Friday, 24 February 2017

Macro and Credit - Pareidolia

"It is discouraging how many people are shocked by honesty and how few by deceit." - Noel Coward, English playwright, composer, director, actor and singer

Watching with interest the continuation in rising asset prices and the relentless tightening in credit spreads, while monitoring our rewarding gold and silver miners exposure and itching to dive into US long bonds, we reminded ourselves for our title analogy of Pareidolia. Pareidolia is a psychological phenomenon involving a stimulus (an image or a sound) wherein the mind perceives a familiar pattern of something where none exists. Most people have never heard of "Pareidolia". But nearly everyone has experienced it. Anyone who has looked at clouds, the Moon and spotted faces or animal forms has felt the pull of "Pareidolia". Pareidolia can be extremely evocative particularly for people having "blind faith" or who are inclined in believing in "miracles" (such as the ones "pulled" by our central bankers). Put it simply, "Pareidolia is the human ability to see shapes or make pictures out of randomness. Pareidolia can be considered a subcategory of Apophenia being the human tendency to perceive meaningful patterns within random data. 
You are probably already asking yourselves where we are going with this, for us it is pretty simple. The much vaunted "Trumpflation" trade which has been put on meaningfully by many pundits, leading to some very distorted and crowded positions (short US Treasuries for instance) appear to us to be a case of "Pareidolia". The idea that the US economy can be "reflated" meaningfully at a rapid pace by the new US administration is an "optical illusion".

While last week me mused around the "Hypomanic" state of the markets and its pervasive elevated euphoric mood and the potential for equities to impact the tightening mood in credit which so far has prevailed, in this week's conversation we would like to cogitate more on the state of the credit cycle, some thoughts on Gold, and the risk of an equity drawdown. Rest assured a correction is most likely to happen in the US in the near future, regardless of the prevailing "Pareidolia" mantra.

  • Macro and Credit - Fooled by "Pareidolia"?
  • Final charts - Credit growth? appearances can be deceiving?

  • Macro and Credit - Fooled by "Pareidolia"?
In our previous conversation we wondered if in 2017 a US equities sell-off could for a change lead credit wider. We also mused around the impact a global rise in core inflation could potentially lead to, when it comes to a "sell-off". Since our last conversation we have seen some interesting developments in credit, namely that the tightening movement has been relatively impressive in the credit space as reported by DataGrapple in their blog post from the 23rd of February providing a good example of "Hypomania" in credit land:
"The credit market felt firm throughout the session. Credit indices would have closed a tad better across the board were it not for a late day wobble following a rapid increase of 4bps to 75bps of the OAT/BUND spread towards the close. The most striking feature of the session was actually the collapse tighter of a number of Crossover names. It is not often that 5 constituents of iTraxx Crossover (ITXEX) tighten in excess 30bps for completely unrelated reasons, while the index is left unchanged. SELNVX (Selecta Group BV) closed 100bps tighter at 700bps after an article published in the Italian press mentioned Argenta – an Italian vending machine operator – could be up for sale and that a merger with SELNVX followed by an IPO of the newly created entity would be on the table. DRYMIX (Dry Mix Solutions) closed 45bps tighter at 139bps, after a refinancing of the company’s debt was announced paving the way for an orphaning of the entity. WINDIM (Wind Acquisition Finance) closed 38bps tighter at 262bps after releasing good numbers and saying that they are happy with their current structure. Without many names widening significantly - New Look was the odd one out -, that was enough to send the basis of ITXEX back to its recent wides despite the absence of any sharp move of the market as a whole, which is fairly unusual." - source DataGrapple
Indeed, it is not often we do see such large movements for an index comprising 75 equally weighted single company credit default swaps. Given we have recently been fairly vocal on our contrarian stance relating to the Trumpflation trade, it seems to us that investors have been blinded by their "Pareidolia". On the subject of this phenomenon and the related "blind faith" of investors, we read with interest Morgan Stanley's recent Cross-Asset Dispatches note from the 20th of February entitled "How Likely Is a Major Equity Drawdown":
"We look at what key indicators say about the likelihood of a recession or a major equity sell-off over the next 12 months. On current readings, the likelihood averages ~20%.

We often hear "a large sell-off is unlikely, based on the levels of 'X'". We decided to test that:
A 15%+ correction for the S&P 500 within 12 months occurred around ~20% of the time when a wide variety of indicators were at the same levels they are now. That's close to average, but higher than options markets imply.
Don't forget, the probability of a 15%+ drawdown over ANY 12-month period is 18%:
We think investors underestimate the unconditional probability of a major drawdown (defined here as a 15%+ decline from current prices within a year).
What's worrying?
Year-over-year changes in oil, real yields and the level of jobless claims are all consistent with a 30%+ probability of a 15%+ drawdown over the next 12 months.
What's comforting?
The level of credit spreads and yield spreads and strong PMIs suggest that this drawdown risk is much lower than normal (in the single digits).
Investment implications – own the tails, hedges in credit: We believe that 2017 has larger-than-appreciated tails on both sides of the distribution. Calls on the S&P 500 are our preferred way of expressing underpriced animal spirits. On the downside, we think credit hedges are attractively priced, and 'lock in' one of the asset classes most relaxed about large drawdown risk.
Investment implications – own the tails, hedges in credit:
We believe that 2017 has larger-than-appreciated tails on both sides of the distribution. Calls on the S&P 500 are our preferred way of expressing underpriced animal spirits. On the downside, we think credit hedges are attractively priced, and 'lock in' one of the asset classes most relaxed about large drawdown risk.
Exhibit 1:
Oil, UST real yields and equities are flagging higher-than-average risk of a large equity sell-off over the next 12 months, while credit spreads remain relaxed
- source Morgan Stanley

Of course the worrying part is of interest to us, not because we suffer from a "pessimism bias" but, mostly because rather than going for the "optimism bias" and "Pareidolia", our contrarian stance pushes us towards "realism bias". 

To that effect, when it comes to the "barbaric relic" aka gold, when it comes to using Continuous Jobless Claims (Z-score 6months - 183 days and data non revised) as an input to generate "alpha" in relation to gold as an asset class, we find of interest that since the 22nd of January 2004 until the 17th of February, this simple input would have generated 0.76 of alpha according to the tool DecisionScreen:

- source DecisionScreen

What is as well very interesting is the sharpe generated by this simple trading rule over one year, we get a sharpe of 1.48 for a cumulated excess return of 23%:

- source DecisionScreen

Obviously as you can see from the tool above, when jobless claims rise, it is bullish for gold. The "barbaric relic" is still a buy. We also indicated dear readers our move towards a positive stance towards the asset class by the end of December 2016 if you remember our previous musings.

Also from a "realistic perspective" relative to gold, we read with interest Mish Shedlock's recent take on  the relation between gold and rate hikes in his post "Rate Hike Cycles vs. the US Dollar: Rate Hikes Bad for Gold?" from the 23rd of February:
"Gold Does Well in These Environments
1. Deflation
2. Hyperinflation
3. Stagflation
4. Decreasing faith that central banks have everything under control.
5. Rising credit stress and fear of defaults
Gold Does Poorly in These Environments
1. Disinflation (1980 to 2000 is a perfect example. There was inflation every step of the way but gold got clobbered).
2. Increasing faith in central banks’ ability to keep things under control (Mario Draghi’s “Whatever it takes” speech triggered a prime example)
Gold does worst in periods of prolonged disinflation and in periods of rising faith in central banks." - source Mish -
So while we don't have rising credit stress and fear of defaults à la 2016, the potential for stagflation should not be discarded with the increasing risks of a global trade war down the line. In that particular case, to repeat ourselves, a slowdown in global trade would evidently be bullish for gold.

But moving back to the subject of the credit cycle and being fooled by "Pareidolia", we keep telling you that we are slowly but surely moving towards the end of the cycle. From the latest Fed Senior Loan Officer Survey and in particularly Commercial Real Estate, to other segments of US consumer credit, but also other parts of the world, there are more signs that the "credit impulse" has been weakening as of late. When it comes to credit and in particular the credit cycle, the growth of private credit matters a lot. On that specific subject, we read with interest JP Morgan's special report from the 22nd of February entitled "Global credit cycle: Divided and sluggish":
  • Globally, the growth of private nonfinancial credit is slowing in a major departure from the past few business cycles.
  • In the DM, credit gains have firmed modestly but banks have recently stopped easing lending standards, giving some caution.
  • The private sector in EM ex China has been deleveraging since mid-2015. The process is broadening and there is no sign this headwind to growth is about to dissipate.
  • EM real GDP growth recently has stabilized as other factors offset the drag from credit. Our economists anticipate some pickup in 2017. Potential downside risks likely would be related to the Fed, US trade and tax policy, or China.

The growth of global private credit is slowing, maintaining a downtrend that took hold in 2014 (Figure 1).

This slowing is a major departure from the past few business cycles, when credit accelerated in the latter phase of the expansion. The dynamic underscores the significant structural forces damping credit growth. Demand for credit remains subdued in the wake of the global financial crisis. Additionally, global banks are becoming less willing to extend new credit. DM banks had been easing standards but stopped doing so in the latest quarter. EM banks have been tightening
standards since 2011 (Figure 2).
Beneath the global surface, the most striking feature of the credit landscape remains the sharp dichotomy between the DM and the EM. The DM private sector underwent a lengthy deleveraging phase from 2010 through 2014. However, DM credit growth has exceeded nominal GDP growth since 2015 and reached an estimated 3.3% in 4Q16. Taken at face value, the credit cycle has swung from being a significant headwind for DM economic growth to a modest tailwind. However, with DM banks turning more cautious, this could cap the recovery in credit growth or even reverse it.
The process is moving in the opposite direction in the EM and the dynamic is sufficiently powerful that it has dominated gains in the DM. EM private credit growth slowed to an estimated 9%oya in FX-adjusted terms in 4Q16, half the peak rate from early in the expansion (Figure 3).

Moreover, China’s double-digit gains strongly influence the EM aggregate. In the EM ex. China, credit growth has receded to just 6.0%oya. The private sector in EM ex. China has been deleveraging since mid-2015 (Figure 4).

These credit dynamics represent an ongoing headwind for EM growth, similar to what happened in the DM from 2010 to 2014. Nonetheless, it is encouraging to see that EM real GDP growth recently has stabilized due to positive offsets from EM policy support, easier financial conditions, rising commodity prices, and stronger DM growth.
Consistent with our previous analysis, this note focuses on broad private nonfinancial credit, which includes bank loans and corporate bonds, as well as other forms of financing. Up through 3Q16, our data are based on statistics provided by the BIS and the central banks (for background, see “The elephant in the room: Taking stock of the surge in EM private debt,” September 2, 2015, and “EM debt overhang is larger but its growth pace is slowing,” J. Goulden and A. Wong, September 30, 2016). We then generate estimates for 4Q16 using data on domestic bank credit, which accounts for the majority of total credit. The data for the EM cover only the subset of countries encompassed in the J.P. Morgan global economic forecast, as shown in Table 2 below.
Leverage cycle slowly turning up in DM
DM private credit contracted in the early days of the economic expansion. When credit growth resumed in 2011 the pace was slow and below that of nominal GDP as households and corporates worked to lower their indebtedness and banks struggled to shore up their balance sheets. A further acceleration in credit took hold in 2014 and beginning in 2015 credit growth moved above nominal GDP growth (Figure 5).
To be sure, any added support for consumer or corporate demand from increased leverage has been modest compared to the 1990s and 2000s cycles. The DM leverage ratio increased 1.7%-pts of GDP per annum on average from 1Q15 through 3Q16, a little more than half the pace from 2002-07.
The recovery in credit growth varies across the DM (Figure 6).

The pickup has been focused in the corporate sector (Figure 7), especially in the US, and so far has not translated into increased capex.

On the contrary, DM capex growth weakened progressively since 2014, falling below zero over the four quarters through 3Q16 (Figure 8).

It appears that many businesses, particularly in the US, used the money they borrowed for stock buybacks or M&A, or else they saved it. After having boomed in the previous cycle, household credit stagnated from 2009-12 as households in the US, the UK, 
By region, credit growth has recovered most strongly in United States, reaching about 4.6%oya in 4Q16—about 1%- pt above nominal GDP growth. By comparison, credit growth was more restrained in the Euro area, the UK, and Japan at 1.8%oya, 3.7%oya, and 1.8%oya, respectively.
We can gain insights into the bank credit cycle, and thus the broader flow of credit, through the quarterly surveys of senior loan officers conducted by the major DM central banks and the Institute of International Finance. These surveys provide information about the supply and demand for bank loans that we show in the form of net percent balances (“PB”).
The latest surveys, which compare banking conditions in January 2017 with those in October 2016, indicate DM banks shifted to a neutral stance on credit standards following an extended easing phase (Figure 11).

This shift includes the US, the Euro area, and the UK, though not Japan. In aggregate, DM banks stopped easing business credit standards in late 2015. They shifted to neutral for consumer credit standards in the 1Q17 survey. This shift in bank behavior is typical of late-cycle expansions and needs to be watched. In these phases, excesses in household or business balance sheets  often occur, or in real estate markets. In addition, these periods normally are characterized by tight monetary policy. Against this backdrop, it is not entirely clear why the DM banks are acting this way.
The relationship generally fits this pattern looking at the available history. The DM originations proxy is positive (the demand PB is positive while standards are neutral), implying DM loan growth should be increasing. This had been the case until DM loan growth stabilized in recent quarters, when it began to undershoot the model’s projection (Figure 12).

Note that since the current-quarter originations proxy tends to lead changes in %oya loan growth, it currently points to a significant pickup in DM loan growth this year. However, since the DM proxy has been over-predicting loan growth for about a year, we would temper its projection." - source JP Morgan
From a "Pareidolia" perspective, one would assume a significant pick up in loan growth, in particular in relation to an expansion in PMIs overall, but as stated by JP Morgan, their DM proxy has been over-predicting, meaning that, in similar fashion to the much vaunted "reflation" play, it can be highly deceptive. It might be a case of us being fooled by "Pareidolia" at this stage of the credit cycle. After all us human beings have a tendency to perceive meaningful patterns within random data. 

Clearly as we pointed out last week, in a state of "Hypomania" and 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. This has been confirmed by Wells Fargo in their note from the 22nd of February entitled "CRE Credit: Fed Concerns, Tighter Standards and Less Demand":
"CRE Credit Availability and Demand WaneThe slowdown in transaction volumes coincides with survey measures reporting weaker demand for CRE loans. According to the Federal Reserve’s January Senior Loan Officer Opinion Survey, CRE loan demand has fallen. Lenders reported weaker demand for multifamily and construction loans on net in the fourth quarter, while nonfarm nonresidential loan demand was relatively unchanged (middle chart).

On the supply side, lending standards for CRE loans continued to tighten across the board in the fourth quarter, marking the sixth-consecutive quarter of net tightening (bottom chart). The Fed survey noted particularly “significant” tightening for loans secured by multifamily properties and construction land development loans, with a net 33.3 percent and 25.0 percent of firms, respectively, tightening standards. Meanwhile, a more moderate 13 percent of banks reported tightening standards for nonfarm nonresidential property loans. The survey data suggest that bank lenders are proving themselves more cautious/selective, consistent with indications that the sector is at a mature stage in the current cycle.

The slower pace of foreign investment in U.S. CRE, and tighter lending standards for CRE loans, should provide some relief to Fed officials that have sounded caution over credit risks associated with the elevated level of CRE pricing." - source Wells Fargo
This brings us to the relationship between credit and more importantly the credit impulse and growth. Credit growth has always been the necessary condition for economic growth. On that subject we read with interest Wells Fargo's Interest Rate Weekly note from the 22nd of February entitled "Credit and Growth: A Partnership Not Opposition":
"Growth and Debt Finance
As illustrated in the graph below, there is a close link between economic growth and domestic nonfinancial debt growth.

This pattern reflects an interaction that works both ways. Economic growth prompts creditors and debtors to accept more debt as the expectation of growth, and thereby the increased financial rewards to creditors and debtors, improves. In turn, the availability of credit opens up opportunities for entrepreneurs to pursue prospects for growth.
During the most recent economic expansion, the very modest pace of debt growth has been associated with a period of subpar real economic growth. In part, this pattern may reflect the emphasis on financial regulation to avoid risk-taking that may have put a damper on financing any idea that carried a perception of risk.
The Specific Issue of Private Credit
During the current economic expansion, the pace of private credit growth to the consumer and business sectors has been particularly weak relative to prior economic expansions (middle graph).

Since both the business and consumer sectors are major contributors to economic growth and job gains, then it is not a surprise that the overall pace of economic growth and job gains have been modest, at best, during the current cycle. This is particularly apparent by the big drag of credit during the first few years of this expansion.
This interaction of private credit and spending imparts a pro-cyclical pattern to both economic activities which, if not carefully monitored, creates an abrupt halt to both when perceptions of risk and economic growth opportunities are altered.
Household Net Worth: A Balancing Act
During the latest downturn, the hit to real estate values was very evident in the steep drop off in household net worth as illustrated in the bottom graph. Each prior recession (1981-1982, 1990-1991, 2001) was associated with a decline in the growth of net worth but not to the extent experienced in the 2007-2009 period. Each recession, including 2007-2009, was associated with a decline in the growth rate of consumer spending.

Once again the link between financial/real asset values and consumer spending emphasizes our view that credit growth and economic growth are closely linked and that this link must be considered by public policymakers.
Current economic policy proposals favor faster growth and easier financial regulation. If thoughtful actions are taken, then faster growth can be achieved within a reasonable risk-taking environment." - source Wells Fargo
Unfortunately, where we disagree with Wells Fargo is that easier financial regulation will not lead to faster growth thanks to faster credit growth (credit impulse) at this stage in the cycle. When it comes to perceiving some form of "reflation" thanks to better credit growth on the back of a stronger credit impulse, we believe our final chart and conclusion below do point out towards yet another case of "Pareidolia".

  • Final charts - Credit growth? appearances can be deceiving?
When it comes to economic growth, private credit matters and matters a lot as we have been discussing to some extent in our conversation. For our final chart we would like to point out to yet another chart from the quoted JP Morgan report relating to the credit cycle. The chart displayed shows indeed that, when it comes to the growth of private credit, appearances can indeed be deceiving such as encountered through "Pareidolia" effects:
"Appearances can be deceiving. The growth of private credit exceeds nominal GDP growth and the debt/income ratio has climbed rapidly since 2012. This sounds similar to what happened in the late 1990s and the 2000s. However, this characterization belies the lopsided nature of the cycle. In the DM, credit growth remains muted and the leverage ratio barely has increased from the lows of a few years ago (Figure 18).

Globally, the EM have accounted for most of the expansion of private credit and leverage. Yet, in the EM, credit growth has been sliding, both outright and in relation to nominal GDP.
In our interpretation, for the DM, the credit cycle gradually has shifted from being a stiff headwind to a modest tailwind. One question is whether even this modest degree of support might be fading. DM credit growth recently has leveled off near 3%. With inflation normalizing, this pace is barely above that of nominal GDP. Moreover, DM banks have unexpectedly turned more cautious. US banks are tightening lending standards while Euro area and UK banks have turned neutral. This shift in the credit supply, especially if it extends  further, could halt the recovery in credit growth or even reverse it.
For the EM, we believe the credit cycle has been damping economic growth. This belief is based on the rapid deceleration in EM credit, especially outside China, to where the majority of the EM countries encompassed in our economic forecast are now deleveraging. From the standpoint of credit supply, the weak banking sector interfered with the transmission of monetary policy. From the standpoint of credit demand, the desire of DM households and corporates to shed debt damped the recovery in consumption and business investment. We think these same dynamics are occurring in the EM. Moreover, the number of EM countries that are deleveraging continues to increase and there is no sign in credit data or in the IIF bank lending conditions surveys that the underlying supply/demand dynamics driving the deleveraging are abating.
EM credit dynamics have damped economic growth but they have not generated recessions, much less a crisis. This is consistent with recent experience in the DM, where the extended deleveraging phase from 2010 to 2014 did not prevent GDP growth but acted as a powerful headwind. In previous research, we found that deleveraging typically occurs during economic expansions, not during recessions (see “A constrained global credit cycle,” October 14, 2016). However, most episodes of significant deleveraging occur during the early phase of economic recoveries, as with the DM during this cycle. In the relatively brief history of our EM data, which extend back about 20 years to just before the Asian financial crisis, a significant number of EM countries delevered nearly all the time except during economic recessions (Figure 19).

However, the more intense phases of deleveraging followed major economic downturns (again, as happened in the DM in the 2010s). In this sense, the current shift toward deleveraging in the EM is unusual." - source JP Morgan
Could it be that the pattern of "deleveraging" in both EM and DM is simply a case of "Pareidolia"? We wonder...

"It is only hope which is real, and reality is a bitterness and a deceit." - William Makepeace Thackeray, English novelist

Stay tuned ! 

No comments:

Post a Comment

View My Stats