Mapping the Muddle

“Pi is (still) wrong”
-Vi Hart, Great Rant via Khan Academy

When I was in middle school, I bet some kid $5 bucks I could memorize Pi up to 314 digits in 3 days & 14 hours.  The first day didn’t go so well so I trimmed the goal to the first 314/3.14 (100) digits and downgraded the bet to $1 … then crushed it! …. then immediately lost the dollar somewhere between my desk and the cafeteria.  

Everyone knows that superfluous quasi-intellectual activity is always better than actual homework.  What everyone needs the occasional reminder of is that over-precision, while situationally and superficially impressive, is limited in its practical value.   

Investors are always enamored with well-manicured, well-marketed analytical sophistry. Partial-truths, cloaked in conviction and precision are understandably and timelessly sirenic.  I’m not immune either.  

Anyway, here’s the point – and I have two broader goals this morning. 

The first is part rant, part didactics.  The second is an extension of the first and roadmaps current conditions.   

We get asked for our GDP forecast all the time (i.e. everyday on The Macro Show) which is a primary reason we actually have a dynamic point estimate.  

Keith may (partially) disagree, but I use the point estimate mostly for expositional convenience.  Our framework is slope dependent, not precision point estimate dependent and operates largely agnostic to whatever the actual print is, within a reasonable range.    

The slope of the growth and inflation lines (whether they are accelerating or decelerating) determines what Quad we’re in and the Quad we’re in – in combination with our quantitative overlay – determines the preferred exposures. 

In mathspeak, were more concerned with accuracy than precision from a modeling perspective and if there is an inherent trade-off, high accuracy and lower precision is always preferable to the converse.   

And the short answer to the question is always that if the estimate refresh is not significant enough to change the slope of the line and push us into a different quadrant, then its largely just high-frequency noise.  This is the case most of the time.  

Of course modelling matters and that “high-frequency noise” ultimately layers and cumulates into signal that is the nowcast, but there’s a general tendency to lose sight of the accuracy forest for the precision tree.   

To be sure, there are easy convictions with tangible investment implications to be had. 

For example:

1. Coming out of the industrial/profit recession in 2016, the comps were so favorable that they all but ensured a multi-quarter stretch of growth accelerating.
2. After the longest stretch of accelerating growth ever from 2Q16 to 3Q18, base effects all but ensure a multi-quarter stretch of growth slowing from here.   

Those were/are easy calls that carry profound investment implications.  And, again, the implication is entirely defined by slope accuracy not point estimate precision.  It didn’t matter if growth accelerated +20bps or +30bps in a given quarter from 2Q16 to 3Q18 …. It’s the fact that it accelerated that matters.  

The Chart of the Day below is a mainstay in our Macro deck.  The highlighted annotation on the side represents most of what I care about.  That is, the model gets the slope of the line right ~90% of the time.  

That doesn’t mean we’ll get a hit every time, but it does mean we’ll be consistently taking high probability swings at pitches over the plate … which is the whole point of risk managed decision making.  

Back to the Global Macro Grind ….. 

Consider the recent market price mosaic: 

• After rallying 19% of the oversold Dec 24^th low, the market has meandered here in March.
• The price meandering has been underscored by Volume trends which, over the last two days, have been -36% below the 1-month average. 
• Equity Benchmarks continue to advance but it’s Utilities that are making higher all-time highs while domestic bond yields continue to backslide and global DM bond yields are again re-flirting with negative.  

In other words, given that the dovish pivot, short-covering reflation rally has already played out and the steady drip of global Quad 4 data continues to cultivate both growth concerns and $USD strength which, in turn, serves to constrain a more prolific breakout in the reflation impetus – the market is clearly struggling to answer the “now what?” question.  

More specifically, the market is locked in an internal struggle to express Quad 3 or Quad 4 as the collective expectation ….. that’s why it’s not doing much at the headline level while the internals (sector/style factors) are trying to kind of discount both simultaneously.    

An easy answer would be that we need more dollar weakness, more incremental stimulus from China (and/or time until China can comp out of the prior stimulus) and more evidence of an L-shaped ‘recovery’ in Europe to further drive the reflation narrative.  Implied in that scenario is an assumption that inflation doesn’t pick up materially stateside, forcing a reversal and collective repricing of the tightening cycle and reinvigorating cross-asset volatility.  

That’s a lot of ‘ifs’.  

Remember, also, ‘American Exceptionalism’ characterized most of 2018.  We called it #GlobalDivergences, a 2018 Macro theme which carried with it clear policy and asset attractiveness divergences.  

Synchronized slowdown, however, is more muddled as it becomes a game of relatives.  It also elevates the importance of mapping the evolution of 2^nd derivative global macro conditions.  

For example, if everyone is slowing but the U.S. is slowing less fast, then dollar weakness is not a reasonable baseline expectation.  

And if significant and/or protracted dollar weakness is unlikely, it carries implications for how long the reflation rally can persist and constrains the outlook for EM assets, commodity prices, etc … 

Here's how we’ll be mapping the muddle. 

1. Domestic Growth is going to decelerate and for multiple quarters.  Current conditions in combination with comp dynamics make that about the biggest gimmee you’ll ever get.  
2. What’s less certain – for all of the dynamics highlighted above - is the path of inflation.    
3. The dollar will remain a fulcrum factor in defining the path of inflation and the global growth path (and the path of the EU and China relative to the U.S.) will remain a defining factor for the dollar.
4. To the extent global Quad 4 persists through most of 1H19, dollar strength/stability would continue to weigh on import prices, commodity prices and goods prices more broadly.
5. That could pull the Quad 3 forecast closer to the Quad 3/Quad 4 line and push out the timeline on a more conspicuous traversal into domestic stagflationary conditions.  

We don’t claim to have a monopoly on good ideas or precision answers to every question. 

We do, however, have a process for mapping and probability weighting evolving conditions in the perpetual service of stalking asymmetric risk.  

The above represents a distilled version of our perspective on current conditions, what we’re monitoring and what we think will matter to decision making as we traverse the back side of the growth curve domestically.  

Our immediate-term Global Macro Risk Ranges (with intermediate-term TREND signals in brackets) are now: 

UST 10yr Yield 2.56-2.68% (bearish)
SPX 2 (neutral)
RUT 1 (bearish)
NASDAQ 7 (bullish)
Utilities (XLU) 56.58-58.85 (bullish)
REITS (VNQ) 83.02-86.52 (bullish)
Housing (ITB) 33.68-35.53 (bullish)
Energy (XLE) 63.86-67.18 (bullish) 
Shanghai Comp 2 (bullish)
Nikkei 20911-21801 (bearish)
DAX 111 (neutral)
VIX 12.75-17.51 (bullish)
USD 95.90-97.70 (bullish)
EUR/USD 1.11-1.13 (bearish)
Oil (WTI) 55.87-58.92 (bullish)
Nat Gas 2.73-2.92 (neutral)
Gold 1 (bullish)
Copper 2.86-2.95 (neutral)

To high accuracy, high precision and highs in the 60’s on tap for next week, 

Christian B. Drake
U.S. Macro analyst