Wednesday, 16 August 2017

Macro and Credit - Experiments in the Revival of Organisms

"Life is obstinate and clings closest where it is most hated." -  Mary Shelley

Looking at the latest gyrations in risky markets including credit markets following our timely musing relating to "Gullibility", while thinking of our next post title we found it amusing the numerous attempts by various financial pundits in "reviving" the Phillips curve which we have been so critical of in recent conversations. For our title analogy we decided to steer towards the creepy given our title analogy refers to a 1940 Soviet motion picture which document Soviet research into the resuscitation of clinically dead organisms. It is available from the Prelinger Archives and is in the public domain. The creepy operations depicted are credited to Doctor Sergei Brukhonenko and Boris Levinskovsky who were demonstrating a special heart-lung apparatus called the autojektor (or autojector), also referred to as the heart-lung machine, to the Second Congress of Russian Pathologists in Moscow. Their modern "Frankenstein" experiments (involving decapitations) were related to the heart-lung machine. It was designed and constructed by Brukhonenko, whose work in the video is said to have led to the first operations on heart valves. The autojektor device demonstrated in the film is similar to modern ECMO machines, as well as the systems commonly used for renal dialysis in modern nephrology. The film depicts and discusses a series of medical experiments. It begins with British scientist J. B. S. Haldane appearing and discussing how he has personally seen the procedures carried out in the film and have saved lives during the war. The experiments start with a heart of canine, which is shown being isolated from a body; four tubes connected are then connected to the organ. 

You are probably asking yourselves already where we going with this creepy analogy of ours (after all being outright "scary" drives traffic for some blog pundits...), but in these Frankenstein markets to say the least, we found it amusing the attempts by so many including the PhDs are the Fed to cling on outdated models that were fit to purpose in their own time but not anymore. As pointed out by our friend Ilya Kislitskiy (@sayfuji on Twitter), resurrection is a complicated business, same thing goes with "Experiments in the Revival of Organisms":
"What can cause global anxiety? Maybe the fact that Phillips curve somewhere is "complicated" and "partially resurrecting" elsewhere? I definitely feel uncomfortable." - Ilya Kislitskiy (@sayfuji on Twitter)
Rest assured we do too, when it comes with macro old-school Phillips-curve-fans. Life is indeed obstinate, to paraphrase Mary Shelley, the author of Frankenstein. This is particularly the case with the Fed and the financial community when it comes to clinging to outdated models. It doesn't necessarily mean that these pundits' Phillips curve "autojektor device" will not eventually lead to a better model. In our book, for the time being, experiments in the revival of macroeconomic models such as the Phillips curve, defeat the purpose but we ramble again...

Also, when it comes to our analogy, Frankenstein markets comes to mind given Bondzilla, the NIRP monster has indeed become insatiable for anything with a yield. In similar fashion to Dr Frankenstein, our generous gamblers aka central bankers have become increasingly nervous with the "creatures" (or bubbles) they have spawned with their various resuscitation experiments and this is even without the exogenous geopolitical turmoil as of late. In recent years they have been doing various experiments in the revival of organisms, some European banks come to mind when we think about "zombies" in homage to recently departed horror film maverick director George A Romero

In this week's conversation, we would like to look at the need to continue building up your defenses in the credit space, meaning further rotation into higher quality in this overextended beta game given the accumulation of late cycle signs we are seeing.

  • Macro and Credit - Frankenstein markets thanks to our Modern Prometheus
  • Final charts - Fun in Funds

  • Macro and Credit - Frankenstein markets thanks to our Modern Prometheus
When one looks at a 70% rise in the VIX index over just three days, a 2% drop in global equities in conjunction with widening credit spreads including of course High Yield and the beta crowd, a young investor would have caught fright. But, in retrospect, this was merely a blip given our Dr Frankensteins are still busy with their central banking experiments. Obviously the big question is surrounding Bondzilla the NIRP monster, given the massive creature has been in growing in size since the start of our central bankers various iterations.

What is already starting to show credit weakness, we think is indeed coming initially from US Commercial Real Estate (CRE). Of course as we pointed out in our most recent conversation, consumer credit is another piece of the credit puzzle we watch carefully as this credit cycle unfolds and is indicative of the lateness of it.  Back in February in our conversation "Pareidolia" we pointed out the following:
"At this stage of the cycle, there is a tendency for excesses (real estate prices, subprime auto loans, etc.) to build up meaningfully. For instance, we have been monitoring the weakening demand for credit but in particular Commercial Real Estate given the latest Federal Reserve Senior Loan Officer Survey has shown that financial conditions have already started to tighten meaningfully in that space" - source Macronomics, February 2017
CRE (Commercial Real Estate) is considered to be a good proxy for the state of the economy. And, if indeed investors are pondering the likelihood that the US economic growth is slowing and that CRE valuations have gone way ahead of fundamentals, then it makes sense to track what is going on in that space for various reasons, particularly when it comes to assessing lending growth and the state of the credit cycle we think. By tracking the quarterly Senior Loan Officers Surveys (SLOOs) published by the Fed you can have a good view into credit conditions. Credit conditions for CRE have already started to tighten since a couple of quarters and from a valuation point of view, it does feel that we are close in many instances to "peak valuation". From a valuation perspective we read with interest Wells Fargo's note from the 14th of August entitled "CRE Deal Volume Shows Late-cycle behavior Q2 Chartbook" particularly the part linked to the Hotel segment of CRE:
  • "Still reflecting the previous headwind of weak corporate profits and a stronger dollar, the seasonally adjusted real revenue per available room (RevPAR), which is the product of occupancy and the real average daily rate (ADR), fell to just 0.2 percent in Q2, year-over-year. That said, corporate profits were up more than 3 percent in Q1, and the dollar is softer suggesting there are some upside risks.

  • Real RevPAR for higher-end hotels (luxury, upper upscale and upscale) posted its fifth straight year-over-year negative reading in six quarters in Q2, while the pace for lower-end hotels (mid- and economy-scale), has slowed significantly from its cycle peak of 8.0 percent in late-2014 to just 0.8 percent.
- Source: STR, U.S. Department of Labor, FRB and Wells Fargo Securities

Are we at "Peak Occupancy" in this rate cycles? It certainly looks this way to us. Also the negative trend in Real RevPAR growth is yet another sign that the credit cycle is slowly but surely turning we think.

It is fairly clear to us that when it comes to credit availability and lending, the CRE space indicates things are indeed slowing as per Wells Fargo's remarks from their report:
  • "Senior loan officers reported that CRE lending standards tightened across the three categories while demand cooled during Q2. A net share of around 17 percent of banks, which is classified as “moderate,” reported tightening standards for construction and land development loans and multifamily.

  • According to the Mortgage Bankers Association (MBA), total commercial/multifamily debt outstanding broke the $3 trillion mark in Q1. At 41 percent, commercial banks and thrifts still comprise the largest share of debt outstanding.
  • Consistent with lending standards, loan growth in multifamily and construction and land development has slowed from its cycle high. Growth in income properties is also moderating.
- Source: STR, U.S. Department of Labor, FRB and Wells Fargo Securities

There are more signs from the CRE market alone that the credit cycle is slowly but surely turning and it is of course showing up in this market first.

When it comes to "Experiments in the Revival of Organisms", clearly the CRE market has reached record valuation levels as pointed out by Wells Fargo. Our Dr Frankensteins at the Fed should clearly be nervous about these lofty valuations levels we think:
"Elevated pricing (Figure 2) has also been a concern for investors, and we suspect the disconnect in pricing and transaction volume is also due to still-elevated levels of cross-border transactions.

Indeed, the low global rate environment and reach for yield, and in some cases, safe-haven plays made U.S. CRE a preferred asset class in many markets by foreign investors. With global economic conditions stabilizing, the risk of global deflation less of a concern, and hence, some major central banks expected to begin tightening monetary policy next year; investors are bracing for higher cap rates in the U.S. and abroad, and in turn, lower valuations. That said, the short and long-end of the curve are largely driven by different determinants. Moreover, cap rates don’t necessarily move in lock-step with the 10-year U.S. Treasury yield. On a global scale, the average cap rate in the U.S. in Q2 was the highest among the 11 countries tracked by RCA (e.g. France, Australia, etc.), with the risk premium still favorable at more than 400 basis points.
- Source: FDIC, Mortgage Bankers Association and Wells Fargo Securities

Sure valuation wise, as we pointed out in our most recent conversation "The Barnum effect", in these Frankenstein markets thanks to the modern Prometheus, already expensive asset classes such as European High Yield can become even more expensive thanks to central banking "Barnum effect". No doubt their Marshall amplifier can reach "11", as in the famous movie Spinal Tap. After all, as long as the Dr Frankensteins are in the game, you got to keep dancing right?

You are probably asking when the tide will be turning when it comes to the "Macro and Credit" picture. On this very subject we read with interest Barclays note from the 11th of August entitled "Macro Credit Framework: Let's be reasonable" where they update their framework for credit returns:
"With credit valuations near post-recession tights, we believe this is a good time to update our macroeconomic framework for credit returns. The core of our framework is that:
  • For the purposes of forecasting credit performance, we can reduce the business cycle into two regimes: a steady state in which the economy is growing consistently and a transitional state that includes recessions plus de 12 months before and after. Macroeconomic indicators can signal which state the economy is in.
  • Spreads evolve differently in the transitional state than in the steady state.
  • Valuations are the key driver of returns for both steady and transitional states, but the distributions shape of those returns varies across regimes. 
- In the steady state, returns are more normal for a given valuation, and treating returns as normal should be adequate for rough estimates of returns and loss probabilities.
- Transitional state distributions are more skewed, making the distribution of expected returns rely more heavily on starting valuations.
Applying the framework to the current environment suggests a 70% chance of positive excess returns for BBB credits, with an expected return near carry (Figure 1 ). 
The macro framework suggests that we should expect 60-70bp of Six-Month Excess Returns for Corporate BBBs

Defining Business Cycle Regimes
For the purposes of making return forecasts, we believe that only two regimes really matter:
  • Steady state is the default, covering periods of consistently positive economic growth that are the “expansion phase” of traditional business cycles. These phases have periods of both slower and faster growth, but without the sustained deterioration that marks a recession.
  • Transitional states have been less frequent (especially in the past 30 years) and consist of recessions plus the 12 months preceding and following them.
There are number of reasons to condense the macro environment to only two regimes:
  • There are clear differences in how credit returns behave in the two regimes with credit more likely to widen, and more likely to linger at wider spreads during transitional states (Figure 2).
    But further subdivision do not seem to add much information - for example, for a given valuation, there is not much difference in returns from "early" in a recession to "late" - so it is sensible to use fewer states.
  • It makes it easier to define the conditions of being in each state, which makes it more likely that we will make the correct call. In general, we think it is relatively straightforward to understand when we are in a recession or its aftermath that means the only difficult call is whether we have moved from a steady to a transitional state.
Economic Indicators - Clue for Transitional Periods
The most challenging part of our framework is determining whether we have entered the transitional state, but we think there are useful indicators for when that happens. There are a number of layers to the process. First, the preconditions need to be in place. Then the timely indicators give us information about whether we are within the 12 months of a recession. Finally, a qualitative evaluation informs us whether our preferred signals are likely to have their usual reliability.
Preconditions for a Recession
Fed hiking cycles and tighter bank lending standards have historically been preconditions for recessions. The past six recessions have been preceded by Fed hiking cycles and leading up to the past two recessions, bank lending standards have tightened (Figures 3 and 4).

Intuitively, this makes sense - the economic benefits of looser Fed policy and accelerating lending conditions lean against the possibility of a recession.
While these indicators are necessary for a recession, they do not by themselves signal that a recession is imminent - both can be in place for multiple years before a recession begins. While these preconditions are currently being met, we must look at more timely measures that signal the economy is entering a transitional state.
Near-Term Indicators 
Jobless Claims
We believe that jobless claims are one of the best indicators of a regime shift, because they generally start to rise about a year before the economy enters a recession. This makes them particularly well timed for the move to a transitional state. Their usefulness as an indicator for credit is supported by the tendency of rising jobless claims to predict negative future returns for credit (Figure 5), consistent with our analysis that spreads generally widen from tights during the transitional state.

The decline in jobless claims has been impressive since 2009, and the measure is at all-time lows after adjusting for total jobs in the economy. With less slack in the series, the decline in claims has lost momentum, suggesting that further improvements could be limited (Figure 6).

However, claims have remained low or flat for multiple years in the past, so flattening alone is not sufficient to indicate a rise in claims. Consequently, we do not believe that jobless claims are currently signaling that we are entering a transitional period (although that assessment could change quickly).
Output Gap
The second timely indicator that we use for our macro credit framework is the output gap - a measure that Barclays Economics uses as a framework for the US business cycle (Figure 7).

The output gap uses a multivariable approach - with inputs such as working hours, output, employment, unemployment, and the labor force - to measure the actual output of the US economy versus the potential output.
We find that the output gap tends to follow the business cycle closely. The indicator typically peaks about three quarters prior to a recession, on average, and starts to roll over during the transitional state (Figure 8).

Coincidentally, spreads also tend to widen when the output gap is declining (Figure 9).

Currently, the output gap has closed and is near the 2005-2006 peak, meaning that the current business cycle is mature. However, the indicator does not seem to be rolling over and is not indicating that we are entering a transitional state yet.
Qualitative factors
In relation to position within the business cycle and whether the economy is entering a transitional period, it also makes sense to monitor qualitative factors that could pose risks to the current steady state of the economy.
Consumer Credit
Trends in consumer credit have become a cause for concern. Consumer debt outstanding has risen for credit cards, auto loans, and student loans, giving consumers less room to use debt to support spending. Delinquency rates have also begun to rise in all three cohorts (Figure 10), with the auto loan sector particularly worrisome.

Along with the significant growth in auto credits outstanding, auto loan quality has worsened, and auto sales have lost momentum. While these trends are somewhat troublesome, the size of the auto loan market pales in comparison with the mortgage market, a major issue in the last recession. And despite some deterioration in the quality of the ABS market, losses so far have been modest. Therefore, we do not believe that concerns in the auto sector will push the economy into a transitional state at the moment, but any further development should be closely monitored.
Retail sector
Weakness in retail has been a theme in 2017, with the sector facing revenue losses to e-commerce competitors and lower returns on assets because of overcapacity. Because the retail sector is such a large employer - up to 10% of American workers are in the sector in some capacity - there is a reasonable question about whether a sector restructuring could spill into the broader economy. We examined this question in greater depth in US Economics and Credit Strategy: Technology-based change leaves retail looking overextended, and our conclusion is that so far jobs are not being lost at a fast enough pace to create exogenous recession risk. For example, the number of retail workers affected by bankruptcies has increased quickly, rising above 200k over the past 18 months (Figure 11), but given that the US sees about 250k new jobless filings a week (a record low), that is not yet enough to cause broader problems.

Returns are more volatile during transitional states 
The next component of our framework is to understand what the most reasonable distribution of returns is likely to be within each state. The first thing we observe is that returns are related to starting valuations in both regimes (Figure 12).

We also note that for a given valuation, they are more volatile, and more likely to be negative, in the transitional state.
We base this analysis on Moody's data on spreads for industrial BBB long bonds. The series starts in 1919, which allows us to calibrate spread performance around 17 recessions. This is a significant advantage over using the Bloomberg Barclays Indices, which cover only three recession cycles. The disadvantage is that the Moody's data do not provide carefully structured return calculations, so we need to estimate returns using carry, spread changes (assuming the same duration as the BBB long index), and historical default rates to account for losses.

Digging a little deeper into the steady state, both returns and volatility rise consistently with starting spreads, in a very orderly relationship (Figure 14).

We also note that they look fairly normally distributed, although with some extra downside tail risk (Figure 15).

By contrast, in the transitional state, returns are less clearly related to valuation (Figure 16).

This seems to be related to more pronounced skew during these periods. But the degree and direction of skew also appear to be a function of valuation.
  • When spreads have been in middle third (between 215 and 285bp), they have widened on average, but usually not enough to offset all returns from carry. As a result, the average estimated six-month return has been about 90bp. At the same time, the occurrences of large gains have been balanced fairly even against the number of large losses.
  • When spreads have been the widest third (widest than 285bp), they have usually tightened, generating an average estimated six-month return of more than 200bp. They have also had higher-than-normal probabilities of large gains or tightening events." - source Barclays.
This analysis in our opinion is very interesting from a Macro and Credit perspective, as posited by our Friend Paul Buigues in his 2013 post "Long-Term Corporate Credit Returns":
"Even for a rolling investor (whose returns are also driven by mark-to-market spread moves), initial spreads explain nearly half of 5yr forward returns." - Paul Buigues, 2013
Returns are related to starting valuations in both regimes, this is a very important point for credit investors we think. Also, according to Barclays, in the transitional state, returns are less clearly related to valuation. As concluded as well in this previous 2013 by our friend Paul:
"Benjamin Graham’s famous allegory of a “Mr. Market” who alternates between periods of depression and euphoria applies especially well to corporate credit investors. In addition to having a bipolar disorder, corporate credit investors are afflicted by a severe case of myopia, as they focus on current default rates, rather than trying to estimate realistic future default rates.
As a consequence, spreads themselves are a very good indicator of long-term forward returns, for both static and rolling investors."  - Paul Buigues, 2013
Do not focus solely on the current low default rates when assessing forward credit risk. Trying to estimate realistic future default rates matter particularly when there are more and more signs showing that the cycle is slowly but surely turning in both CRE and consumer credit. Peak jobless claims and peak hotel occupancy might after all show us that we are indeed moving towards a transitional state in these Frankenstein markets. One thing for certain, the energy sector credit woes in 2016 and outperformance in the second part of 2016 was a sign that returns can be clearly related to valuation or when credit spreads reached do not make economic sense that is when average high-yield spreads above 1000bp can be irrational.

Our core thought process relating to credit and economic growth is solely based around a very important concept namely the accounting principles of "stocks" versus "flows". We have used this core principle in the past when assessing the issues plaguing Europe versus the United States as per our September 2012 conversation "Zemblanity

We encountered previously through our readings an essential post dealing with our core concept of "stocks versus "flows" from Mr Michael Biggs and Mr Thomas Mayer on entitled - How central banks contributed to the financial crisis which explains precisely why both Friedman, Keynes and the central banks have been behind the curve in preventing the previous financial crisis and potentially the next one: 
"We have argued at some length in the past that because credit growth is a stock variable and domestic demand is a flow variable, the conventional approach of comparing credit growth with demand growth is flawed (see for example Biggs et al. 2010a, 2010b).To see this, assume that all spending is credit financed. Then total spending in a year would be equal to total new borrowing. Debt in any year changes by the amount of new borrowing, which means that spending is equal to the change in debt. And if spending is equal to the change in debt, then the change in spending is equal to the change in the change in debt (i.e. the second derivative of the development of debt). Spending growth, in other words, should be related not to credit growth, but rather the change in credit growth. 
We have called the change in debt (or the change in credit growth) the 'credit impulse'. The credit impulse is effectively the private sector equivalent of the fiscal impulse, and the analogy might make the reasoning clearer. The measure of fiscal policy used to estimate the impact on spending growth is not new borrowing (the budget deficit), but rather the change in new borrowing (the fiscal impulse). We argue that this is equally true for private sector credit." - Mr Michael Biggs and Mr Thomas Mayer on
 We have always wondered in relation to the global rounds of quantitative easings the following:
"Does the end (lowering unemployment levels) justify the means (increasing M) or do the means justify the end (deflationary bust)?"
Credit dynamic is based on Growth. No growth or weak growth can lead to defaults and asset deflation. The change in credit growth is a flow variable and so is domestic and global demand!

As a reminder, in our part 2 conversation "Availability heuristic" from September 2015, the liabilities structure of industrial countries is mainly made up of debt (they are “short debt”), in particular in Japan, the US and the UK. In contrast, the international balance sheet structure of emerging markets is typically composed of equity liabilities (“short equity”), which is the counterpart of strong FDI inflows that contributed to improve emerging markets’ external profile in the last decade. With a rising US dollar, what has been playing out is a reverse of these imbalances hence our "macro reverse osmosis" discussed in past conversations to explain violent rotations in flows from Emerging Markets. 

As we have seen on numerous occasions, and as we have remarked in our conversation  "Critical threshold", higher yields can lead to material fund outflows. As a reminder, more liquidity = greater economic instability once QE ends. As we move towards the transitional state described by Barclays in their note, one way of "mitigating" dwindling policy support would be to "embrace" a barbell strategy. 

For our final chart, no doubt to us than in these Frankenstein markets and experiments in the revival of organisms, it has been a "Goldilocks environment" for US credit.

  • Final charts - Fun in Funds
Bondzilla the NIRP monster has been created by our Dr Frankensteins in various central banks. Clearly the instigation of NIRP has put the demand for US credit from foreign investors into overdrive and it has been as we pointed out recently "Made in Japan". Our final charts come from Wells Fargo Credit Connections report entitled "Correction mode" from the 11th of August and displays fund flows relative to net share buyback:  

"Follow the Money
"Nearly $1.2 trillion of new money has flowed into U.S. Taxable Fixed Income mutual funds and ETFs since 2009. Much of that money came from Money Markets as central banks slashed cash rates and investors rotated from cash to bonds to enhance their yield. International investors jumped on the bandwagon in 2014 following the Fed’s taper tantrum, and accelerated their buying last year after European and Asian central banks cut cash rates to negative. Throughout the entire period, the lion’s share of money poured into U.S. IG credit funds. An upsurge in demand for corporate bonds allowed corporations to ramp up gross bond issuance to record levels of nearly $1.5 trillion per annum over the past six years. Much of that bond issuance was used to fund share buybacks, which helped buoy stock prices. All of this is shown in the charts above and below.

So, the $1.2 trillion question is, when does this virtuous cycle end? The short answer is, not yet as the world remains awash in excess savings and global central banks continue to pump about $200 billion of new cash into the system each month via their existing QE programs. But, we suspect the slow march back towards cash has begun. In the U.S., the Fed has raised the Fed Funds rate by 100 bps over the past 20 months to 1.25% (upper bound) and remains in tightening mode. The Wells Fargo Securities’ economists expect another hike this year, and three more next year. However, cash rates in Europe, Switzerland and Japan remain deeply negative. Until these rates are greater than zero, global excess cash will likely continue to hunt for yield enhanced investment opportunities around the world, a portion of which should flow into U.S. credit and serve as a backstop against spikes in volatility." - source Wells Fargo
As long as our Dr Frankensteins keep pumping into the system and NIRP stays firmly into play, there is no reason why asset prices cannot go even further towards the 11th mark on the Marshall amplifier of our central banks. The only creepy and scary thing with experiments in the revival of organisms is when our mad scientists will lose control of their Bondzilla, but we guess that's a story for another day while we keep monitoring the credit impulse globally and weakening signs from the US.

"Invention, it must be humbly admitted, does not consist in creating out of void, but out of chaos. " - Mary Shelley

Stay tuned ! 

No comments:

Post a Comment

View My Stats