Tim Duy's Fed Watch

Was Kevin Warsh Really A Fed Governor?

Former Federal Reserve Governor Kevin Warsh’s column in Tuesday’s Wall Street Journal was so riddled with errors and misperceptions that it is hard to believe he was actually a governor.

Warsh wants the Fed to announce a “practicable long-term strategy and stick to it,” claiming they have offered many such plans but never stuck to them. I don’t agree. The Fed has a plan, but Warsh just refuses to see it.

The former governor’s first critique:

A year ago around this time, the U.S. stock market fell about 10%. The Fed reacted precipitously, reversing its announced plan for 2016 of four quarter-point rate increases. But when prices rallied near the end of the year, the Fed decided it wouldn’t look good to let the moment pass without raising rates. It raised its key interest rate by a quarter point in December.

The Fed did not reverse its announced plan. The rate forecast contained within the Fed’s Summary of Economic Projections is just that, a forecast, not a plan. Incoming data triggered a revision of that forecast. And, contrary to the narcissistic belief of many market participants, it wasn’t all about them. Falling equity prices were just one of many data points that changed the course of policy. The Fed faced a very real slowdown in activity. Andrew Levin, former Fed economist, for instance, described the economy as operating at stall speed. Note that output growth slowed markedly during 2015 and into 2016:

The unemployment rate stalled out:

And inflation remained tepid:

If the Fed updated their forecast, it was for good reason.

Warsh follows with another instance of the Fed supposedly reversing course:

In late October, Fed Chair Janet Yellen expressed willingness to run a “high-pressure economy” to push the unemployment rate lower and inflation higher. Yet in a speech two weeks ago, she said that allowing the economy to run “persistently ‘hot’ would be risky and unwise.”

Yellen never expressed a willingness to run a high-pressure economy. That was always a complete misrepresentation of her comments. In that speech, she was simply proposing a research agenda for macroeconomists, including the topic of the influence of aggregate demand on supply. Had anyone actually read the speech (and I have to assume Warsh did not), they would see that she did not provide any policy proposals. Her subsequent comments were nothing more than an effort to set the record straight, not a shift in position.

Warsh uses the above episodes to claim that the Fed lacks a strategy:

Changes in judgment should be encouraged, but they ought to indicate something other than day trading or academic fashion. They must be rooted in strategy. Otherwise, the real economy winds up worse off…

…The Fed’s technocratic expertise is no substitute for a durable strategy. This make-it-up-as-you-go-along approach causes many Fed members to race to their ideological corners, covering themselves as hawks and doves…

The problem here is that the Fed does have a strategy, but Warsh refuses to see it. Specifically, incoming information alters the Fed’s economic forecast and, in accordance with a basic Taylor Rule, the Fed’s rate forecast with the goal of meeting the dual mandates over the medium term. Indeed, San Francisco Federal Reserve economists Fernanda Nechio and Glenn Rudebusch show that the Fed altered its rate forecast systematically in response to incoming data in this manner. That data brought the Fed’s original forecast for four rate hikes down to the actual one rate hike.

There is indeed a strategy. It is not the Fed that is too focused on the near-term. It is Warsh that is too focused on the near-term.

Warsh proposes five reforms for the Federal Reserve, beginning with:

First, the Fed should establish an inflation objective of around 1% to 2%, with a band of acceptable outcomes. The current 2.0% inflation target offers false precision. According to the Fed’s preferred measure, inflation is running at 1.7%, only a few tenths below target. The difference to the right of the decimal point is too thin a reed alone to justify the current policy stance. It also undermines credibility to claim more knowledge than the data support.

This one reveals Warsh’s true intentions – the current inflation target does not support his desire for higher rates, so he wants to move the target! Moreover, the current target does not offer false precision. No one at the Fed believes they can consistently hit two percent. And being below two percent has not stopped them from raising rates. In practice, the Fed will tolerate misses within a reasonable (25bp) range around two percent as long as forecasted inflation is trending toward target. That’s the medium-term strategy Warsh claims to be so concerned about.

Next:

Second, the Fed should adjust monetary policy only when deviations from its employment and inflation objectives are readily observable and significant. The Fed should stop indulging in a policy of trying to fine-tune the economy. When the central bank acts in response to a monthly payroll report, it confuses the immediate with the important. Seeking in the short run to exploit a Phillips curve trade-off between inflation and employment is bound to end badly.

It is a misperception that the Fed acts impetuously on the most recent data. They use that data to update their forecast in a systematic fashion (see above). I generally excuse most people from not understanding this distinction. It is a difficult concept and, quite frankly, one that the Fed does a poor job communicating. Warsh, however, has no such excuse.

For his third reform:

…the Fed should elevate the importance of nonwage prices, including commodity prices, as a forward-looking measure of inflation. It should stop treating labor-market data as the ultimate arbiter of price stability…A material catch-up in wages after a long period of stagnation need not trigger a panicky response.

I don’t think Warsh understands the foundations of Fed’s approach to inflation forecasting. From Yellen’s lengthy discussion of the topic in September 2015:

To summarize, this analysis suggests that economic slack, changes in imported goods prices, and idiosyncratic shocks all cause core inflation to deviate from a longer-term trend that is ultimately determined by long-run inflation expectations. As some will recognize, this model of core inflation is a variant of a theoretical model that is commonly referred to as an expectations-augmented Phillips curve. Total inflation in turn reflects movements in core inflation, combined with changes in the prices of food and energy.

Wages aren’t in that description. Wages are primarily a guide to estimating full employment (economic slack). Rising wage growth indicates the economy is approaching full employment. Wage growth in excess of the inflation target plus productivity growth raises warning signs that the economy is operating beyond full employment. Also, commodity prices are included as idiosyncratic shocks. Finally, the labor market data is very clearly not the ultimate arbiter of price stability. In the long-run, inflation expectations are the ultimate arbiter of price stability.

Warsh’s next reform:

Fourth, the Fed should assess monetary policy by examining the business cycle and the financial cycle. Continued quantitative easing—which Fed leaders praise unabashedly—increases the value of financial assets like stocks, while doing little to bolster the real economy. Finance, money and credit curiously are at the fringe of the Fed’s dominant models and deliberations. That must change, because booms and busts take the central bank farthest afield from its objectives.

Note that earlier Warsh complained that the Fed reacted to the financial cycles – easing policy when equity prices were falling and vice-versa. Now he wants the Fed to react to those cycles? In actuality, the Fed does take financial considerations seriously. See, for example Governor Jerome Powell here. Vice Chair Stanley Fischer here. Governor Daniel Tarullo here. Go back to the work of former Governor Jeremy Stein. And with regards to quantitative easing, the Fed would argue that their actions have indeed bolstered the real economy. And while busts in particular do take the Fed far away from its mandate, so too would strangling the economy to address a theoretical financial risk. Incorporating a financial stability term in the Taylor Rule is easier said than done.

Finally:

Fifth, the Fed should institutionalize its new strategy and boldly pursue it with a keen eye toward the medium-term.

As I noted earlier, the Fed does have a strategy, the focus of that strategy is the medium run forecast, and the Fed changes their behavior in a systematic way to pursue that strategy. In short, the Fed’s already acts in accord with this supposed “reform.” Move along, folks, nothing to see here.

Bottom Line: If you want to understand the Federal Reserve and monetary policy, I don’t think reading Warsh’s op-ed gives you much to work with.

Alarm Over the Fed's $4.45 Trillion Balance Sheet Is Silly

The Federal Reserve is laying the groundwork for shrinking its $4.45 trillion balance sheet. But don’t panic yet, bond traders. This isn’t 2013.

Continued at Bloomberg Prophets...

FOMC Preview

The Fed will take a pass at this week’s FOMC meeting. The median policy participant forecasts just three 25bp rate hikes this year and incoming data offers no surprises to force one of those this month. March, however, remains in play.

The three forecasted rate hikes is not a promise. It could be one hike or could be four or more. The actual outcome will depend on the path of actual economic outcomes and what those outcomes imply for the forecast.

The Fed is aware that crosscurrents in the economy – such as potentially significant changes to fiscal and economic policy – create substantial uncertainties about the course of monetary policy this year. From the most recent minutes:

…many participants emphasized that the greater uncertainty about these policies made it more challenging to communicate to the public about the likely path of the federal funds rate.

Translation: The Fed’s crystal ball is as cloudy as everyone else’s, but that’s hard to explain. For example, the potential positive demand shock from expected deficit spending could be overwhelmed by a potential negative supply shock from an increasingly xenophobic Trump Administration.

What does this mean for March? Currently, market participants place low odds of a March rate hike. The underlying bet is that if the Fed moves three times this year, the most likely timing will be June, September, and December. I think this is reasonable; bringing March into that mix requires a change in the tone of the data.

Specifically, to pull a rate hike forward, the Fed needs evidence that either inflation is firming more than anticipated or that unemployment is more significantly undershooting its natural rate. Both would be cause for concern for policymakers. But, in practice, given the inflation inertia evident in recent years, the labor market would most likely be the driving force of behind a March rate hike. From the minutes:

Several members noted that if the labor market appeared to be tightening significantly more than expected, it might become necessary to adjust the Committee's communications about the expected path of the federal funds rate, consistent with the possibility that a less gradual pace of increases could become appropriate.

The December labor report was largely consistent with the Fed’s forecasts, and thus will have little impact on the March meeting. The same is true for the GDP report for the final quarter of 2016. Notable is that domestic demand has held up well the last three quarters:

They will also be heartened that equipment investment, broke a string of four consecutive negative quarters with a 3.1 percent gain. Also, note that core durable goods orders are finally back to making year over year gains:

This too is consistent with the small uptick in growth anticipated by the Fed for 2017.

But we still have plenty of data between now and March. In particular, watch incoming data and how they impact the forecast of key variables such as unemployment and inflation. The Fed will pay close attention to:

• Nonfarm payrolls. They expect payroll growth to continue slowing to something close to 100k a month. A re-acceleration would raise eyebrows on Constitution Ave.

• The unemployment rate. The Fed’s estimate of the natural rate of unemployment firmed over the past year. Hesitation to drop it lower means that surprise falls in unemployment would prompt more aggressive Fed action.

• Faster wage growth. Some policy makers argued in December that subdued wage growth gave them more time to respond to an unemployment overshoot. But wage growth accelerated in December to 2.9% annually, the highest pace since 2009. Watch this space (and see this from Bloomberg on competition for workers in the fast food industry).

• Inflation numbers. Although, as noted earlier, inflation has been fairly inertial, that could change at the economy settles further into full employment/potential output.

• Acceleration? The Fed is not anticipating a large acceleration in activity this year, so any indication that activity is picking up more than expected will be watched with wary eyes.

At this point I still do not anticipate a March hike. And note that a March move doesn’t guarantee a faster pace of rate hikes; it could be largely pre-emptive, just displacing a subsequent rate hike. But if they could justify a March move and you were anticipating two to three rate hikes this year, you should probably be thinking of three to four. Not to mention some action on the balance sheet added to the mix.

Beware of Hard-Money Advocates

Federal Reserve Chair Janet Yellen made it clear in a recent speech that monetary policy wouldn't immediately be affected by the changing of the guard in Washington. It isn't the short-term that’s worth worrying about, but the long-term and the potential for new Fed governors to be neither objective nor divorced from political pressures...

Continues at Bloomberg Prophets...

Quantifying The Changing Rate Forecast

In my last post, I asserted:

The actual amount of tightening will ultimately depend on the evolution of the forecasts for unemployment and inflation. If the expectation for unemployment drifts lower for this year, for instance, the median dots are likely to shift higher to ensure that the Fed continues to meet its mandate.

Can we quantify the impact of a changing economic forecast on the projected amount of tightening this year? Yes, using the methodology of Federal Reserve Bank of San Francisco economists Fernanda Nechio and Glenn Rudebusch. In a recent article, they argue the change in the Federal Reserve’s 2016 projected rate increase from 100bp to 25bp was consistent with a simple extension of a Taylor-type policy rule, specifically:

Funds rate revision = neutral rate revision + (1.5 × inflation revision) – (2 × unemployment gap revision).

Recall that the interest rate projections contained in the Fed’s Summary of Economic Projections (SEP) are not policy commitments. They are forecasts that we should expect to change with evolving forecasts of key variables, notably inflation and unemployment. The Fed’s credibility should not be judged on the accuracy of its rate forecast. It should be judged on its ability to meet its mandate. Actual policy should shift relative to the rate forecast as economic conditions change.

We can look to the December 2016 SEP as an example of the Nechio-Rudebusch approach. The Fed’s median rate forecast for 2017 rose 0.3 percentage points relative to the September SEP (Note: There is a rounding issue here. Effectively, the forecast changed from two to three 25bp hikes). The median neutral rate estimate (the longer run forecast of the funds rate) rose 0.1 percentage points. The inflation forecast (Nechio-Rudebusch use core inflation) was unchanged. The unemployment forecast fell 0.1 percentage points while the estimate of the natural rate of unemployment (the longer run forecast of the unemployment rate) remained unchanged. Thus the Fed revised down the unemployment gap estimate by 0.1 percentage points.

Applying the Nechio-Rudebusch policy rule:

0.3 percentage points = 0.1 percentage points + (1.5 x 0.0 percentage points) – (2 x (-0.1 percentage points)

In other words, the changing economic forecast for 2017 explains the magnitude of the change in the 2017 rate projection. Thus, we should watch incoming data for its impact on the forecasts for key variables to estimate its impact on policy.

In practice, we might expect minimal revisions of the longer run rates of interest and unemployment over the course of 2017. The estimate of the neutral interest rate declined substantially in 2015 and early 2016, but I suspect that pattern will not repeat in 2017. The estimate has already held fairly steady since the June 2016 SEP. Also note that the Laubach-Williams estimates of the natural rate of interest are now edging upward:

The era of declining estimates of the natural rate of interest may be over. Likewise, the estimate of the natural rate of unemployment remains at the March 2016 SEP. Assuming these parameters remain constant in 2017, the variation in the fed funds rate projection will depend on the unemployment and inflation and inflation forecasts.

As a practical example, I view the December employment report as consistent with the December SEP economic forecasts. The pace of underlying job growth continues to slow toward a range that is likely consistent with a steady unemployment rate after the demographic impacts on labor participation reassert themselves. This suggests the Fed’s forecast for a fairly small (0.2 percentage points) fall in the unemployment rate is largely unchanged. Hence, the employment report should not impact the median rate forecast for 2017.

Bottom Line: The Fed’s policy stance shifts in consistent manner. Important to understanding these shifts is estimating the impacts of incoming data on the Fed’s medium term forecast. In my opinion, the policymakers spend too little time discussing the impact of incoming data on their forecasts, leading to the perception that policy is more backward than forward looking. The above example illustrates, however, the importance of the latter for monetary policy.

Solid Employment Report Keeps Fed On Track

The labor market finished out the year on a solid note. Solid, not spectacular, and largely consistent with the Fed's expectations. Consequently, the final employment report for 2016 should not impact the Fed's median forecast for 75bp of rate hikes in 2017. 

Payrolls rose 156k in December and jobs gains the previous two months were revised upwards by 19k. While good numbers, job growth continues to slow:

Since January 2015, the 12-month moving average of monthly job growth slowed from 262k to 180k. Still, that remains greater than the pace necessary to hold the unemployment rate constant once the demographic impacts again dominate the cyclical factors (Federal Reserve Vice Chair Stanley Fischer estimates that number to be 65k-115k). But the economy continues to trend toward that pace.

Supported by an increase labor force participation, the unemployment rate ticked up to 4.7%, holding just below the Fed's estimate of the natural rate of unemployment:

Measures of underemployment are now again showing signs of improvement, albeit the pace of improvement has slowed along with the pace of job growth:

The pace of wage growth accelerated to 2.9%, the highest rates since 2009:

Overall, the report should win hearts and minds on Constitution Ave. The economy looks to be tracking exactly where the Fed expects it to go, with job growth slowing sufficiently such that the unemployment rate holds steady just below full employment. Such a situation would allow for continued improvement in measures of underemployment while maintaining healthy but not excessive pressure on wage growth. In contrast, recall the concerns about about undershooting the natural rate of unemployment that surfaced during the December FOMC meeting. From the minutes:

...In discussing the possible implications of a more significant undershooting of the longer-run normal rate, many participants emphasized that, as the economic outlook evolved, timely adjustments to monetary policy could be required to achieve and maintain both the Committee's maximum-employment and inflation objectives.

 ...Several members noted that if the labor market appeared to be tightening significantly more than expected, it might become necessary to adjust the Committee's communications about the expected path of the federal funds rate, consistent with the possibility that a less gradual pace of increases could become appropriate...

The economy is now at a point where a sudden boost in activity would prompt the Fed to accelerate the pace of rate increases. This employment report, however, suggests this isn't happening just yet. 

One note of caution, though. Manufacturing employment rose in this latest report by 17k, the largest gain since January 2016. This comes on top of improved data from the manufacturing sector:

This serves as further evidence that the inventory correction process over the past year has run its course. Note also the improvement in the service sector in recent months:

This suggests to me that risks for growth and hence rates are currently weighted to the upside.

Bottom Line: A solid report largely consistent with expectations among monetary policymakers. Hence it should have little impact on interest rate forecasts for the coming year. But watch out for upside risks to the outlook; the economy gained some traction in the final months of 2016. It is reasonable to believe that traction will hold in 2017.