Cracks in the Fed

The stock market correction may not be over until it becomes clear the Fed will hold back from tightening further -- either by skipping a hike in December or suggesting it will be the last for a while.  Upcoming inflation data may very well hold the clue.  While the risks are for softer-than-consensus wage inflation data this week, the October Core CPI (due November 14) may be more important in influencing market expectations regarding the Fed.

At this point, Fed officials are adhering to their expectation of gradual tightening into 2020, including a December hike and 3 hikes in 2019.  However, cracks in their determination are beginning to appear, with the acknowledgement by some that inflation remains subdued and potential economic growth could be stronger than expected.  There seems to be growing consensus among officials that future rate decisions will be more data-dependent than in the recent past -- implying not necessarily a 25 BP hike per quarter. 

Last week, news stories headlined Fed Vice Chair Clarida reaffirming a commitment for more rate hikes.  In fact, most of his speech, academic in nature, focused on reasons not to hike.   He emphasized the improvement in productivity growth, the flatness of the Phillips Curve and a lower-than-thought natural rate of unemployment to suggest inflation will remain soft.  He argued for further rate hikes based on the idea that keeping the funds rate below r* --  the theoretical and unobservable neutral rate consistent with steady growth and inflation -- would raise the risk of higher inflation in the future.  

The problem with r* is that it is a "steady state" concept -- economic growth and inflation are assumed to be independent of initial conditions (e.g., current state of the economy, stringency of bank lending standards) and exogenous factors (e.g., fiscal policy, foreign growth) among other idealistic assumptions.  The Fed's econometric model of the US economy has run steady-state simulations, but this required some tinkering and many years of simulation for the model to reach that state.  In the real world, initial conditions and exogenous factors matter.  Currently, fiscal stimulus would argue that r* should be higher than otherwise, while softer foreign growth would argue for the opposite.  Bank lending standards, while easing over the past few quarters, remain tight enough to suggest the funds rate should not be raised further.

In a recent speech, Cleveland Fed President Mester (a hawk) almost admits that forward guidance on the course of the funds rate is no longer relevant.   She says with the funds rate now close to its neutral level, rate decisions will be more data dependent than in the recent past.

Similarly, Atlanta Fed President Bostic thinks the Fed is still a few rate hikes away from a neutral level.  But, he "intends to weigh the risk of acting too swiftly and choking off the expansion against the risk of having the economy overheat and get into a situation with rising inflation and inflation expectations that would necessitate a muscular policy response. My thinking will be informed by the evolution of the incoming data and from what I'm able to glean from my business contacts."  

All the recent Fed speeches have turned cautious because of a recognition that inflation remains subdued despite the strong economic growth and tight labor market.  Upcoming wage/price inflation data may very well be the clue to whether the Fed will pull back from its forward guidance on the funds rate at the December FOMC Meeting.  This week's inflation data risks underscoring this recognition:

Monday --  Consensus reasonably looks for a soft 0.1% m/m September Core PCE Deflator.  The y/y is expected to fall to 1.9% from 2.0%.

Wednesday -- Consensus looks for a speedup in the Q318 Employment Cost Index to 0.7% (q/q) from 0.6% in Q218.  This risks being too high.  While Average Hourly Earnings (AHE)  sped up by 0.1% pt in Q318, it is possible Benefits slowed.  For example, news reports made much of Amazon's boosting its minimum wage, but not all mentioned that it cut bonus payments simultaneously.  Bonuses impact the ECI's Benefits component but have no impact on AHE.

Thursday -- Consensus looks for +1.2% (q/q, saar) Q318 Unit Labor Costs.  This estimate risks being too high.   ULC is calculated as the difference between Compensation/Hour and Productivity.   Roughly a near-trend 3.0% increase in Compensation/Hour looks reasonable, based on the GDP data.  But, Productivity could come in higher than the consensus estimate of an above-trend 2.1%.   A flattish ULC would be important for two reasons.   First, it is the broadest measure of labor costs.  Second, it would show that the pickup in nominal wage inflation just reflects improved productivity and thus is not inflationary.   Clarida made a big point about this in his speech.

Friday -- Consensus looks for a high 0.3% m/m in October AHE.   Risk is to the downside.  Calendar considerations point to 0.1% m/m.   Also, composition shifts could have contributed to the high 0.3% September print and could disappear in October.

Next week's real-side data should not be soft enough to convince the Fed to stop tightening, even if they print below consensus.   

Wednesday -- ADP Estimate risks printing below the consensus estimate of +189k m/m, as there could be payback for ADP exceeding Payrolls in September.  October ADP reversed its September relationship to Payrolls in each of the past 6 years.  So, the risk is that whatever prints will underpredict Friday's Nonfarm Payrolls.

Thursday -- Evidence is mixed with regard to the October Mfg ISM, but favors an increase.   The Richmond Fed fell, but this might be catch-up to the Mfg ISM after it diverged in direction in September.  Also, Richmond Fed could have been impacted by the hurricane.   Mfg ISM does not tend to be affected much by weather factors.  The Phil Fed Mfg Index (pushed ahead one month) points to an increase.  It correctly predicted the direction of Mfg ISM in each of the past 6 months.

Friday --  The consensus estimate of +190k m/m October Nonfarm Payrolls risks being too high.  While the factors that held down September Payrolls -- an unwinding of summer jobs or drag from the hurricane -- should either fall out or reverse in October, there could be a drag from another hurricane in October.  A lot of attention will be placed on the Unemployment Rate, after it fell to 3.7% in September.  Unfortunately, there is no reliable evidence regarding the risks surrounding the consensus estimate of a steady 3.7% in October.