Macro and Credit - Ballyhoo

"Chaos is inherent in all compounded things. Strive on with diligence." - Buddha

Watching with interest recent market gyrations, with the intervention of China in the mix to calm down the turmoil in its equities market, when it came to selecting this week's title analogy, we decided to go for the word "Ballyhoo":

1. : a noisy attention-getting demonstration or talk
2. : flamboyant, exaggerated, or sensational promotion or publicity
3. : excited commotion

A "Ballyhoo" is as well a "publicity, hype" from circus slang, "a short sample of a sideshow" used to lure customers (1901), which is of unknown origin. The word seems to have been in use in various colloquial senses in the 1890s.  In nautical lingo, ballahou or ballahoo (1867, perhaps 1836) was a sailor's contemptuous word for any vessel they disliked. There is as well a 2009 book entitled "Heroes and Ballyhoo" by Michael K. Bohn about sports stars during the period 1919-30s an "era of wonderful nonsense", when sport-crazed public demanded spectacles instead of just matches. Given this golden crazy age lasted 12 years long and many pundits are indicating that a recession in the United States could happen in the next two years, we are indeed wondering when this period of "irrational exuberance" to paraphrase former Fed supremo Alan Greenspan will end. On a side note, for sports fanatics out there, baseball legend Babe Ruth personified the Golden Age of the roaring "Ballyhoo" twenties, a close second was the boxer Jack Dempsey. 

In this week's conversation, we would like to look at housing as yet another sign that we think we have reached "peak" US economic activity.

Synopsis:

• Macro and Credit -  The real state of Real Estate in the US and the consequences
• Final chart - Beware of the velocity in tightening conditions

• Macro and Credit -  The real state of Real Estate in the US and consequences

Back in April 2012 we indicated the following relationship with the housing bubble: 

"The surge in the Baltic Dry Index before the start of the financial crisis was a clear indicator of cheap credit fueling a bubble, which, like housing, eventually burst. In the chart below, you can notice the parabolic surge of the index in 2006 leading to the index peaking in May 2008 at 11,440; with the index touching a low point of 680 in January 2012"  - source Macronomics, April 2012

- source Macrobond

The Baltic Dry Index, a gauge of rates to transport dry-bulk commodities including grains and coal by sea. Dry bulk cargo represents the largest part of the $380 billion shipping industry. Container shipping traffic is driven by consumer spending as it is dominated by consumer products. Container volumes to the United States are dependent on the housing market. Furniture and appliances are some of the top freight categories imported in both the United States but, in Europe as well from Asia. 

Any changes in consumer spending trends are depending on the health of the housing market:

- source Macrobond

With the Fed on its hiking mission, house affordability is being impacted through rising mortgage rates. Housing is getting more expensive in conjunction with labor shortages and rising costs linked to some extent to tariffs such as those on imported steel.

Basically it seems that the housing market in the United States seems to be stalling as affordability is becoming an issue:

- source Macrobond

Making a quick detour to shipping, there are as well signs that global trade is indeed cooling off. 

Another indicator other than the BDY is the Harpex Shipping Index. It is considered a good indicator of global economic fleet shipping activity since it tracks changes in freight rates for container ships over broad categories. It is slightly different than the BDY. Harpex weights average daily charter rates across eight size classes of vessels to formulate its index. A vessel containing dry bulk generally transports a single load type. Containers ship, by comparison, usually transport a wider variety of finished goods, which makers therefore the Harpex Shipping index a more accurate indicator for measuring global trade:

- source Macrobond

We can clearly see a deceleration in global trade happening at the moment thanks to this index.

But, let's return to US Housing. 

The housing market has clearly been the weak spot in the “strong economy” narrative. The Fed’s hiking path is leading to rising 30 year fixed mortgage rate towards 4.90%, the highest level touched since April 2011:

- graph source Macrobond

Single-family homebuilding is the largest share of the US housing market and fell by 0.9%. Housing affordability is becoming a challenge. 

At the same time, US housing prices are now 6.3% higher than their peak in July 2006 and 46% above their trough in February 2012:

- graph source Macrobond

On the subject of housing being a cause for concern, we read with interest Bank of America Merrill Lynch's US Economic Weekly note from the 19th of October entitled "Will housing hurt?":

"Will housing hurt?

• We have made a number of changes to our housing forecasts to reveal a weaker trajectory of sales, starts and home prices amid rising rates.
• We think home price appreciation is set to slow but not fall negative absent a recession in the overall economy.
• Housing is no longer a tailwind for the economy, but the headwinds are blowing very gently.

Home prices: from boom to bust

Home prices nationally, as measured by the S&P CoreLogic Case-Shiller index are running at 6.0% yoy as of the latest data in July. Assuming some modest slowing into the end of the year, we believe we are on track for home prices to end up 5.0% this year, as measured by 4Q/4Q change. As we look ahead into next year, we expect the slowing in home prices to persist, leaving home price appreciation (HPA) of 3% at the end of 2019 (Chart 1).

Thereafter we expect home price appreciation to hold at that 3.0% pace in 2020.

Back to econ 101, home prices should be a function of housing supply and demand. As we argued in Home sales: the peak has been reached, we think existing home sales peaked at the very end of last year and have since been moving sideways in a choppy fashion. This is a function of affordability which has been challenged from rising mortgage rates and elevated home prices. Inventory levels have remained extremely low, but since we look for some continued growth in single family housing starts but little change in home sales, we could start to see the supply of homes increase. The modest shift in the demand curve and out of the supply curve naturally implies slower  home price appreciation. As Chart 2 shows home price appreciation typically peaks along with the peak in home sales.

With mortgage rates heading higher, the challenges with affordability will continue. As a simple rule of thumb based on the NAR’s affordability index, we find that a 50bp increase in mortgage rates would need about a 5.5% offsetting drop in home prices in order to keep affordability unchanged. Of course, this does not account for the rise in income which provides an additional modest offset. Plugging in forecasts for mortgage rates based on our rates strategy call for the 10 year to end this year at 3.25% and 3Q 2019 at 3.35% – which implies close to 5.15% and 5.25%, respectively, for the 30-year fixed-rate mortgage – we would see affordability continue to slip lower (Chart 3). 

While affordability would still be above the historical average, it would still be more challenging than the past several years.

Another important aspect when thinking about the trajectory of home prices is an idea called “mean reversion”. Home prices are ultimately anchored to a fair value which is a function of income growth. Based on the OECD’s methodology, we compare nominal Case-Shiller home prices with disposable income per capita, indexed to 100 in 1Q 2000 (Chart 4) which shows the overvaluation during the housing bubble given the irrational exuberance in the market and easy credit conditions.

The housing bust left prices to tumble back below fair value. Based on our calculation, prices are once again overvalued on a national level, albeit not nearly as much as during the bubble period. Over time the overvaluation can be solved in two ways: 1) home prices grow at a rate below income for a period of time to close the gap; 2) home prices decline to correct the valuation difference. The pull to fair value can be quite strong.

Regional realities

We have been discussing the national outlook for the housing market but the dynamics will vary on a regional basis. Focusing on the top 20 metropolitan statistical areas (MSAs), we find that all 20 are still witnessing positive YOY home price appreciation, ranging from a low of 2.8% in Washington DC to a high of 13.7% in Las Vegas.

Generally speaking the West Coast has seen stronger home price appreciation relative to other regions. This reflects the fact that the West has enjoyed robust economic growth, supported by the thriving tech sector, which has led to greater income and wealth creation. This subsequently feeds into housing demand and a bid on prices (Chart 5).

At the same time, the West has also suffered from greater building constraints and a more severe housing shortage, owing to restrictive land-use regulations and zoning laws. This has contributed to home prices well outpacing income growth. Unsurprisingly, a regional analysis of price/income ratios finds the greatest levels of overvaluation in  Western MSAs (Chart 6).

Conversely, the Midwest cities were generally undervalued.

The higher prices rise in overvalued regions, the harder they may fall. So outright price declines could be seen as demand pulls back, though as discussed earlier we think this is less likely for aggregate national prices. Meanwhile, more affordable areas should continue to see price gains assuming healthy regional economic growth.

Sales and starts are a bit weaker

While existing home sales have peaked and will continue to hold around 5.5 million through next year, we see further upside for new home sales, albeit only modest. We forecast new home sales to edge up to 665K next year from our forecast of 640K this year, which is up from 612K last year. Why would new home sales increase while existing home sales move sideways? The recovery in new home sales was much slower since builders were hesitant to add supply to a challenged market, particularly in the early stages of the recovery.

We have revised down our forecast for starts this year and next. We expect 1.260 million starts this year and 1.30 million next year. The gain will be entirely in single family construction as multifamily has little upside.

- source Bank of America Merrill Lynch 

While we expect single family starts to edge higher – consistent with continued elevated levels of NAHB homebuilder sentiment and low levels of inventory – we think builders will be cautious in the face of rising mortgage rates." - source Bank of America Merrill Lynch

Unfortunately we do not share Bank of America Merrill Lynch's optimistic view. That would not make us "perma-bears" but we do not fall easily prey to "Ballyhoo" games namely sensational promotion.

No offense to Bank of America Merrill Lynch but, Main Street has had a much better record when it comes to calling a housing market top in the US than Wall Street. 

If you want a good indicator of the deterioration of the credit cycle, we encourage you to track the University of Michigan Consumer Sentiment Index given the proportion of consumers stating that now is a good time to sell a house has been steadily rising:

- graph source Macrobond

Maybe after all, they are spot on and now is a good time to sell houses in the US? Just a thought. Main Street was 2 years ahead of the 2008 Great Financial Crisis (GFC) as a reminder. Many pundits are predicting a recession in the US economy in the next two years.

As we have stated before, the Fed will continue its hiking path, until something breaks, and we have already seen some small leveraged fish coming belly up when the house of straw build up by the short-vol pigs blew up and when during the summer the house of sticks of the macro tourist carry pigs blew up (Turkey, Argentina, etc.). We keep pounding this but, Fed's quarterly Senior Loan Officer Opinion Survey (SLOOs) will be paramount to track going forward as the credit noose tightens.

Furthermore, it’s isn’t only residential housing which is a concern, in recent years Commercial Real Estate prices have gone through the proverbial “roof”:

- graph source Macrobond 

We think that "housing is no longer a tailwind for the economy" and that "headwinds are blowing very gently" is in this case a "Ballyhoo".

If one looks at US Homebuilders index versus the S&P500 that cyclicals matter when it comes to assessing the rising probabilities of a US recession:

- graph source Macrobond

This is telling you that housing activity is leading overall economic activity, housing being a sensitive cyclical sector. We have reached "peak" everything when it comes to US economic activity. It might be very well all downhill from there. We are already seeing signs in Europe with the latest PMIs of global trade deceleration, and not only from shipping mentioned above.

Also, if one looks at the S&P500 versus US Regional banks, one could conclude that "misery loves company":

- graph source Macrobond

The Regional Banks index has fallen 16.58% from its high back in early June and has fallen 7.05% since the start of October. Bank OZK’s stock dropped nearly 24% on the 19th of October after Commercial Real Estate (CRE) write-offs. The Arkansas-based bank is one of the largest condo construction lenders in Miami, NYC and LA. You would be wise thinking about selling your condo in Miami according to Main Street's predictive history.

As indicated by Bank of America Merrill Lynch in their weekly Securitized Products Strategy weekly note from the 22nd of October, bank stocks and MBS basis have a strong relationship since 2015:

"How Q3 bank earnings inform us

Domestic bank demand is key to agency MBS valuations; one simple relationship we have ascribed to is the strong relationship between bank equity valuations and the current coupon mortgage basis. Even recently, lower bank stock valuations have coincided with the basis widening. The underlying logic tying these two together is the outlook for bank balance sheets, reflected in stock prices, suggesting a technical backdrop for bank demand for agency MBS.

Many individual moving parts, however, come into play on the various pieces of bank balance sheets. For example, theory suggests deposits are impacted by the Fed’s balance sheet runoff. Appetite for securities relies on tolerance for capital volatility related to AOCI (all other comprehensive income), which changes with rate views and duration appetite. Finally, loans funded vary based on credit risk appetite and industry competitiveness, such as non-bank participation and accessibility to the high grade, high yield markets. These moving parts change, dampening or expanding bank demand for securities. We leverage the 3Q18 earnings call transcripts of the largest banks to extract takeaways on driving factors influencing these trends.

Lower tolerance for incurring AOCI risks – Tax reform, lower tax rates specifically, has reduced bank tolerance for incurring AOCI risks. AOCI losses have led to a larger tax deductible historically than what the current lower tax regime offers. The outlook for higher rates this year, and the potential for even higher rates ahead, has dampened enthusiasm for banks to take duration risk.

Cash is king, a compelling alternative, only getting better– Cash yielding 2+% compared to a post-crisis era of offering nothing raises the bar for investing in securities and taking on duration risk. Projecting returns on cash, along the forward path, only further stands to enhance the appeal of this strategy. Indeed, this is how the Fed’s tightening of policy works its way through banking channels, essentially raising the risk-free rate!

Yes, higher base case NIMs, but a few IF’s echoed – The selloff in rates highlights better NIM opportunities presented today, as deposit rates undershoot model forecasts. However, it is far from being just this simple. Convexity concerns and volatility ahead pose risks, along various forward paths. The outlook for loan growth, hinging on whether the economy keeps expanding, dictates securities demand, be it for HQLA/LCR reasons or for NIM/earnings.

The big question is can the US economy continue to expand as such a pace when housing is already struggling and even if FICO scores get lowered to facilitate credit card use by a pressurized US consumer?

There are indeed some implications down the line as highlighted by Bryce Coward from Knowledge Leaders Capital in his blog post from the 19th of October entitled "More evidence of a slowing housing market, and its implications":

"The slowdown housing activity leads overall economic activity by eighteen months. Housing, being one the most cyclically sensitive sectors of the economy, often feels the impact of higher rates well before other areas. This alone implies a peaking of economic activity right about now, leading to persistently slower growth rates through Q1 2020. 

Not coincidentally, a peaking of economic activity about now is also consistent with the 1.2% fiscal stimulus boost we’re getting in 2018. Incremental stimulus for 2019 drops to .4%, with the potential of that entire stimulus being negated by dead weight losses from tariffs, but that is a topic for another day." - source Knowledge Leaders Capital

This ties up nicely we think with Main Street sanguine view of the housing market, namely that it's less and less the time to buy a house and more and more the case of selling a house as per the previous credit cycle call Main Street made. The credit cycle is no doubt turning regardless of the "Ballyhoo" put forward by some pundits.

Sure overall, the latest quarterly Fed Senior Loan Officer Opinion Survey (SLOOs) points towards gradual tightening of financial conditions overall, yet the recent move based on the sensitivities of major market variables points towards an accelerating trend as per our final chart.

• Final chart - Beware of the velocity in tightening conditions

Our final chart comes from Morgan Stanley US Economics note from the 11th of October and indicates how using a more real-time look at financial conditions points towards a higher velocity in the tightening trend of financial conditions:

"An updated view on financial conditions indices shows a mixed picture, with the
Chicago Fed’s FCI actually easing further in the week ending October 5, while other alternative financial conditions metrics show a more considerable tightening in recent days.

The Chicago Fed updated its weekly FCI this morning. The latest update covers through last Friday, October 5, so it’s quite lagged. Somewhat surprisingly, the index eased 0.026 points – the largest one-week easing since the week ending August 10 and the 13th consecutive week of easing for the index.

The index now stands at a level of -0.76, a low since July 2015, driven by lower readings on the risk, credit, and leverage subcomponents. 49 underlying indicators tightened in the last week and 56 loosened – some of the biggest contributions to easier conditions were the Markit IG 5-yr senior CDS index, HY 5-year senior CDS index, and the 3-month TED spread.

An alternative metric that we look at for a more real-time look at financial conditions has shown a greater tightening in financial conditions so far this week. This metric tracks financial conditions based on the sensitivities of major market variables in the Fed’s FRB/US macro model, and we express it in a fed funds rate equivalent.

By this approach, financial conditions have tightened about 10bp from last Friday and about 50bp from the end of September. That compares with the experience from early February this year when financial conditions tightened about 80bp over a two week period." - source Morgan Stanley

Is this velocity seen in greater tightening of financial conditions a case of "Reflexivity", being the theory that a two-way feedback loop exists in which investors' perceptions affect that environment, which in turn changes investor perceptions, or is it simply a case of "Ballyhoo" at play? We wonder...

"Civilization begins with order, grows with liberty and dies with chaos." - Will Durant, American historian

Stay tuned!

Macro and Credit - Gullibility

"Mystical references to society and its programs to help may warm the hearts of the gullible but what it really means is putting more power in the hands of bureaucrats." -  Thomas Sowell

Watching with interest the 10th day of record highs for the Dow Jones and the on-going rally in other asset classes and in continuation to our previous theme of "Barnum statements" and "Woozle effects", we decided to pursue on the path of "tricks of the mind" in our title analogy by selecting "Gullibility".

"Gullibility" is a failure of social intelligence in the sense that a person or an investor gets tricked or manipulated (by central bankers) into ill-advised course of action (insatiable yield chasing). While it is closely related to credulity, many investors in the past have been vulnerable to this form of exploitation. If you would like to make a small experience on the subject you could try this popular test of "gullibility" by telling one of your friends that the word gullible isn't in "the" dictionary (earliest dictionaries did not by the way). If your friend is a gullible person he might respond "really?" and he will go and look it up. 

Some writers on gullibility have focused on the relationship between the negative trait of gullibility and positive trait of trust. The two are related, as gullibility involves an act of "trust", same goes with central banking (remember: "believe me it will be enough"). On this subject we find it amusing that a certain Stephen Greenspan (not Alan) in 2009, in his book "Annals of gullibility: why we get duped and how to avoid it" presented dozen of examples of gullibility in literature and history to name a few: The Adventures of Pinocchio, Little Red Riding Hood, The Emperor's New Clothes, The Adventures of Tom Sawyer, Romeo and Juliet, Macbeth, Othello and even in the classic The Art of War, The Prince or The Trojan Horse story. Greenspan argues that a related process of self-deception and groupthink factored into the planning of the Vietnam War and the Second Iraq War in his book. In science and academia, gullibility has been exposed in the Sokal Hoax which we referenced to in a previous post of ours "The Sokal affair" describing the acceptance of early claims of cold fusion by the media. 

In society, tulip mania and other investment bubbles (Bitcoin comes to mind) involve gullibility driven by greed, while the spread of rumors involves a gullible eagerness to believe (and retell) the worst of other people. You might be wondering where we are going with this but hearing the other Greenspan recently, Alan that is, making yet another claim that there is a bond bubble in the making, although he has done so in recent years made, it interesting to say the least.  Two years ago, when the 10-year US yield was 2.44% the other Greenspan told Bloomberg TV that "we have a pending bond market bubble." In a Bloomberg TV interview in July 2016, he again expressed concerns about stagflation and said "we're seeing the very early signs of inflation beginning to tick up." He also told us at the time with 10-year Treasury yield at 1.50% thanks to Brexit concerns, that he was "nervous" bond prices were too high. In terms of "gullibility", this is the same Greenspan that told us that no one saw the previous crisis coming. Is third time a charm for the "Maestro"? We wonder. Stephen, the other Greenspan, in 2009 wrote that exploiters of the gullible "are people who understand the reluctance of others to appear untrusting and are willing to take advantage of that reluctance." In 1980, Julian Rotter wrote that the two are not equivalent: rather, gullibility is a foolish application of trust despite warning signs that another is untrustworthy. When we see European Junk Bond yields falling to another record low of 2.33% and closing on 10-Year US Treasury Yields at 2.26%, we think that "gullibility" applies for European high yield investors as a foolish application of trust despite warning signs that our central bankers are untrustworthy hence our chosen title. To check your investor gullibility factor you could simply ask yourselves if you would rather hold US Treasuries for the next 3 years or European High Yield but we ramble again...

In this week's conversation, we would like to look again at the potential risk for a convexity event we discussed previously in our conversation "Bond ruck" from a NAIRU ((non-accelerating inflation rate of unemployment) perspective. 

Synopsis:

• Macro and Credit - Convexity - A NAIRU headache?
• Final charts - The barrel in the credit revolver is getting empty

• Macro and Credit - Convexity - A NAIRU headache?

Traditional measures of monetary policy such as the Taylor Rule that are based on growth and inflation and a nonaccelerating inflation rate of unemployment, or NAIRU, of 5 percent. NAIRU is the lowest unemployment rate an economy can sustain without spurring inflation according to the definition. The name "NAIRU" arises because with actual unemployment below it, inflation is expected to accelerate, while with unemployment above it, inflation decelerates. One practical use of this model was to provide an explanation for stagflation, which confounded the traditional Phillips curve, hence the concerns of the "Maestro" and his bond bubble fears. The "NAIRU" analysis is especially problematic if the Phillips curve displays hysteresis, that is, if episodes of high unemployment raise the NAIRU. This could happen, for example, if unemployed workers lose skills so that employers prefer to bid up of the wages of existing workers when demand increases, rather than hiring the unemployed. As we posited in our June 2013 conversation "Lucas critique":

"As far as we are concerned when unemployment becomes a target for the Fed, it ceases to be a good measure. Don't blame it Goodhart's law but on Okun's law which renders NAIRU, the Phillips Curve "naive" in true Lucas critique fashion." - source Macronomics, June 2013

Also as we pointed out in our July conversation "The Rebound effect" in our final chart relating to the death of the Phillips curve, the framework is still adhered to by the Fed, meaning that they should be very slow in removing the credit punchbowl. We also pointed out that subdued job switching is due to a mismatch between jobs and worker skills. To repeat ourselves, what matters is the quality of jobs but we should add that to ensure Americans are great again, they need to get better skills for the jobs being advertised and that goes through training. The Fed's models are built on past relationships. Yet, what seems of key importance to us is that many believe the NAIRU unemployment rate to be around 4.5%, while inflation and wage pressures remain muted because as we pointed out in recent conversation demographics mismatch in the labor market is affecting wage growth. 

You might already being asking yourselves what it has to do with convexity and bond bubble fears for the "Maestro"? It all has to do with "gullibility". On the subject of the non linearity of the Fed's hiking path and therefore convexity risk, we read with interest Bank of America Merrill Lynch's take in their Liquid Insight note from the 4th of August entitled "Who doesn't love 90s rerurns?":

"Key takeaways

• The Fed is facing a similar situation to the 1990s; the unemployment rate is uncomfortably low, but so is core inflation.
• In the '90s, there were data head-fakes and confusion in the Fed's reaction function, resulting in a "fits and starts" cycle.
• This sounds familiar. We think we should prepare for the possibility of further pauses in the hiking cycle, akin to the '90s.

A not-so-distant memory

Nostalgia for the 1990s is tangible: the economy was in a prolonged expansion, the budget deficit was turned into a surplus and, importantly, the quality of television improved with Seinfeld, The Simpsons and Friends. We are now facing a feeble recovery from the Great Recession, a swelling budget deficit and reality television. Don’t fret; there are some things in common. The Phillips Curve faced similar challenges in the late 1990s as it does today (Table 1). Today’s Fed should keep in mind the lessons of the 1990s:

1. Relying on the Phillips Curve relationship can prove dangerous
2. It is essential to maintain credibility as a defender of price stability
3. Hiking cycles can have fits and starts

We think the 1990s episode puts the spotlight on a risk to our forecast. Although we have been arguing that the risks are skewed toward a slower cycle, our baseline expectation is still for the Fed to follow their dots (a hike in December and three more hikes next year). Given the lessons learned from the experience of the 1990s, we should prepare for the potential of a pause in the tightening cycle early next year.

Measured pace is the exception, not the rule

Many investors tend to look back at the 2004-6 “measured pace” episode as the norm for Fed tightening cycles. In reality, the 1990s offers a much better comparison. In the 1990s, there were many head-fakes in the data and changes in the Fed's reaction function, which led to a “fits and starts” hiking cycle. As Chart 1 shows, the Fed delayed hiking rates until three years into the recovery; did a relatively normal tightening cycle, but then did two mini-cycles of cuts followed by another hiking cycle.

On paper, this seems like a Federal Reserve that was quite confused. Why start the hiking cycle in 1994 with inflation still low? Why the abrupt reversal and then the delayed resumption of hikes? It helps to think about Fed hikes back then as a complicated interactive process rather than a pre-ordained plan.

Stage 1 (1994-1995): preemptive tightening of monetary policy against the "inflation scare" in bond markets. The unemployment rate at the onset of the hiking cycle was only 6.5% while wage and price inflation was subdued. But the 30-year Treasury rate soared to 8.1% in November 1994 from 5.9% in October 1993. Fed officials became concerned about losing their inflation-fighting credibility. In the February 1994 minutes, Fed officials noted that “a relatively small move would readily accomplish the purposes of signaling the Committee’s anti-inflation resolve” and in the March 1994 minutes argued that “a stronger policy action …would serve to underscore the Committee’s commitment to its price stability.”

Even within this apparent steady tightening cycle there were plenty of head-fakes. Three times the Fed signaled that it might be done and skipped a meeting. Each time they came back with a “catch up” hike of 50bp, 50bp and 75bp, respectively. This stopand- start occurred for two reasons. First, they expected hikes to slow the economy, but the labor market continued to motor ahead. Second, most Fed officials came into the period thinking NAIRU was around 6.5% but as inflation failed to respond, they revised their estimates lower.

Stage 2: Recalibrating an overshoot: By July 1995, the Fed seemed to regret those hikes, noting in the minutes that “some modest easing was desirable now that the growth of the economy had slowed considerably more than anticipated and potential inflationary pressures seemed to be in the process of receding.” This led to 75bp of cuts over the next several months.

Stage 3: Watchful waiting: The Fed went on hold for roughly two years. While dissents were minimal (because of Greenspan’s domination of the Committee at that time), there was growing pressure from the Committee to resume hiking. Skimming through the transcripts, we find that Janet Yellen was one of the voices advocating for the Fed to hike rates. Her arguments echo those that she is making today.

In the September 1996 transcripts Yellen gave two arguments for hiking:
"First, an unemployment rate of 5.1% lies near the lower end of almost anyone's estimated NAIRU range. Second, whatever the NAIRU, the unemployment rate does have predictive power for changes in the inflation rate...for that reason, I would conclude that the risk of inflation has definitely risen, and I would characterize the economy as operating in an inflationary danger zone."

"My concern is that a failure to shift policy just modestly in response to shifting inflationary risks could undermine the assumptions on which markets' own stabilizing responses are based."

Stage 4: Crisis mode: The FOMC did eventually hike by 25bp in March 1997, but it proved to be a one-off as the economy faced a variety of exogenous shocks thereafter, including the Asian currency crisis in 1997, the Russian default and failure of LTCM. Through it all, the unemployment rate continued to glide lower, but yet the "inflationary danger zone" proved benign. As such, if the Fed actually engaged in a hiking cycle, the economy might have suffered, as would have the Fed's credibility. It turned out that Greenspan was right to hold off as he correctly saw that stronger economic growth was generated by strong productivity gains, which were disinflationary.

Stage 5: The real deal: Heading into 1999, the Fed was facing a tight labor market with the unemployment rate hovering just above 4%, wage growth of about 3%, but core PCE growth of only 1.4%. Asset prices were elevated with signs of "irrational exuberance" in the markets: by the time the Fed hiked, the cyclically adjusted P/E had climbed to historical highs. This sounds quite familiar to the situation we are in today with a tight labor market, but muted price pressures with the exception of assets.

A big theme then and now: shifting views of the Phillips curve

Janet Yellen was not the only one who believed that inflation would accelerate as the unemployment rate fell further. In fact, it was very much the consensus view. According to the Blue Chip consensus survey, economists in 1996 were looking for CPI inflation of 3.0% for 1998 while it ended up coming in at 1.6%. Core CPI similarly came in below 3%, averaging 2.3% for the year. Economists reacted to the data and within a year were bringing forecasts lower, coming in close for 1999 (Chart 2).

The challenge is that models are imperfect. If you were sitting in September 1996 with an unemployment rate just above 5%, a standard Phillips Curve would estimate 3.5% for core CPI over the next 12 months, which was about 1.3pp too high relative to the actual data. At the time, the weakness in inflation was explained by external factors such as globalization and technological improvements that could not be captured properly by the Phillips Curve. There was quite a lot of awareness at the time that the Phillips Curve was failing. In September 1999, John Williams and others at the Board authored a piece called “What’s Happened to the Phillips Curve,” which looks at various specifications of the Phillips Curve to attempt to improve the fit in the 1990s.

Finding the right re-runs

It is important to focus on the experience of the 1990s, rather than the 2000s, when considering risks to monetary policy. Fed officials struggled with the same questions two decades ago as they are today. How tight is policy? Does the Phillips Curve work? Have we maintained credibility as defenders of price stability? How should we handicap market risks?

The result in the 1990s was a hiking cycle that was not linear. It may end up being the same story today. The Fed had a 12-month pause between the first and second hikes before slowly delivering another 75bp of hikes. We could be reaching a crossroads once again as wage and price inflation remain subdued. After another hike, the Fed will be a hair away from neutral (Chart 3), which means the next hike would no longer be considered a removal of accommodation, but rather a true tightening of policy. 

We think a considerable risk to our forecast is that the Fed pauses the hiking cycle early next year, slowing the pace of rate hikes." - source Bank of America Merrill Lynch

The consensus view it seems is that inflation will accelerate as the unemployment rate falls further, in similar fashion to the 90s. We discussed convexity in our conversation "Bond ruck" recently, given that as well, some have been arguing that in a rising rates environment you would be better off with RMBS thanks to negative convexity features. We indicated there was a catch because of the significant need in "delta hedging". Maybe the Maestro's fear comes from the potential for a "Convexity event" fueled by MBS hedging, particularly in the light of the balance sheet exercise soon to be taken by the Fed. As we pointed out in the final point of this particular conversation, the unemployment rate is currently far below NAIRU. The risk again, is therefore a policy mistake in similar fashion to what was avoided by the Maestro in 1997, but, eventually burst the bubble in stage 5 as pointed in Bank of America Merrill Lynch above. If indeed Janet Yellen still believes in an acceleration of inflation as per the September 1996 transcripts, there is indeed a cause for concern from a "convexity perspective". Yet, when it comes to central banks' "Barnum statements" we are not rest assured that "gullible" and like any good behavioral psychologist we tend to focus on the process rather than on the content of their narrative.

What we think is very similar to the 90s so far is indeed the non linear hiking path chosen by the Fed. On that note we agree with HSBC's Steven Major take on disinflation sinking further Treasury Yields as reported by Bloomberg in their article from the 7th of August entitled "HSBC's Steven Major Sounds a Bearish Alarm on European Credit":

"Major, HSBC Holdings Plc’s head of fixed-income research, is a key proponent of the view that global interest rates can stay low for longer, he says investors aren’t being paid for the risks they are taking in corporate debt markets in the euro area, with yields a whisker away from post-crisis lows.

"Low volatility across asset classes may give a false sense of security and bond markets may be caught napping," HSBC strategists led by Major wrote in a note published Monday. "The risk is that with increasingly interconnected capital markets, driven by years of international spillover from quantitative easing, local triggers can have a more global impact than before.

Major’s bearish case: The European Central Bank’s asset-purchase program won’t be as large over the next year as the 12 months prior, while there’s a natural cap on credit demand as yields sit "materially below" typical expense ratios for retail funds, and to a lesser-extent insurance funds. "Gross yields leave very little left for income-seeking savers," he adds. "Spreads could widen with a volatility shock."

The strategist also takes exception with the argument that the euro-area economy is ensnared by Japanese-style stagnation, which would keep spreads in check for decades to come, amid low interest rates and a lifeless corporate sector." - source Bloomberg

We also agree with his forecast for 10 year UST yield of 1.9%, that still make us part of the bond bulls after all. As we pointed out at the beginning of the conversation, at this stage we would rather hold US Treasury notes than European High Yield no offense to the gullible investing crowd busy picking the remaining basis points in front of the credit steam roller. We also agree that the US treasury market and the US dollar are reflecting a weaker economy. If there is a market which is less gullible to the "Barnum statements" from the Fed, it is the US bond market.

The big "Gullibility" question is one of returning duration risk through a sudden rise in bond volatility and we are not even taking into account exogenous geopolitical risk at this stage (which would be bond and gold bullish and oil bullish). On this subject we read with interest Nomura's take in their Inflation Insights note from the 7th of August entitled "Returning real duration risk to the market":

"A central banker’s headache

There are still many scenarios in which the gradual normalisation in the real monetary stance pursued by central banks might not have a gradual impact on financial markets. Uncertainty still looms large on key variables conditioning a balanced path to “normal.” Separately and/or collectively, the lack of term premium in inflation valuations, the level of the natural real rate of interest, the anchoring of inflation expectations, and flow dynamics from the unwinding of large-scale asset purchases can often have outsized impacts on global bond markets. Of course, central banks might find a perfectly balanced path, yet risks remain that real rates could overshoot. We suggest short real rate trades as a hedge against an environment of fast-rising nominal rates and slow-rising inflation.

“Taper tantrum” vs. “gradualism”

In Inflation Insights - Mini “taper tantrum” vs. “dear Prudence”, we argued that risk of a large, unexpected upswing in real yields remained. Despite normalisation in the real stance of policies being widely expected and advertised by central bankers, some mechanisms might facilitate unusually large yield changes.

Several factors potentially generating unexpectedly large yield moves

These mechanisms originate in both cyclical and structural features of economies. At a conceptual level, they stem from the possibility of a sharp rebuilding in term premia accompanying higher interest rates. The inflation premium – the price of a risk of inflation overshooting – remains low on most metrics. The term premium on the inflation component of yield remains also very subdued, as we have shown in Inflation Insights - A short trip to the long end . Therefore it is from the real component of yields that the risk of fast-rising rates comes from.

At a practical level, factors potentially resulting in an unexpected and uncontrolled rise in real yields are:

• The non-linear dynamics between the level of real rates and the real term premium, as short-term nominal rates retrieve some margin away from the effective lower bound.
• The re-pricing of the natural real rate of interest (NRRI) until recently, financial markets have rather followed central banks in their downside revision of the natural real rate and accepted the consequences on fixed income valuations. Even though there is an active debate among Fed members about NRRI, most still believe it will be heading higher as they continue with their normalization. A repricing could happen if surprises on activity remain to the upside, e.g., as they have been in the euro area.
• The complex interplay between segmentation across market participants and largescale asset purchase programmes conducted by central banks. For example, the transfer back to private investors of MBS paper would probably be accompanied by reintroduction of interest rate hedges. Another example is the behaviour of insurance companies and pension funds in the euro area increasing their purchases of fixed income assets with the ECB purchase programme – against the intuition of the balance portfolio channel pushing such investors to invest more in riskier securities.
• The discontinuity of asset purchases and the “signalling channel” – an abrupt stop to bond buying or re-investment would result in the pricing in of a much denser sequence of “taper” and then rate hikes. In our opinion, the long lead time of preparing markets for the Fed’s eventual balance-sheet unwind could mean that it is rather unlikely in the US (that said, there is a risk to some US rate vol if the Fed gets new leadership). However, it is made possible in the euro area by the perception of technical constraints on the PSPP.

Low inflation duration risk, high real duration risk

All these factors would allow for potentially large and sudden nominal yield changes, translating into large and sudden real yield changes through the elasticity (“beta”) of real yields to nominal yields. However, we think there are reasons to expect real yield changes to be larger than nominal yield changes.

One feature of the environment of low inflation has been that inflation valuations are unusually highly sensitive to current realized inflation trends (both markets and consumer prices). It is not only the expectation component of inflation valuations that has remained subdued, it is also the inflation premium embedded in valuations that has only marginally corrected from its negative levels. The lack of any increase in inflation premia and also inflation term premia would push real yields upwards, even faster than nominal yields.

As Figure 1 shows, it was the case in the early stages of the “taper tantrum” in 2013. The 5y5y real rate increased by about 150bp initially - the 5y5y inflation rate lost about 40bp over the same period. With markets remaining sceptical about the sustainability of higher inflation, it is very possible that the next big bond market sell-off will be a real-rate story.

 Away from the Effective Lower Bound (ELB): a distributional change

The risk of non-linear dynamics on rates stems above all from the proximity of the ELB. The effectiveness of monetary policy is conditional on the ability of the central bank to change the real rate – in other words, to increase the nominal rate faster or slower than the expected change in inflation. At the ELB, the short-term real rate is fully determined by the (negative of the) level of expected inflation. Central banks must resort to other instruments to lower real yields further on the curve – a methodology much advertised in Japan for instance for QQE.

Credible alternatives to the use of the nominal short-term rate might lower real rates further along the yield curve. Their credibility rests to a large extent on the notion that long-term inflation expectations do not fall too much – otherwise the objective of lower long-dated real yields is difficult to achieve. This is why the focus on inflation valuations by central bankers increases – and not only the focus on realised inflation and inflation scenarios.

Figure 2 shows that 5y5y inflation rates remain above half a standard deviation below their 2014-17 average. 

The 1-year inflation rate in 9 years is 30bp below the 2% target in the euro area. In the US, it is 2.60% - which is only just consistent with the Fed’s long-term PCE target and the CPI/PCE basis. In the UK, the 1y in 9 years is 3.45% - the only inflation rate above the central bank’s target of about 3% in RPI-equivalent terms ( the target is set in terms of CPI inflation, UK markets quote RPI inflation). Yet the spread to the target has been contracting since the beginning of the year. In our view, inflation markets have not priced in risk of inflation overshoots that these numbers suggest, consequently as a result the right-side of the inflation distribution remains thin.

A particular risk arises in case of a “spurious” surge in inflation expectations – for example acceleration in prices stemming from rising energy prices. In this case, central banks could increase the pace of nominal rate normalisation on the back of a temporary increase in inflation and despite a low probability of protracted and self-sustained inflation backdrop. Although central banks in general try to “look through” temporary factors, the distinction between “temporary” and “persistent” is not always easy to make in the heat of the moment. In addition, one consequence of the effective lower bound is to skew the distribution of inflation expectations – the normalisation of the distribution to the right can easily be triggered by a temporary factor such as large oil price increases.

Re-pricing the natural real rate of interest

The consequences of a lower natural real rate of interest for monetary policy are very far-reaching, in our opinion, and the uncertainty around the potential evolution of these rates would suggest in and of itself, an increased cautionary outlook for central banks in the pursuit of “normalisation”. An important aspect of this issue is that financial markets have priced in much lower levels of the “equilibrium” real rate. We focus on a simple statistical result – the mean-reversion level of the 5-year real rate, 5 years forward.

Figure 3 shows the contraction in this market’s estimate of the natural real rate. The magnitude is 150bp between our two samples across markets – the common magnitude of the contraction suggests a global issue rather than a country-specific problem.

The 5y5y real rate using swaps is currently about 0.40% in the US, 0% in the euro area and -1.40% in the UK. These levels are very close to the “mean-reverting” level that has prevailed since 2012 – but still very far from the levels before 2008. They are also not too far in the US and the euro area from estimates of the natural real rate provided by Laubach and Williams – respectively 0.45% for the US, -0.30% for the euro area and 1.50% for the UK. A very large discrepancy remains for the UK. In addition, these numbers suggest that “real normalisation” towards the natural real rate has already occurred for some version of the natural real rate – especially in the US.

When looking at the real rate curve in the US, given where 30-year TIPS have been trading, it seems as if the markets have taken to heart the idea that the long-run real rate is somewhere around 1% (2% inflation minus 3% median long-run dots). Down the curve is where the debate over how high the natural real rate could go, in our view. Further increase in real yields would therefore potentially put policy at or above the neutral level – a result that currency markets have not missed in the case of the euro area, with the consequence of a sharp euro appreciation. A sudden and sharp revision in the neutral level advertised by central banks could result in much higher real rates, with the risk to push policy into restrictive levels, which in turn would add to the negative pressure on inflation and activity. This would happen at a time when short-term rates are only some basis points away from the ELB, with limited restored potential for accommodation.

Rushing to the exit: the segmentation of investors

Another factor creating potential non-linear dynamics in real rates stems at a technical level from the segmentation of investors – a feature that accounted for the unusually large changes in US real yields in 2013 referred to as “taper tantrum.” We have explained this consequence of segmentation in Inflation Insights - Mini “taper tantrum” vs. “dear Prudence”.

In the US, a structural and potential destabilising feature for rates market dynamics is the process of MBS holdings of the Federal Reserve System slowly returning to private investors. For years, the Fed has basically warehoused US rate vol (see link) by holding so many MBS securities out of the private sector. MBS hedging in the past was an issue for global markets during quick rises in US rates (where MBS hedgers can push nominal US rates up even faster – known as “convexity hedging”). However, rates/spreads can also rise over time as private investors begin to hedge their MBS for interest rate risk.

In the euro area, the ECB provided evidence that the implementation of the PSPP resulted mostly in the selling of euro fixed income assets by non-residents to the eurosystem, while insurance companies and pension funds did not change their investment pattern. Such investors given their high fixed-income exposure would typically react to a change in the direction of monetary policy towards faster normalisation and create “rush-to-the exit” dynamics." - source Nomura

As we pointed out, a "Convexity event" would be possible depending solely on the "velocity" in the rise in US rates. Given the Fed's recent "Barnum statements" and dovish tone, we think that the Fed's first priority will be to go ahead with the initiation of its balance sheet reduction in September, before following up on additional rate hikes we think. There is a possibility though that, given the current NAIRU discrepancy, Janet Yellen travels "back to the future" so to speak, to the 90s and this, would of course generate significant instability and the aforementioned "bond" volatility which is so feared by the beta crowd and carry players alike. "Velocity" will induce "volatility" and therefore weigh on fund bond flows. Obviously, the risk is a significant rise in real rates, which would be simultaneously bond and gold bearish. We might have been gullible in some instances, but it isn't our core scenario for the time being. Not until we see the facts change, namely less muted wages growth, that we will change our mind to paraphrase Keynes.

For our final charts, back in our conversation "Orchidelirium" we asked ourselves if the US consumer was "maxed out". We noticed at the time that consumer loan demand, a finding consistent with the weaker spending in Q1 had been cooling. We mentioned it was a significant indicator to monitor in the coming months.

• Final charts - The barrel in the credit revolver is getting empty

In our book, consumer credit year-over-year change needs to be monitored very closely. Our final charts comes from Wells Fargo Economics Group note from the 7th of August and shows that Consumer Credit Growth has been slowing to end Q2. We are getting worried as well that total consumer credit momentum is decelerating:

"Consumer Credit Growth Slows to End Q2

Consumer credit rose $12.4 billion in June. Nonrevolving credit growth has been slowing for the past couple years, and with the recent halt in revolving credit’s momentum, total consumer credit is decelerating.

 Nonrevolving, Revolving Credit Both Slowing

• Steady growth in nonrevolving credit earlier in the economic expansion helped spur total consumer credit higher despite anemic growth in the revolving space.
• More recently, however, revolving credit growth has stalled. Net-charge offs have risen and delinquencies are up slightly, although both of these series have risen from historically low levels.

- source Wells Fargo

In June this year we mentioned our concern relating to the US consumer getting close to being maxed out in our conversation "Potemkin village":

"While it might be premature to pull the curtain on the Potemkin village, if indeed we break the 5% level for nonrevolving credit and continue to see a deteriorating trend in the coming months, then it will be a cause for concern. For credit markets at the moment, it's pretty much "carry on", though we are clearly tactically more cautious with High Yield and high beta in general." - source Macronomics, June 2017

If "Gullibility" is a failure of social intelligence in the sense that an investor gets tricked or manipulated by central bankers in buying European High Yield for example, then again, the US yield curve has been immune to the old tricks played by the Jedis at the Fed. As the economic cycle matures, so do credit standards and credit demand that simple. Most recent evidence indicates us that both consumer credit and business lending demand is slowing regardless of low unemployment and NAIRU readings. On top of that nonfinancial domestic corporate profit margins peaked in early 2014 and have declined in the last two years. This means, dear credit friends, that you need to have a more cautious stance going forward. Some investors might still be gullible enough to believe in "Barnum statements" and "Woozle effect", to paraphrase Jabba the Hutt in the Return of the Jedi, dear central banker your old Jedi mind tricks don't work on us.

"Most people are sceptical about the wrong things and gullible about the wrong things." - Nassim Nicholas Taleb

Stay tuned!

Charts of the Day - Positive correlations and large Standard Deviation moves

"The least initial deviation from the truth is multiplied later a thousandfold." - Aristotle

The Charts of the Day we have chosen to display reflects our concerns relating to the relationship between rising positive correlations due to the "divine" intervention of our "generous gamblers" aka "omnipotent" central bankers and the significant rising "instability" à la Minsky it entails. This can be ascertained from the rise in +/-4 standard deviations moves or more in various asset classes. 

The significant rise in these large standard deviation moves can be seen in the chart below from Bank of America Merrill Lynch European Credit Strategist note from the 11th of August entitled "The liquidity paradox":

"Chart 1, above, refreshes our “correction counter” that we first published in May. It sums, over time, the number of instances that assets in our sample register +/- 4SD moves. The chart shows that asset classes more directly impacted by central bank policy (such as government debt and currencies) are indeed seeing a relative rise in the number of “corrections” over time. But assets where central bank policy is a less direct driver of performance (such as equities, where the growth outlook is arguably key) are seeing “corrections” increase in a more linear fashion.Why might the era of high central bank help be, perversely, resulting in more market corrections? We argued in May that the cumulative result of so much monetary policy support is that the market’s emotional gap between fear and greed has narrowed." - source Bank of America Merrill Lynch

This pattern was expected given following the Great Financial Crisis (GFC), positive correlations have been rising between different asset classes, a subject we discussed again in May this year in our conversation "Cushing's syndrome" and leads us to use once more to use the below IMF chart in the light of Bank of America Merrill Lynch standard deviations move chart above:

"Cushing's syndrome" aka central banking "overmedication" leads to a rise in "positive correlations. There is a growing systemic risk posed by rising "positive correlations.

Since the GFC (Great Financial Crisis), as indicated by the IMF in their latest Financial Stability report, correlations have been getting more positive which, is a cause for concern:

- source IMF, April 2015

This "overmedication" thanks to central banks meddling with interest rates level is leading to what we are seeing in terms of volatility and "positive correlations", where the only "safe haven" left it seems, is cash given than in the latest market turmoils, bond prices and equities are all moving in concert.
"Positive Correlations" is a subject we touched in our conversation "Misstra Know-it all" back in September 2013 and we referred to Martin Hutchinson's take on these correlations:
"Negative real interest rates are correlated both with a rise in stock valuations (because dividend yields decline) and with a rise in earnings themselves, as the corporate cost of capital declines. Earnings are now at record levels in relation to US GDP, two or three times the deflated level that would be suggested by the current anemic rate of growth. However valuations continue to increase in relation to these inflated earnings, driving stock prices into the stratosphere. Since central banks worldwide are now pursuing the same easy-money policies as the Bernanke Fed, the same correlations are appearing elsewhere, with the exception of the majority of emerging markets, where economic reality remains in play." - source Asia Times, Martin Hutchinson
We could not agree more with the above. Regardless of their "overstated" godly status, central bankers are still at the mercy of macro factors and credit (hence the title of our blog). When it comes to rising risk and the threat of "positive correlations" and Cushing's syndrome we read with interest Nomura European Strategy note from the 6th of May 2015 entitled "At the mercy of macro":

• "The correlation between macro variables (eg, bund yields, FX and oil) and equity market factors (Momentum, Value, Growth, Risk) is now higher than the correlation between macro variables and the market. This is the first time this has happened since 2006 and the difference between the two correlations is the largest than at any point in the post 2000 era.

• The average pairwise correlation between stocks in Europe is close to its post Lehman low. However, we do not think that this heralds the return to some kind of stock-picking nirvana (if such a thing exists).

• The rapid move up in bund yields and EUR-USD reversals of recent weeks has been felt in some sharp factor reversals, most notably an underperformance of Momentum both across the market and within sectors.

• We have moved away from a world where changes in macro variables cause short-term rallies and corrections in the overall index level, to a situation where the same macro variables are the driving force for groups of stocks within the market. So understanding the macro risks is no less important. This perhaps represents a market where the main focus is on the nature of the recovery and the timing and type of earnings growth rather than macro developments prompting a continuous existential crisis as we have seen in recent years. We take this as a positive development." - source Nomura

No offense to Nomura but, we do not take it at all as a "positive development". On the contrary, we think it is representative of the excess of "alkaloids" use leading to Cushing's syndrome and rising instability as posited by the great Hyman Minsky." - source Macronomics, May 2015.

As a reminder:

"The greater the volatility, the greater the disadvantage of owing negative convexity bonds like you find in the High Yield space. In the current low yield environment, both duration and convexity are higher, therefore the price movement lower can be larger..."

There lies the crux of central banks interventions. There is now deeper inter-linkages in the macro economy as well as financial markets globally post crisis.

We would like to repeat what we indicated back in January 2012 in our conversation "Bayesian thoughts" when we quoted Dr. Constantin Gurdgiev, from his post entitled "Great Moderation or Great Delusion":

"when investors "infer the persistence of low volatility from empirical evidence" (in other words when knowledge is imperfect and there is a probabilistic scenario under which the moderation can be permanent, then "Bayesian learning can deliver a strong rise in asset prices by up to 80%. Moreover, the end of the low volatility period leads to a strong and sudden crash in prices." 

Expect more violent moves going forward as a consequence. For us, there is no "Great Rotation" there are only "Great Correlations"...

"History has not dealt kindly with the aftermath of protracted periods of low risk premiums." - Alan Greenspan

Stay tuned!