rates Mean Lower Equity Multiples? Sep 2014).
■ Why is higher inflation negative for equity valuations? When inflation moves up, the hurdle rate for all nominal investments moves up and in turn bond yields and earnings yields (inverse of the equity multiple) move up.
■ A 1pp rise in inflation compresses multiples by 1 point. A majority (70%) of the historical variation in the S&P 500 multiple is explained by its drivers: earnings/normalized levels (-); payouts (+); rates broken up into inflation (-) and real rates (+); and macro vol (-). Our estimates imply that a 1pp rise in inflation lowers the equity multiple by 1 point or a 5% decline in prices from the recent peak. Our house view and the consensus sees a somewhat smaller rise in inflation over the next 2 years. These ranges of increases in inflation imply a modest pullback in equities that would put it within the bands of normal 3-5% pullbacks that have historically occurred every 2-3 months.
Is the inflection in inflation a leading indicator of the end of the cycle? How long is the lead? On average 3 years, but the Fed’s reaction is key
With an average correction in equities of 21% around recessions, the timing of the next one is obviously key. If the recent uptick marks the typical mid- to latecycle inflection up in inflation, how long after did the next recession typically occur? On average 3 years, which would put it in late 2020. But the timing is likely determined critically by the Fed’s reaction. Historically, a Fed rate-hiking cycle preceded most recessions since World War II, with recessions occurring only after the Fed moved rates into contractionary territory. Arguably the Fed did this only after it was convinced the economy was overheating and it continued hiking until the economy slowed sufficiently or went into recession. At the current juncture, core inflation has remained below the Fed’s target of 2% for the last 10 years and several Fed officials have argued for symmetry in inflation outcomes around the target, i.e., to tolerate inflation above 2%. It is thus likely that the Fed will welcome the rise in inflation for now and simply stick to its current guidance, possibly moving it up modestly.
How high will inflation go? If inflation expectations remain range bound, core PCE inflation will stay within its narrow band of 1-2.3% in which it has been for the last 23 years
Outside the Great Inflation of 1968-1995, core PCE inflation has remained in a remarkably narrow band (Six Myths About Inflation, Oct 2017). The period since 1996 encompassed 3 business cycles that saw unemployment fluctuate between 3.8% and 10%; the dollar rise and fall by 40% more than once; oil prices rise 7-fold and almost completely reverse. Yet inflation remained in a narrow band unusual for an economic time series. Indeed, with a standard deviation of 35 bps, much of the range of variation in inflation since 1996 cannot be differentiated from the normal noise inherent in macro data.
The stability of inflation across large business- dollar- and oil-cycles in our view reflects the stability of inflation expectations which are the only driver of inflation over the long run. Inflation expectations have been stable since the mid-1990s, fluctuating for most of the last 23 years in a tight 50bps range and for most of it in an even narrower 30bps range. Following the dollar and oil shocks of 2014-2015, inflation expectations fell out of and are still 20bps below this range and 50bps below average. Absent large unexpected and persistent shocks, inflation expectations evolve slowly. It has in fact been difficult for policy makers to effect changes in inflation expectations as the recent experience of Japan and  the 10-year miss on the core PCE inflation target in the US illustrate (Six Myths About Inflation, Oct 2017).
What about all the stimulus? The impact of the stimulus will follow a pickup in growth with long lags (1½ years)
It is well known that inflation responds with long lags to growth, a tightening labor market and the dollar. Consider that the correlation between real GDP growth and core CPI inflation is a modest positive 5%. But when GDP growth is lagged by 6 quarters, the correlation jumps to a much stronger 80%. The lagged relationship implies that a sustained 1pp increase in GDP growth raises core inflation by 20bps after 1½ years. Our house forecast for GDP growth which is above consensus implies GDP growth of near 3% and core inflation peaking around 2.2% in 2020.
Growth outcomes significantly above our house view would need to materialize and sustain to raise inflation above and outside the band of the last 23 years. Moreover there would be plenty of lead time with growth needing to sustain at high levels for a prolonged period (1½ years) before it moved inflation up."  - source Deutsche Bank
As we repeated in numerous conversation, for a bear market to materialize you would need a significant pick-up in inflation for your "buckling" to occur and to lead to a significant repricing of risky asset prices such as equities and US High Yield. But what is very interesting to us is that the buildup in the trade war rhetoric coming from the US could be a harbinger for higher inflation down the line given that companies would most likely increase their prices with rising import prices that would be passed on already stretched consumers thanks to solid use of the credit cart (nonrevolving credit). 

In our recent conversation "Bracket creep", which describes the process by which inflation pushes wages and salaries into higher tax brackets, leading to a fiscal drag situation, we indicated that with declining productivity and quality with wages pressure building up, this could mean companies, in order to maintain their profit margins would need to increase their prices. To repeat ourselves "Protectionism", in our view, is inherently inflationary in nature. 

To preserve corporate margins, output prices will need to rise, that simple, and it is already happening. This can have a significant impact on earnings particularly when the S&P 500 Net Income Margins LTM is at close to record levels as indicated in Deutsche Bank's note:
"Inflation and earnings
Is inflation bad for margins and earnings? Historically, higher inflation has been associated with higher margins and strong earnings growth
Conceptually, higher inflation is ambiguous. From a pricing vs cost perspective, whether higher inflation leads to higher or lower margins depends on the relative strengths of price vs wage and other input cost inflation. It depends on the relative importance of variable vs fixed costs. And on the extent to which corporates can increase productivity in response to cost pressures. It is notable that while markets seem to have been surprised by the recent uptick in wage inflation, corporates have been noting it for at least a year. Finally, inflation does not occur in a vacuum. The drivers of higher inflation matter and when it reflects strong growth, it implies not only higher sales but operating leverage from fixed costs can raise margins and amplify the impact on earnings.
Historically, the empirical evidence is unambiguous. Higher inflation was clearly associated with higher margins and strong earnings growth.
- source Deutsche Bank

With the S&P 500 Net Income Margins LTM close to record levels and with the recent rise in prices operated by companies recently, it remains to be seen how long can margin levels remain this elevated. Sure, the fiscal boost provided by the US government should provide additional support yet the big question for us is relative to the US consumer and its sensitivity to rising prices as we discussed in the final point of our conversation "Harmonic tremor". Have we seen peak "Consumer confidence" and peak PMIs recently? One thing for certain is that Citigroup’s US Economic Surprise Index (CESIUSD Index) as an indicator of economic momentum has started to "buckle" recently. There is a clear relationship between the CITI's Economic Surprise Index and the Fed's monetary policy. When the Fed is in tightening mode, good news such as rising inflation expectations is generally seen as bad news. In spread terms, only high yield is sensitive to macro surprises. Moreover, the response of high yield spreads to macro surprises is "monotonic" in ratings: the lower the rating, the stronger the response. In our conversation "A shot across the bows", we indicated the following when it comes the Citi Economic Surprise Index (CESI). It could potentially indicate that economic fundamentals are trading ahead of themselves and could portend some credit spreads widening in the near future given there is a reasonably strong relationship between the inverse of Citigroup Economic Surprise Index and both the IG CDX and HY CDX. So all in all, you want to watch what the CESI does in the coming weeks and months. 

But moving back to the impact of the "Big Bad Wolf" aka inflation on equity multiples, we read with interest as well the other part of Deutsche Bank's report on the impact inflation can have:
"Inflation and equity multiples
Does higher inflation mean lower equity multiples? By how much? A 1 pp rise in inflation compresses equity multiples by 1 point or a decline in prices of around 5% from recent pre-correction levels
■ The correlation between bond yields and equities depends on the driver: inflation (-) or real rates (+). Contrary to popular notions that higher bond yields mean lower equities, the historical relationship between bond yields and equities has been ambiguous (Long Cycles In The Bond-Equity Correlation, May 2014). Instead, the impact of higher yields on equities depends on whether they reflect higher inflation (-) which has always been negative for equities; or whether higher yields reflect higher real rates (+) which have always been positive for equities until real rates reached very high levels (greater than 4%--seen only once during the Volcker disinflation) (Do Higher Rates Mean Lower Equity Multiples? Sep 2014).
Why is higher inflation negative for equity valuations? When inflation moves up, the hurdle rate for all nominal investments moves up and in turn bond yields and earnings yields (inverse of the equity multiple) move up.
■ A 1pp rise in inflation compresses multiples by 1 point. A majority (70%) of the historical variation in the S&P 500 multiple is explained by its drivers: earnings/normalized levels (-); payouts (+); rates broken up into inflation (-) and real rates (+); and macro vol (-). Our estimates imply that a 1pp rise in inflation lowers the equity multiple by 1 point or a 5% decline in prices from the recent peak. Our house view and the consensus sees a somewhat smaller rise in inflation over the next 2 years. These ranges of increases in inflation imply a modest pullback in equities that would put it within the bands of normal 3-5% pullbacks that have historically occurred every 2-3 months.
- source Deutsche Bank

Obviously from a "buckling" perspective the big question is whether higher yields reflect higher real rates (+) which have always been positive for equities until real rates reached very high levels, or, are they reflecting a higher inflation risk, in which case the repricing could be more severe as the Fed would probably step up on its hiking gear. For the positive momentum to hold and goldilocks environment to continue, you would need inflation and growth not running too hot, so that the Fed can gradually hike rather than stepping up its hiking pace. This is as well clearly highlighted by Charlie Bilello from Pension Partners in his blog post from the 15th of February entitled "Inflation, deflation and stock returns".

Again, it is a matter of "velocity" in the movement. An exogenous factor such as a geopolitical event that would trigger a sudden and rapid rise in oil prices would of course upset the situation and be much more negative for equities as we saw with the huge rise in oil prices prior to the Great Financial Crisis (GFC) of 2008.

One might therefore rightly ask if indeed inflation could be a leading indicator for recession. This is also a point which has been discussed in Deutsche Bank's very interesting note:
"Inflation as a leading indicator of recession
Is the inflection in inflation a leading indicator of the end of the cycle? How long is the lead? On average 3 years, but the Fed’s reaction is key
With an average correction in equities of 21% around recessions, the timing of the next one is obviously key. If the recent uptick marks the typical mid- to latecycle inflection up in inflation, how long after did the next recession typically occur? On average 3 years, which would put it in late 2020. But the timing is likely determined critically by the Fed’s reaction. Historically, a Fed rate-hiking cycle preceded most recessions since World War II, with recessions occurring only after the Fed moved rates into contractionary territory. Arguably the Fed did this only after it was convinced the economy was overheating and it continued hiking until the economy slowed sufficiently or went into recession. At the current juncture, core inflation has remained below the Fed’s target of 2% for the last 10 years and several Fed officials have argued for symmetry in inflation outcomes around the target, i.e., to tolerate inflation above 2%. It is thus likely that the Fed will welcome the rise in inflation for now and simply stick to its current guidance, possibly moving it up modestly.
- source Deutsche Bank

It is most likely that the Fed's hiking process was due to its fear of not being behind the curve when it comes to rising inflation. Yet with a yield curve flattening and loose financial conditions in conjunction with renewed fear of a trade war that would entail pricing pressure and imported inflation with a bear market in the US dollar, there is indeed a big risk in having the Fed having to move at a more rapid pace than it would like to. The balancing act of the Fed is incredibly difficult but, it boast a first mover advantage other the likes of the ECB and the Bank of Japan. Volatility might have been repressed but in all honesty, it is in Europe where the repression has been the most acute as it can be seen in government bond yields.

The big question surrounding the potential lethality of the "Big Bad Wolf" aka inflation lies in the velocity of inflation expectations. On that specific point, Deutsche Bank gives us additional food for thoughts in their lengthy note:
"Inflation and inflation expectations
How high will inflation go? If inflation expectations remain range bound, core PCE inflation will stay within its narrow band of 1-2.3% in which it has been for the last 23 years
■ Outside the Great Inflation of 1968-1995, core PCE inflation has remained in a remarkably narrow band (Six Myths About Inflation, Oct 2017). The period since 1996 encompassed 3 business cycles that saw unemployment fluctuate between 3.8% and 10%; the dollar rise and fall by 40% more than once; oil prices rise 7-fold and almost completely reverse. Yet inflation remained in a narrow band unusual for an economic time series. Indeed, with a standard deviation of 35 bps, much of the range of variation in inflation since 1996 cannot be differentiated from the normal noise inherent in macro data.
■ The stability of inflation across large business- dollar- and oil-cycles in our view reflects the stability of inflation expectations which are the only driver of inflation over the long run. Inflation expectations have been stable since the mid-1990s, fluctuating for most of the last 23 years in a tight 50bps range and for most of it in an even narrower 30bps range. Following the dollar and oil shocks of 2014-2015, inflation expectations fell out of and are still 20bps below this range and 50bps below average. Absent large unexpected and persistent shocks, inflation expectations evolve slowly. It has in fact been difficult for policy makers to effect changes in inflation expectations as the recent experience of Japan and the 10-year miss on the core PCE inflation target in the US illustrate (Six Myths About Inflation, Oct 2017).
- source Deutsche Bank

As long as growth and inflation doesn't run not too hot, the goldilocks environment could continue to hold for some months provided, as we mentioned above there is no exogenous factor from a geopolitical point of view coming into play which would trigger an acceleration in oil prices. Though, in similar fashion to volatility, the game can continue to be played provided "implicit inflation" or "inflation expectations" remain below "realized" inflation. In similar fashion to the demise of the short-vol trade, if there is a change in the 23 years narrative and suddenly "realized" inflation is above "expectations" then obviously this would be another grain of sand that could trigger some new avalanches in financial markets. We are not there yet we think.

Finally in our final chart below, given the late stage of the credit game, we think it is becoming essential to track any changes in credit availability in the months ahead given our loose financial conditions have been and the flattening of the US yield curve.

  • Final chart - Afraid of buckling? Watch credit availability
We have long posited that "Credit availability" is essential and a good predictor of upcoming defaults as far as US High Yield is concerned. 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 quarterly Senior Loan Officer Opinion Surveys (SLOOs) published by the Fed are very important to track. 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. Tightening in credit standards in conjunction with rate hikes will eventually weight on High Yield, and we are already seeing some fund outflows in the asset class (15th consecutive week). Our final chart comes from CITI Global Economics View note from the 23rd of February entitled "How Could Equity Sell-offs Affect Global Growth" and displays US Non-financial corporations Debt Outstanding as a percentage of GDP and AAA-BBB Effective Yield Spread for Industrial Corporate Bonds (1997-2017):
"What to watch?Given that a tightening in financial conditions poses a risk to the outlook, we would monitor:
  • The durability of the sell-off: that’s rather obvious – a brief period of financial tightening is unlikely to have any material implications on the real economy.
  • Credit availability and credit spreads: given the stage of the business cycle, prospects for higher inflation, and lower monetary accommodation in advanced economies, we think credit availability and credit spreads amid high leverage across some sectors and economies are key indicators to assess whether financial conditions are starting to feed through to economic activity (Figure 6).

  • Sentiment measures: measures of household and business sentiment are at very high levels across most AEs. A decline in sentiment would probably be a precursor to some moderation in spending intentions, even though the relationship between consumer sentiment and real consumption appears to have declined in recent years." - source CITI
If further loading of the credit mouse trap will eventually cause significant and somewhat unpredictable deformations, possibly leading to complete loss of the member's load-carrying capacity, low recoveries and significant losses for credit investors, when it comes to assessing a potential "buckling" in the credit markets, apart from the "Big Bad Wolf" aka inflation being the enemy of volatility and leverage, credit availability is an essential part of the credit cycle.

"Prepare for the unknown by studying how others in the past have coped with the unforeseeable and the unpredictable." - General George S. Patton

Stay tuned ! 

Sunday 18 February 2018

Macro and Credit - Structured Criticality

"Bright light is injurious to those who see nothing." -  Prudentius

Looking at the relief rally that followed the tragedy that followed the "first to default" wipe-out of large swaths of the short volatility complex and given that we think that a large part of the continuation in the rally in equities is supported by the $171 billion in YTD stock buyback announcements, when it came to selecting our title analogy we reminded ourselves of "Structured Criticality" which is a property of complex systems such as financial markets. In complex systems such as financial markets, a small event may trigger larger events due to subtle interdependencies between elements. In our previous conversation we mentioned the pile of sand analogy with the additional grain of sand that triggers the avalanche as per the demise of the "first to default" or equity tranche short-volatility complex in the capital structure (or cone shape) of financial markets. Though the pile has retained its shape following the avalanche which has caused some number of grains to slide down the side of the cone (short volatility funds blowing up), it is nearly impossible to predict if the next grain of sand will cause an avalanche and where this avalanche will occur on the pile and how many grains of sand will be involved (risk parity, vol control products?). However, the aggregate behavior of avalanches can be modeled statistically with some accuracy. For example, some can reasonably predict the frequency of avalanche events of different sizes. The avalanches are caused when the impact of a new grain of sand is sufficient to dislodge some group of sand grains. If that group is dislodged then its motion may be sufficient to cause a cascade failure in some neighboring groups, while other groups that are nearby may be strong enough to absorb the energy of the event without being disturbed. Each group of sand grains can be thought of as a sub-system with its own state, and each sub-system can be made up of other sub-systems, and so on. In this way you can imagine the sand pile (or financial markets) as a complex system made up of sub-systems ultimately made up of individual grains of sand (yet another sub-system). Each of these sub-systems is more or less likely to suffer a cascade failure. Those that are likely to fail and reorganize can be said to be in a critical state. Put another way, the likelihood that any particular sub-system will fail (or experience a particular event) can be called its criticality. So then, the pile of sand (or financial markets) can be viewed as a network of interconnected systems, each with its own criticality. The relationships between these groups impose a structure on this network which has a profound effect on the probability and scope of a cascade failure in response to some other event. In other words - structured criticality. Given most buybacks have been funded by debt, we wonder when the next grain of sand will trigger the next avalanche. For now the complex system is benefiting from an unhealthy support coming from the flurry of buybacks announcements we think so caveat emptor ("let the buyer beware") with U.S. stocks recording the strongest weekly performance since at least 2013.

In this week's conversation, we would like to look at how volatility is the enemy of leverage and the on-going repricing of financial markets including volatility forcing markets to re-adjust to a loosening of financial repression and what it entails. There are as well many young market practitioners today that have never traded through a rising rates environment or seen what renewed inflationary pressure means for risky asset prices. 


Synopsis:
  • Macro and Credit - Volatility is the enemy of leverage
  • Final charts - US Dollar ? Twin deficits and inflation matter

  • Macro and Credit - Volatility is the enemy of leverage

As we pointed out in our previous conversation "Harmonic tremor", the regime change in volatility and the effect of "Who's Afraid of the Big Bad Wolf?" aka "inflation expectations" thanks to rising wage inflation expectations have already claimed the small fishes such as some players in the short volatility leveraged and crowded complex. Leverage and rising positive correlations not only reduces the benefit from diversification but the jump in global risk premiums meant that the sell-off episode has shown us that this time was indeed different in the sense that what could be seen as "antifragile" havens in a true Taleb fashion such as US long bonds, gold and Swiss franc did not played their defensive purposes, only cash mattered, or having had sufficient downward protection strategies in this small avalanche that clearly put into the limelight the brewing instability in market structures. 

Whereas recently the markets have rebounded significantly thanks to the impressive support from additional buyback announcements, one should clearly be wised in  trying to understand the "Structured Criticality"  and vulnerabilities which have been highlighted by the VIX episode. The anomaly was obvious to many, namely that financial repression has led to volatility being repressed beyond anything reasonable thanks to central banking intervention. Repricing was way due for a reality check and of course as one might correctly opine, volatility is always the enemy of leverage (ask LTCM). It should not come as surprise therefore with the return of volatility to a more normal stage to see Global Macro Hedge Funds staging a comeback. As we pointed out in our November 2012 conversation "Why have Global Macro Hedge Funds underperformed", the main culprit was the lack of volatility. 

Obviously the biggest question following the "repricing" of volatility to a more "normal" state after many years of "financial repression" led by central banks is the risk in the change in the narrative we warned about in so many conversations. This is leading to unpredictability making a return into what have been "predictable" markets for so many years. On this subject we read with interest Deutsche Bank's Special Report from the 16th of February entitled " Undoing the unstrange  - The problem of re-emancipation of the markets" which we think is a great illustration of the change in the narrative we are seeing first hand:
"After years of calm and predictable markets, suddenly there seems to be many things going on at the same time. As recently as early January, the incoming vol supply could not find a buyer as vol selling and carry trade remained the dominant themes. This changed practically overnight as rates broke through significant technical levels, which triggered a spike in gamma, which quickly spread across all market sectors. With every new installment of stimulus unwind, it seems as if things are moving in reverse, but not to where we left them, rather towards what appears to be an unknown and unfamiliar destination. This is proving to be a highly unconventional tightening cycle and recovery. After years of forced hibernation, brought about by suspension of traditional trading rules by the central banks, the markets are facing a painful process of re-emancipation. This is causing considerable confusion and anxiety. Last time we saw a recovery from a conventional recession was about 14 years ago (for many, this is longer than their entire professional career). Things are different this time. Both the 2008 financial crisis and subsequent policy response were highly unconventional, and therefore there is no reason to expect that recovery and unwind of the policy response should be conventional either. We believe that the following three observations summarize the ongoing complications associated with stimulus unwind and the conflicts they create in the context of economic recovery.
1) Unwind of stimulus is a mirror image of the QE trade. It is a de-risking mode and, as such, it goes against the grain of recovery. This is in sharp contrast with conventional unwind of the recession trade, which is the risk-on mode.
2) Risk is asymmetrically distributed between rates and risk assets. There are two distinct paths to higher rates (through higher real rates or wider breakevens). They mean two different things for bonds and stocks. For bonds, the distinction between these two paths is a matter of degree between a mild and a moderate selloff. For equities (and USD), on the other hand, the effect is binary – it means a difference between a modest rally and a substantial selloff.
3) Volatility plays an essential role in the policy unwind. It is one of the key decision variables in the process of portfolio risk rebalancing -- higher volatility causes complications. However, unlike traditional recoveries, which are collinear with the unwind of the recession trade -- and, as such, volatility-reducing – the unwind of financial repression is withdrawal of convexity supply and a vol-enhancing mode.
The main diagonal: Conventional recovery from a conventional recession
To visualize the problem, we start with a figure that illustrates how recoveries from conventional recession used to play out in terms of the interplay between yields and equities. We start at point 1: Recession typically begins with a steep decline in risk assets and allocation to bonds. Monetary policy intervenes with rate cuts, which slows the selloff in risk, with rate cuts continuing until the economy stabilizes and the market turns around (2).
The recession-recovery path in the figure moves along the main diagonal (between the 1st and 3rd quadrants) -- recovery is a mirror image of the recession. As the defensive position (long bonds/short risk) is rebalanced, it moves the market naturally into the risk-on region (3) with more aggressive allocation to risk assets and underweight in bonds continuing typically until rate hikes slow the rise in equities (4). Unwind of the recession trade (in the conventional setting) goes along the grain of the market trade – its inertia leads naturally into the recovery trade. Because of this, past recoveries have been generally accompanied with lower volatility.
The agony of the off-diagonal: Rise of the unconditional
The current policy unwind is qualitatively different from traditional recoveries. The underlying complications can be traced back to the later installments of QE, around 2011, which signaled the beginning of a new regime of market functioning, an utterly new mode rarely seen to persist beyond transient episodes. The figure illustrates a longer history of the three assets in question, USD (in terms of TWI index), S&P levels and 10Y UST yield, indexed to their Jan-200 levels.
The letters S and W stand for strong and weak. Typically, stocks, bonds and currency cannot all rally at the same time for a prolonged period of time. Generally, they support each other conditionally: For two of them to rally, one has to sell off (and the other way around). This is seen in the picture during first decade of this century. 2011 signals a structural shift to a new regime: Between 2011 and 2016, the three assets supported each other unconditionally – they rallied simultaneously. This was a result of continued QE against the background of threat of sovereign risk overseas, which created positive externalities for both USD and US stocks, and it represents the other side of the state of exception created by the extended influence of central banks.
This outlines the essence of the problem of policy unwind. While central banks actions and the market environment had clearly created optimal conditions where, for many years, every asset class made money at the same time, the natural question one had to ask is: What to expect after that? If unwind of the stimulus is its mirror image, where does one go when everything sells off?
Monetary policy pharmakon of why does it hurt when we unwind?
The figure below illustrates the recession-recovery path post-2008. It starts, as usual, with a selloff in risk assets and a rally in bonds (1 & 2), but as the crisis deepens and QE gets deployed (3), the action moves (and stays) on the off-diagonal where both bonds and equities rally. Unwind of QE now becomes essentially a de-risking move -- it goes against the grain of recovery.
Currently, we are heading towards point 4, beginning to catch sight of the bifurcation point (5) from which the market could either sink into the “stagflationary” trap (6: everything: stocks bonds and currency, sell off) or move to the 1st quadrant if the Fed and Congress manage to engineer a turnaround and we get catapulted towards what looks like a traditional recovery. This is the biggest challenge for the Fed at the moment, which is further complicated by the ongoing rise in volatility. This complication, which appears to come naturally in this context, is further amplified by the Fed’s negative convexity exposure to inflation.
Inflation is producing an Icarus effect: Although negative convexity of inflation is a far OTM risk, it is significant even at remote distances from the strike, due to its enormous size. The accumulation of relatively illiquid long-dated bonds on retail balance sheets is at toxic levels and a substantial rise in inflation, to which there is no adequate policy response, could threaten to trigger a bond unwind that the market would be unable to absorb.
Locally, the main problem for risk assets is a rise in real rates: Having UST bonds with strong dollar or high real yields will be more attractive than holding US stocks, which means accelerated de-risking and higher volatility in the stock market. Higher inflation, on the other hand, would be supportive for equities and could cause another leg of selloff in bonds. What complicates things is that the behavior of real rates at this point is also a function of expected inflation: Higher inflation warrants a more hawkish Fed and therefore pricing in higher real rates. The reaction of stocks is a non-linear function of inflation – although risk assets might “like” higher inflation, this would remain true only up to a certain point.
Unwind of financial repression: Volatility is the key variable
Traditional recoveries have not been very sensitive to volatility behavior. In fact, volatility showed a tendency to decline in those cases. This follows almost automatically because of collinearity of recession unwind with the risk-on trade. The role of volatility in the current context is a novelty. An exit from almost a decade of financial repression has another dimension defined by volatility. This is a consequence of both the nature and the duration of the stimulus and subsequent  addiction liability that central banks run at the moment. The subsequent three figures illustrate how volatility enters the play during different stages of stimulus and its withdrawal.
Step 1: The recession starts at elevated volatility levels. The solid line represents the efficient frontier of a portfolio on the risk-return plane. Changing the risk causes a repricing of the frontier. This is shown by the dashed lines which reflect the levels of the existing market volatility. The two-sided arrow represents the risk limits of a given portfolio. This is kept constant through different stages of rebalancing. For a given risk limit, one finds a place on the frontier that fits inside the dashed lines (“VaR limits”).

Step 2: Response to crisis through QE consists of constraining the rates at the long end and therefore reducing the market volatility. As volatility resets lower, investors can afford to move further out along the risk curve until their risk limits are compatible with new volatility levels. This is the asset misallocation trade (one does things that one regrets later). This persists for years after the initial decline of volatility from crisis levels in late 2009. The new position is shown with the red double-sided arrow (the initial one is shaded).

Step 3: Unwind of stimulus and Fed exit is also a withdrawal of convexity supply. This implies higher volatility, which means that the prior portfolio is now operating above the risk limits. As a consequence we have a risk rebalancing towards the left (point 3 or the green arrow).

In the subsequent months, a particular pattern of volatility, in terms of its breakdown across different assets, will determine the mode of risk rebalancing. In that context, volatility will play a decisive role in determining the success and timing of the recovery and a particular economic trajectory.
Trades
Money market repricing
Inflation or no inflation in the short run, with continued push towards easy fiscal policy, financial conditions are unlikely to tighten. In that context, inflation risk could become more acute than currently perceived. At least, this is what history  would suggest. When markets operate close to full employment, further easing of financial conditions could create an explosive response in the economy. In that environment, the Fed is likely to stay the course and continue to hike, especially in light of the realization of the actual threat of inflation getting out of hand. In the meantime, the question is more about the Fed path rather than about its stance. A possibility of frontloading some of the hikes implies a flattening between the red and green sectors of the money market curve. This mode is not yet being priced in by the curve and vol. We are buyers of conditional bear flatteners at the short end of the curve." - source Deutsche Bank

As we pointed out in our previous conversation, there lies the risk ahead for financial markets when it comes to "inflation expectations", "realized inflation" could prove to be a significant grain of sand in terms of "Structured Criticality" particularly with a potential acceleration in trade wars and geopolitical exogenous factors coming into play (both are bullish gold by the way):
"Believing that the spread between implied and realized volatility would persist has indeed been a dangerous proposal with rising positive correlations. In similar fashion believing that "implied inflation" could persist remaining below "realized inflation" could become hazardous in the coming months, particularly with growing geopolitical exogenous risks around. Whereas QE was deflationary, QT could prove to be inflationary but we ramble again..." - source Macronomics, February 2018
Also, what we re-iterated in our conversation "Who's Afraid of the Big Bad Wolf?", with inflation, the only issue is when the "Inflation Genie" is "Out of the Bottle" as warned by Fed's Bullard in 2012, it is hard to get it back under control:
“There’s some risk that you lock in this policy for too long a period,” he stated.  ”Once inflation gets out of control, it takes a long, long time to fix it”
As pointed out by Christopher R. Cole, CFA from Artemis Capital Management latest note entitled "Volatility and the Alchemy of Risk - Reflexivity in the Shadows of Black Monday 1987",  the rise of the Big Bad Wolf aka inflation was what started a liquidity fire in credit that spread to equities before the 1987 volatility explosion. As we pointed out in our recent musings, when it comes to "Structured Criticality", for a "bear market" to materialize, you would indeed need a return of the Big Bad Wolf aka "inflation", being probably one of the most dangerous grain of sand around when it comes to "avalanches" in the conic structure of financial markets we think. If volatility is the enemy of leverage, then again, inflation is the enemy of volatility. 

If indeed as pointed out by Christopher Cole, volatility is the brother of credit and volatility regime shifts are driven by the credit cycle, we have yet to see in earnest a significant tightening in financial conditions, yet from the buybacks frenzy to the current M&A craze, everything points towards a late credit cycle in our playbook. Yet when it comes to pressure on credit spreads, as seen during the energy crisis with the fall in oil prices leading to the blow-out in credit spreads, things can turn "south" as for the short-vol sellers faster than a rat on roller skates. What we think is of interest is that finally the much vaunted "Great Rotation" by some sell-side pundits, has finally somewhat started to materialize slowly in terms of fund outflows but this time where all the "fun" has been running thanks to low volatility and low interest rates, namely in the bond markets thanks to "goldilocks" environment enabling the "beta game" for the carry tourists. Fund outflows also point towards "Structured Criticality" in the sense that the shape of the conic structure in credit has been heavily skewed in recent years by the significant inflows into corporate bonds including the ETF complex. On the subject of fixed income outflows and the growing importance of the ETF complex we read with interest Bank of America Merrill Lynch's take from their Situation Room note from the 15th of February entitled "Position reduction":
"Following the recent equity market correction and equity and rates vol spike, investors reduced positioning in risk assets across the board this past week ending on February 14th. Outflows from equities continued for a second week at $3.55bn from $29.54bn. US-domiciled high grade bond funds and ETFs reported the first weekly outflow since December 2016 at $1.52bn, following a $5.28bn inflow the week before. HG funds had an outflow of $0.28bn after an inflow of $4.54bn, and HG ETFs had a $1.25bn outflow – highest since June 2013 – following a $0.74bn inflow one week earlier. Short-term HG held up comparatively well with an inflow of $0.12bn, down from $2.15bn the week before, while HG outside-of-short-term lost $1.65bn after gaining $3.13bn the prior week.

High yield also experienced a flows exodus of $6.33bn – the second highest weekly outflow on record – after a $2.34bn outflow the prior week, with HY funds and ETFs losing $3.58bn and $2.75bn in redemptions (-$1.37bn and -$0.97bn one week ago), respectively. Leveraged loans also had an outflow of $0.27bn from an inflow of $0.50bn the week before. Global EM reported an outflow of $2.87bn following an almost flat prior week of $0.02bn inflow. Outflow from munis accelerated to $0.66bn from $0.47bn, while inflow to money market funds slowed to $0.02bn from $27.80bn. Mortgages experienced a $0.18bn outflow following a $0.08bn inflow one week ago. On the other hand, government bonds continued to report decent inflows at $1.73bn this past week following a $1.75bn inflow the week before. The net effect on the all fixed income category was a significant $8.21bn outflow from a $4.02bn inflow a week earlier.

IG ETFs vs. bond funds
ETFs are becoming increasingly important vehicles in fixed income and inside we provide a discussion of trading volumes relative to the IG corporate bond market. Today we fielded a number of questions about yesterday’s record ~$924mn outflow from the largest IG corporate bond ETF (LQD) and whether we are concerned about it. We are not as, while the importance of ETFs in IG credit is growing, they are still relatively small. About 20% of US corporate bonds (IG+HY) are held by bond funds and ETFs (Figure 13), which applied to the size of the index eligible IG market comes out to $1.27tr.

However, we estimate that ETFs hold only $190bn of IG corporate bonds, or 2.9% of the market (Figure 16). Hence bond funds – not ETFs – are the elephant in the room as they hold more than six times as many IG corporate bond assets relative to ETFs. Even with the more recent shift to passive investment (see piece below) inflows to HG bond funds were four times ETF inflows in 2017.

The particular ETF in question (LQD) had about $34bn of assets – or 0.5% of the size of the IG market - before suffering a 2.6% outflow, which is a drop in the bucket. This ETF has suffered outflows all year totaling about $4.7bn as bond prices declined (entirely due to higher interest rates as credit spreads are flat on the year), which is normal (Figure 11).

However, we estimate that high grade bond funds and ETFs overall (a category that includes LQD) have seen inflows of $47bn this year. Hence the big story is one of very large inflows as opposed to ETF outflows. Now most IG bond funds/ETFs buy other IG assets in addition to corporate bonds - such as Treasuries, mortgages, etc. Focusing on dedicated corporate bond IG funds/ETFs (again including LQD) we estimate a $1bn inflow this year.
Recent daily outflows from HG bond funds/ETFs
However, we are starting to see small daily outflows from high grade bond funds and ETFs recently – specifically Friday-Wednesday (Figure 14).

This is to be expected given that the three main drivers of inflows to high grade bond funds/ETFs are 1) good total return performance (instead IG corporate bonds have lost 2.74% so far this year), low interest rate vol (Instead the move index has jumped to 70bps from 47bps) and equity outperformance (instead stocks corrected recently). For more details see: Inflows to taper 26 January 2018. For us to be concerned about large overall HG outflows – i.e. from bond funds as well - we need to see a much bigger increase in interest rates.
ETF liquidity injection
Fixed income ETFs are getting increasingly popular and, as a result, are adding liquidity to the mostly illiquid corporate bond market. In particular dedicated IG corporate bond ETF trading volumes are about 5.6% of cash bond trading volumes LTM – on adding the corporate bond portion of fixed income ETFs with broader mandates – such as agg-type funds - that number increases to 7.5% (Figure 15).

Trading activity in IG corporate bond ETFs is highly concentrated with the largest fund accounting for about 60% of volumes (Figure 17).

Trading volumes for the most active bonds in the corporate bond market are comparable, although slightly lower (Figure 18). In terms of AUM ETFs rose from 0.9% of the high grade index market value in January 2010 to 2.9% currently (for both corporate and high grade bond ETFs, adjusting for the share of corp. bonds, Figure 16)." - source Bank of America Merrill Lynch
While the slow movement in outflows has not reached the "Structured Criticality" level that would mean another "avalanche, these grains of sand do start to add up. Whereas foreign investors were responsible for the big acceleration in HG bond fund/ETF inflows in recent years thanks to a big decline in the cost of dollar hedging, retail in many instances have taken over from these foreigners particularly in the High Yield ETF space, rendering them more prone to volatility thanks to the feeble nature of these investors. While tracking bonds ETFs is of interest, it is of course not the best great gauge of real health in credit markets we must confess, though from a short term perspective, it might indicate some weakness in the near term. The correlation between oil prices and High Yield is much more interesting from a "monitoring" perspective. What you should be concerned about is that the switch from a negative real yield regime to a more normal, positive real yield regime might spark a big non-financial credit crisis because this time around leverage is higher now compared to history. If you believe in a "stagflationary" scenario unfolding à la 70s, the major difference is that leverage was falling during the rapid credit cycles of the 70s, with the biggest spikes in yields taking place at the end of the period.  There also a phenomenon that needs to be taken into account and it is that the current Boomers are more leveraged than previous generations were ahead of retirement as per the final points we have shown in our March 2017 conversation entitled "The Endless Summer". We concluded our missive at the time asking ourselves how many hikes it would take before the Fed finally breaks something.

But before your worries get ahead of you, in terms of credit matters, from an allocation perspective, if indeed slowly but surely rising outflows pressure from the Fixed Income space, we got interest by the suggestion made by Deutsche Bank in their Credit Bites note from the 16th of February entitled "The Resilience of Loans":
"In the aftermath of the recent inflation induced spike in volatility we analyse the impact it has had on the relative performance of HY bonds and leveraged loans. One of our key relative value views in our 2018 outlook is that loans would fare better than bonds if we did indeed see an inflation/rising yields led move higher in volatility that puts pressure on credit spreads.
When we published our outlook back in November one of our key relative value views was that loans would fare better than bonds if we did indeed see an inflation/rising yields led move higher in volatility that puts pressure on credit spreads. Given recent events we thought it would be worthwhile taking stock of where we stand and how the recent bout of volatility has impacted the relative returns between loans and bonds.
In Figure 1 we look at the cumulative YTD returns for the HY bond and leveraged loan indices.

We can see that in the early weeks of the year with spreads generally trending sideways to tighter bonds had fared fairly well. However with Bund yields generally moving higher from the second week of January loan returns started to bridge the performance gap. Then the inflation induced spike in volatility pushed credit spreads wider and helped to accelerate this trend. At the time of publication loans have outperformed bonds by 1-1.5% across the rating bands as we can see in the right hand chart of Figure 1.
 
In Figure 2 we run the same analysis for the USD market. We can see the relative performance dynamics are very similar to what we have already shown for the EUR market.

After the initial spread tightening and associated outperformance of bonds, the combination of higher bond yields and the spike higher in volatility has seen loans notably outperform. In fact the level of outperformance is slightly more impressive in the USD market. At the time of publishing loans had outperformed bonds (at an index level) in the 1.5-2% range across the rating bands (right hand chart of Figure 2).

We would additionally argue that it is not just the obvious outperformance of loans that has been impressive but also the general stability of loan returns. This highlights a key factor in why we think loans will outperform this year as they are generally less susceptible to day to day market volatility as well as having negligible exposure to rates duration.
If spread weakness in 2018 is driven by macro factors such as higher inflation and rising bond yields leading to higher volatility and wider spreads then the recent trend in performance makes us more comfortable with the view that loans will outperform bonds this year. We would be more concerned about this view if spread widening were to be driven by fundamental credit factors that pushed us towards the next default cycle.
Near-term we might see some reversal of this loan outperformance if volatility continues to settle down, equities continue to rebound from the recent correction and credit spreads continue to reverse some of the recent widening. However over the medium term we expect higher inflation and yields to keep volatility elevated above the lows of 2017 and therefore credit spreads to maintain a widening bias which should benefit loans over bonds." - source Deutsche Bank
What we don't like right now in the Leveraged Loans market is that Lower-rated deals (and covenant-lite transactions) are driving it at the moment. As indicated by S&P Leveraged Loans:
"There's $970B of outstanding US Leveraged Loans and more than 75% of that is covenant-lite" - source S&P LCD News
It might be more appropriate from a defensive perspective to play the Leveraged Loans game through large "Senior Tranches" in CLOs, ensuring you have a high attachment point, should defaults make a return at some point. Yet no doubt the low volatility of the asset class is compelling. Also in the US, managers of open-market CLOs have received a waiver from retention risk from the part of the US Court of Appeals for the DC circuit recently. This decision opens the door to other markets such as RMBS, CMBS and ABS to issue with the new rule in place. With a slower pace of issuance taking place over the next few months following the ruling, the asset class could benefit from a "technical bid". 

While many continue to be puzzled by the weakness in the US Dollar as per our final charts, we do think that when it comes to the long term direction of the currency, the twin deficits matter, and matter a lot.

  • Final charts - US Dollar ? Twin deficits and inflation matter
Economies that have both a fiscal deficit and a current account deficit are often referred to as having "twin deficits." The United States has fallen firmly into this category for years. According to Nomura FX Insights report from the 16th of February entitled "Twin deficits + inflation = weak dollar, current account balances are good explainers for FX performance. Both the Twin deficits in the US in conjunction with rising inflation expectations are good reasons to put forward for the weakness in the US dollar according to their report:
"USD/JPY’s plunge to levels last seen in late 2016 has caught many by surprise, but it fits neatly into a dollar downtrend narrative. Indeed, EUR/JPY has broadly been in a range since September last year, suggesting that we are not seeing a yen- or euro-specific move, but rather a dollar move. Remember, the euro is also seeing new highs – it has recently touched its highest level since late 2014.
We wrote recently that growing twin deficits in the US typically see the correlation between yields and the dollar breakdown and also that the dollar fares poorly during actual hiking phases. Another way of looking at this is correlations of G10 FX performance against current accounts or shifts in interest rates. Here we find that current account balances have asserted themselves as the best explainer of relative FX performance just as monetary policy has lost its grip on markets (Figure 1).


As for inflation, we also have written that the current combination of higher US inflation and loss of momentum in US growth surprises should weigh on the dollar. This has panned out. Again looking at correlations across a range of indicators, we find that FX is now negatively correlated with inflation levels.
In a world where twin deficits and inflation matter, the yen stands out. Japan has the lowest expected inflation in the G10 world. Core inflation is currently an anemic 0.1% compared with the recent 1.8% in the US. Japan is running a sizeable current account surplus and its fiscal balance is improving. Meanwhile, the US’s trade deficit is widening fast and the US is set to see its worst deterioration in its fiscal balance outside of a recession in modern history. All this suggests that dollar weakness could continue. We need to monitor the pace of the move, and Fed actions (more tightening to slow the economy) or even BoJ/ECB actions, but for now we’d look for USD/JPY to breach 100 and the euro to breach 1.30 in coming months." - source Nomura
In addition to the above interesting points made by Nomura, if the US dollar tends to weaken when inflation worsen, it also tends to strengthen when oil prices fall. When it comes to "Structured Criticality", the rapid fall in oil prices in 2015 was the grain of sand that led to the "avalanche" in risk-off and the significant widening in credit spreads that led to the weakness seen in equities in early 2016. Right now, the market has regained some posture thanks to financial engineering in the form of renewed buybacks and a strong M&A pipeline, until we get another unforeseen grain of sand, but that's a story for another day it seems.

"The epitaph on the grave of our democracy would be: They sacrificed the long-term for the short-term, and the long-term arrived" - Sir James Goldsmith

Stay tuned !  
 
View My Stats