Brad DeLong says that I confuse him, but nicely places the blame on Federal Reserve Chairman Ben Bernanke. And, truth be told, I have to admit to some confusion, as it it difficult to see how we arrived at this whole tapering debate in the first place. The data simply doesn't lead you to it. So how did the Fed get here?
It is difficult to say the path of activity has meaningfully accelerated. For all of the drama that can surround the quarterly GDP report, the economy has been growing at a pretty steady pace since the recession ended:
And the improvement in the job market in recent months is really less than meets the eye. The twelve-month moving average of job growth is pretty stable and arguably lackluster:
But at 7.6 percent still too high, as too are measures of underemployment. Moreover, one would think that the recent path of inflation would in and of itself end any discussion of tapering:
It looks as though we started the tapering discussion solely on the expectation that the data would soon improve as the impact of fiscal contraction waned. Yet that seems very premature given the levels of inflation and unemployment. There really doesn't seem to be a reason that we needed to have this discussion now rather than at least until after we saw third quarter GDP.
So, again I ask why are we having this discussion? The plot thickens further as Bernanke has altered the nature of the discussion recently. The proposed taper is no longer about reducing policy accommodation. From the Reuters transcript of Bernanke's House testimony:
You talked about the weakness of the economy, I think that's evidence that we need to provide continued accommodation even if we begin to change over time the mix of tools that we use in providing that accommodation.
Bernanke is talking as if the goal is to change the mix of monetary policy but not the level of accommodation, essentially trading some reduced accommodation from ending asset purchases for additional accommodation by extending the forward guidance on interest rates. But why? If the level of accommodation is the same, does the mix matter?
That's an interesting question - does the Fed have research saying the mix matters, and why? I can see two reasons. One is that somehow asset purchases have a more negative distortionary impact. Another is that there exists an internal bias in the FOMC against expanding the balance sheet. Arguably, some elements of both where on display in Bernanke's testimony today:
The second reason for increases in rates is probably the unwinding of leveraged and perhaps excessively risky positions in the market. It's probably a good thing to have that happen, although the tightening that's associated with that is unwelcome. But at least the benefit of that is that some concerns about building financial risks are mitigated in that way and probably make some FOMC participants comfortable with this tool going forward.
Some FOMC participants are not comfortable with asset purchases because the fear financial instability. And maybe they should, but at first blush I am hard pressed to say why shifting the accommodation to interest rates doesn't simply trigger another instability elsewhere. But what I believe is not important; what is important is that some FOMC members believe it, likely enough that reaching consensus is increasingly difficult as the last minutes suggest.
As that discomfort grows, so too does the risk of premature policy withdrawal. So Bernanke needs to find a way to appease both factions within the FOMC, those that want to end asset purchases and those that worry that the exact opposite, an increase in asset purchases, is needed. The solution is this "let's change the policy mix" argument. Give the hawks a bone by giving a time line to ending asset purchases but take it away by waffling on the Evan's rule to push back the timing of the first rate hike.
The trouble is that handing off policy from asset purchases to interest rates is not as simple as it seems. One is a current action, the other is a promise about the future. As is often said, talk is cheap. Actions speak louder than words. Forward guidance is not as powerful as asset purchases in shaping expectations, thus investors,regardless of the promise of a longer period of low short term rates, repriced debt when asset purchases were clearly revealed as temporary .
In other words, if Bernanke had a plan to change the mix of accommodation without changing the level, that plan failed to some degree as financial conditions tightened. To what extent this has a feedback effect that forces the Fed to delay tapering is unknown, but I suspect whatever economic damage may occur is already on the way. For instance, even if higher rates don't undermine housing purchases, they will reduced refinancing activity, something that may have become more important as home prices rise and equity levels increase. But without getting the hawks on board with an expansion of asset purchases, it is difficult to see how Bernanke talks rates back down to 2 percent. Again, talk is cheap. The data would need to start deteriorating much faster to given Bernanke's word additional purchase.
Bottom Line: Trying to reconcile the tapering talk with the data is not easy. I think that's why many analysts missed the direction the Fed was heading. It was somewhat incomprehensible based on the Fed's previous guidance. It could be the effort to change the policy mix but not the level of accommodation was Bernanke's attempt to gain consensus within the FOMC and maintain accomodation, but he may have underestimated the importance of asset purchases in holding rates low.