Here is a survey of primary dealers from early and late June:
As usual, there is fodder for more than one point of view.
When asked in early June why rates increased in May (question 5), dealers gave low importance to inflation and economic growth and high importance to changes in Fed posture and uncertainty about Fed policy.
Before the June meeting, dealers expected tapering to begin in December (question 6). After the June meeting, they expected tapering to begin in September/October (question 6, appendix).
The conventional point of view is that the Fed lowers interest rates by buying bonds, so that this reduction in the expected rate of bond buying would cause rates to rise.
My contrarian opinion includes the notions that:
(1) the Fed has been too tight since 2007
(2) especially in the current context, the liquidity factor of Fed open market operations is vastly overemphasized, and the inflation factor is much more important, especially at the long end of the yield curve (which is the only end fluctuating right now), and further, that since the Fed has been too tight, we would expect looser policy to increase inflation expectations and expectations of real economic growth, ceteris paribus, so that when the Fed buys bonds, we should see rates rise.
(As an analogy, I would compare the Fed manipulating the money supply by buying bonds with created money to the Fed manipulating the rental market by buying apartments and upgrading them to upscale condos. (Remember prices fall when rates rise.) The conventional approach to the idea seems to be that the Fed would be causing rental rates to increase because they would be pulling apartments out of the market. The typical approach is that upgrading the condos takes time, so that initially rental prices do increase, and eventually they fall as the effect of the condos is felt. First, I think that empirically since 2007, at least, the liquidity effect has been fleeting, at best. Secondly, what we have now is the Fed prospectively announcing changes in their bond buying activities, and pundits are explaining market movements as a result of future expected liquidity effects from OMO the Fed hasn't even completed yet. I can't imagine how a coherent liquidity effect could play out as a forward looking phenomenon, but I admit that I haven't reviewed the academic work on the matter.)
So, conventional Fed interpretations would say that the Fed will gear down QE3 more quickly than previously thought, and that this caused rates to rise, due to some notion of a forward looking liquidity effect. That is what the primary dealers appear to be saying in this survey.
The fact that equities declined immediately after the June Fed policy announcement (around June 19), is being taken as evidence of this interpretation. But, I will point out for those who believe that the Fed is being recklessly loose, then a tightening of policy should, if anything, skew toward higher equity prices, as decreased uncertainties about future inflation would improve outlooks on real economic growth.
But, I think there are some interesting nuggets in that survey that bear me out. First, I'll note that in question 3, which asks about expected future Fed Funds rate levels, the expected rates remain basically unchanged. The rate in 2017 is 3 to 3.25% both before and after the Fed announcement. Actual market rates (adjusting from June 2017 Eurodollar futures) moved in May from about 1.5% to 2.25%, when this question was first asked, and moved to around 3% after the Fed announcement. So, if the dealers held the marginal market opinion about these rates, why was the market discounting June 2017 Eurodollars by almost 1% in late May? And why did that discount go away?
Other than the expectations for earlier tapering, the only other change I saw in the survey was a decrease in uncertainty about the Fed's balance sheet. Question 8 shows that on the range of the possible size of the SOMA portfolio at the Fed at the end of 2014, probabilities for a portfolio under $3.5 trillion remained the same, but at the other end of the probability distribution, the probability of an exceptionally large balance sheet decreased, due to more certainty about balance sheet remaining under $4 trillion.
So, the dealers themselves, while saying that the change in the tapering schedule somehow created a huge increase in interest rates across the yield curve, also say that the change in the tapering schedule (1) didn't change their long term Fed Funds rate expectations, (2) didn't increase the risk that the Fed would shrink the balance sheet any faster than it might have before, (3) and did reduce the level of uncertainty about the Fed balance sheet remaining too large.
So, the dealers' and the Fed's own opinions aside, I will not be moved. I propose that the dealers are wrong - Fed uncertainty was making rates too low. Improving economic indicators and a slightly tightening Fed with less Fed uncertainty meant inflation expectations decreased but real growth expectations increased from April to late June, with rates boosted by a reduction in uncertainty about the Fed balance sheet. (Rates may have also been boosted by central bank developments in China, Japan, and Europe.)