Puzzling differences between QE at the Fed and the Bank of England

On Tuesday this week, the FOMC decided not to begin to reduce the scale of its monthly asset purchases, as Bernanke had previously signalled it might.  This highlighted for me the starkly different QE policies that the Fed has compared with the UK.

In the UK, the LEVEL of assets purchased will be kept constant at least until unemployment falls to 7%.  In the US, the FOMC will keep the RATE of purchases constant until it sees adequate improvement in the economy.  The FOMC had a similar unemployment criterion to the UK, but appears to have reduced the emphasis placed on this now.  One might think of this as a necessary criterion for both policymakers, but less likely to be sufficient for the Fed than the BoE.

What could possibly explain why one central bank operates a rate of change policy, and another operates a levels policy?

One possibility is that in the US, the effects of a given quantity of asset purchases are mostly temporary, while in the UK, their effects are mostly permanent.  In this case, for the Fed to apply the same stimulus as the UK's MPC, it would have to buy about the same quantity of gilts each month as the UK's MPC has accumulated throughout the entire life of its Asset Purchase Facility. (And perhaps more, in $s, as you might think, especially from what policymakers have said about how the benefits of QE come about, that purchases should be proportional to GDP, and the US economy is roughly x times larger than the UK).  The Fed is not buying anything like that quantity.  So if this were the explanation, (QE has temporary effects in US, permanent effects in UK) then, other things equal, we ought to see the economy in the US entering free-fall, and term premia steadily rising in the US relative to the UK.  Which doesn't seem to be the case.  The Fed is also buying a lot of agency debt.  Since these agencies are now backed by the Government explicitly, this ought to be a close substitute for plain vanilla US government securities.  But they aren't, at least according to the markets.  If the stories about how the policymakers say QE work are to be believed, the Fed's QE purchases ought to have more of an impact than those of the MPC, not less, and there is nothing about this debt that would suggest that its impact effect, whether larger or not, should be more temporary.

Another possibility is that the two central banks plan to do drastically different things with the interest rate instrument to compensate for the very different QE plans, leaving the overall monetary stimulus roughly the same.  This doesn't wash.  Both from the yield curves in the US and the UK, and the forward guidance chatter from each set of policymakers, we can see that the paths for the instruments are roughly the same.

Yet another explanation is that the size of the recessionary impulse the Fed is trying to offset is much greater, and, moreover, is an impulse that is still gathering pace, compared to the UK, where the impulse was smaller, and has peaked.  If the recessionary shock were still growing in the US, this would require ever looser monetary policy, a growing stock of assets held by the Fed.  This seems unlikely to be the explanation.  The fall from pre-crisis peak to trough has been of the same order of magnitude in the two countries.  The recovery from the trough has been more marked in the US.  Strictly speaking, we can't refute this answer by looking at past output, because we don't know what would have happened to the two countries' output in the absence of the monetary and fiscal response.  The impulse imparted by the financial sector one would have guessed to be larger in the UK, where the balance sheets of the financial sector are larger relative to GDP.  So that goes the wrong way.  The US has had to contend with the fiscal cliff, and is looking forward to another one.  But the UK has had less stimulatory fiscal policy throughout.  So this explanation doesn't seem to work.

Of course it's possible that QE has the same effects in both countries, it's just that either one or both policymakers has the wrong view about this, and is doing either too much or too little.  This is hard to square with the facts.  Again, we would expect to see one economy taking off (say the Fed mistakenly believing that QE's effects are temporary), or the UK going into free-fall (the MPC mistakenly believing the effects of QE are permanent).   Unless there were drastically different fiscal policies to compensate.  Which isn't the case.  There's also no great evident divergence in the policymakers views about the effects of QE.   Both central banks have conducted research using event studies of QE announcements, and there is a broad consensus on its impact, and no evidence so far published that the effects are much more temporary in the US than the UK.  Although it would be hard to disprove this, given how many other things affect yields, asset prices, and, further down the chain, the rest of the economy.  Central banks tend to avoid overtly criticising other central banks, so it's conceivable that there is a big disagreement behind the scenes, but I doubt it.

One version of this argument that does fit the facts is that QE has no direct effect whatsoever, except in so far as it signals what policymakers intend to do with the policy rate itself!  This is a rough approximation of Woodford's view, expressed, for example, in his Jackson Hole paper of 2012, and paraphrased in a previous post on this blog.  If this were the case, then two economies hit by the same shock, and responding in roughly similar ways with fiscal and interest rate policy, could sustain totally different QE policies.  (In the same way that they could sustain totally different policies regarding the typeface the policy announcements are made in.)  Bernanke could announce triggers that would keep any derivative of QE constant until certain conditions were met - the level, the rate of change, the rate of change of the rate of change, or the rate of change of the rate of change of the rate of change - and it would not make a blind bit of difference.  Except in so far as what was inferred about the path of the policy rate.

This way of reconciling UK and US QE, also has to answer a lot of questions.  First, why did the event studies of the effects of QE record any effect?  Answer:  either i) because markets mistakenly thought it should, then later realised their portfolios were not much different, and so the effect dissipated, or ii) the effect was only an effect brought about by signalling news about future policy rates.  ii) on its own is tricky, because some of the work showed that the effects of QE were bigger in the asset classes in which the operations were carried out.  But i) and ii) together might work as an explanation.

This nihilistic view about QE, though it reconciles the coexistence of starkly different rates of change (Fed) and levels (BoE) policies, begs further questions.  If the effects of QE are simply due to what is being signalled about future policy rates, how come markets come to infer signals about the policy rate via the change in QE in the US, but via the level of QE in the UK?  According to this view, this is roughly what must be happening, if you think the two economies were hit by roughly the same size and kind of shock, and the path for expected future rates is roughly the same.

Well, we could rationalise a situation like this.  Suppose that those pricing the yield curve know that QE is pretty much impotent, but just try to put themselves in the mindset of the Fed, even if they don't believe it.  They learn that the Fed thinks a given monetary stance is maintained by a given rate of change of QE , and they infer from what the Fed says about its likely path for purchases what the Fed's view of the state of the economy is.  They make some calculation about whether the Fed will ever learn the truth (hypothetically, here, that QE has no effect at all), and work out what the Fed will eventually do to the policy rate.  Those same investors listen to the BoE swear that it's a particular level of assets held that maintains a given monetary stance, and infers from the level it says it wants what it thinks is happening to the economy, and so on.