Speed Read: The Fed has made three mistakes. Its first mistake is in continuing to dampen volatility knowing full well that it cannot be kept low, only postponed. In doing so, it has ignored the US ‘monetary trinity’ – a balance between the front-end, the 10-year yield and the USD. Second, it has misread the signal from aggregate inflation and therefore run pro-cyclical monetary policy. But the one that has the potential to damage policy and communication is the assertion that the Fed, and Janet Yellen in particular, have thrown their weight behind is arguing the Equilibrium Real Interest Rate is zero. The error lies in the ERIR. We’ll call it r*, but we couldn’t resist that.
Market implications: Yields and the USD could rise faster if the Fed hikes by less, not more. A solid US economy and no large deflationary impulse outside the US are pre-conditions.
“So Baldrick, what you’re telling me is that one shade of blue you’ve never seen before is a little bit darker than another shade of blue you’ve never seen before either?” – The Black Adder
Whether it is the Fed or the PBoC, the people running monetary policy at such prominent institutions are incredibly smart. But they can make mistakes if their radar goes awry, or if their behaviour changes.
The ‘common agency’ problem at the Fed. Bernheim and Winston (1986) argued that when more than one person (a “principal”) is given more than one task to manage on behalf of someone else, they end up acting in a more risk-averse fashion. That risk-aversion increase with the number of principals and tasks. The Fed has seen this very problem take hold – with uncertainty creating clusters within the FOMC and their tasks increasing from a trade-off between price and employment to include financial stability as well.
The Fed has made three mistakes thanks to a misdirected radar and the common agency problem.
I. In trying to dampen volatility, they have shifted it into the future (again)
In February 2015, Vice Chair Stanley Fischer suggested that the Fed may need to add a little mystery in its communication with markets. “There’s no good reason that I can see for us to telegraph every action that we have to take”. Instead, the Fed has tried its hardest to go in exactly the opposite direction, to try and dampen volatility at every cross-road by delaying hikes, and trying to talk down the dollar and yields, and therefore market expectations. It has thus ignored a key lesson of the last decade – it cannot lower volatility permanently, it can only displace it intertemporally.
That’s where the US ‘monetary trinity’ kicks in. Conventional wisdom would argue that the yields and the US dollar would behave as the Fed does. But if the Fed has some reaction function and yields and FX are endogenous, surely all three of them are jointly determined – particularly over longer stretches of time.
The Fed (setting the local price of money), 10-year yields (setting the global price) and the US dollar (relative price) collectively are the ‘monetary trinity’ in the US that determine the monetary stance. When the Fed stays ‘behind the curve’, the yield and the USD have to adjust faster.
Should the Fed have hiked much earlier than it did? Surely promising one hike and no more a year would have done very little damage, but collectively it would have generated a better balance. Had the hikes come, perhaps the Dollar Tantrum might have not occurred – we explain below.
So why isn’t this the big mistake? Only because the economy hasn’t been derailed yet. The Fed still has an opportunity to rectify its error. But if the US and other large economies maintain course, we believe that anything less than 3 hikes this year could push the 10-year and the USD much higher, not lower.
II. Misreading Inflation: The US wasn’t disinflating
The FOMC looks below the surface of real activity but never (at least publicly) disaggregates inflation. Labour market activity, manufacturing, housing… its all very well done. When it comes to inflation, however, the only points of public discussion are about aggregate inflation and core inflation at times.
Had the Fed had focused on one simple decomposition of inflation – domestically generated services vs mostly imported goods, the source of the disinflationary impulse in the US economy would have been abundantly clear (Ex 1). Domestically generated inflation (ex-healthcare) has been positive and rising for 3 years now thanks to strong domestic activity. It is the externally exposed parts (shale, manufacturing, exports) that have done badly and deflated.
Pro-cyclical policy: The overwhelming urge to dampen volatility and the mis-reading of inflation together generated negative and falling real interest rates for the part of the economy that was doing well already, and positive and rising real interest rates for the externally exposed segment that needed help.
But this too can still be rectified: Imported inflation is now rising too, but overall inflation is still low enough that past errors can be rectified.
III. But arguing that r* is zero is the one argument that could hurt
One of the first models I estimated as a young (ok, I’m exaggerating a bit) analyst at my last stint was Laubach and Williams’s (LW here) natural rate model. It is THE model that everyone from Janet Yellen to Larry Summers quotes when it comes to knowing where the natural rate it.
There’s one serious problem – the LW model that generates the zero r* estimate is based on a key assumption that has little support empirically.
The LW model uses Frank Ramsey’s framework to assume that both potential output growth and r* are driven by a common factor – say productivity. This is a sound theoretical proposition to make, but Hamilton et al (2016) show decisively that data going back to the 19th century produce no strong link between growth and real interest rates. They also conclude that r* in the US is probably positive and will remain so.
And now, the turn in demography will raise, not lower, r* – at the very least, it has most likely already stopped falling. We have argued as much with Charles Goodhart, our Senior Advisor. Demographics pushed both inflation and the real interest rate lower and made central bankers look like rock stars. The going will be much harder from here on.
If r* isn’t zero and is going to rise, it complicates both policy and communication tremendously.
Arguing that r* is zero has been very helpful. The ZLB and low inflation had limited the extent of monetary accommodation that could be provided, but arguing r *=0 gave a boost to accommodation by shifting the goalposts.
But payback may be on the way. Surely r* moves at a glacial pace, so arguing that it has now rather rapidly become positive is difficult without admitting an error in judgement that would rattle markets even more.
But even worse, if the Fed truly believes that r* is zero and it isn’t (we believe it isn’t), then it will not worry about the increase in inflation too much. That’s the scenario – along with macro stability outside the US – under which yields could really jump higher globally, sowing the seeds of the next recession.
In summary, the difference between knowing which of the two shades of blue is a bit darker, when neither of them can be seen, could determine the future path of policy, yields and this expansion.