Andrew Haldane’s spin

Andrew Haldane’s first speech as a monetary policymaker is built around a cricket metaphor.  It seemed de rigeur that any response followed suit.  But all I could manage was this pretty silly and mostly inappropriate pun in the title – the cricketing equivalent of bowling wide of the wicket – which I will explain later.

The speech makes for interesting reading on many counts.

AGH chooses to label himself with hawkish tendencies, in deciding that  those who favour playing on the ‘front foot’ have the balance of the argument.  Batsman who move forwards to the ball are likened to policymakers who seek to raise rates to tame the boom sooner rather than later.  If we think of him as a like-for-like replacement for Spencer Dale, whose inclinations were similar throughout his tenure on the MPC, then this itself won’t change the complexion of the Committee.  This opening contribution chimes with what I digested from his writing during his financial stability days.  From that I took that he put weight on the idea that monetary policy had too narrowly focused on short term macroeconomic indicators, and neglected the evidence of a low-frequency money, liquidity and credit boom.  Extrapolating somewhat, we might deduce that he thought monetary policy, if not a main contributor, at least a partly amiss in not being tighter soon enough before the bust, and that that history risks repeating itself.

By resolving in favour of playing on the ‘front foot’ Haldane judges that the costs of finding out that one has raised rates too late are greater than the costs of discovering they have been raised too early.

For me those arguments tilt pretty clearly on the other side.  The cost of raising rates too early is in tipping the economy into a protracted period – perhaps inescapable – at the zero bound, repeating Japan’s experience since the mid 1990s.  The cost of moving too late is stoking up another boom, and perhaps a bust akin to the one we saw in 2008 [followed by a protracted period….].  Both seem pretty disastrous, but the asymmetry comes in the efficacy of the tools we have to combat them.

The costs of moving too early weigh more heavily for me because of the lack of powerful, well understood means to stimulate the economy in the absence of being able to cut short rates further.   There would be broadly 4 options.  Loosen fiscal policy, engage in further QE, indulge in more forward guidance, undertake credit easing.  Looser fiscal policy on any significant scale would be possible (politically, supposing that the time would come after the expiry of the Coalition), but not particularly desirable with finances already strained.  For me, QE, despite all the intriguing event study analysis (and not all of it given the central bank gloss), remains an uncertain instrument.  We don’t know if its effects persisted;  or if they lowered spreads on risky rates;  even if they did we don’t know for sure that this amounts to observing a policy that has imparted a benefit [QE has costs].  We also don’t know for sure that its effects weren’t simply about causing changes in private forecasts of future central bank rates.  Also, we already have large stocks of assets on the central bank balance sheet:  could we double them without generating (albeit irrational) fear about the integrity of monetary and fiscal policy?  Forward guidance might be possible, despite all the missteps in the last year, but rates are already low right along the curve, and when the time came to manage them lower, they might be squeezed even harder against their expected future floor.  More credit easing would be possible and desirable, but the existing FLS has not been a spectacular success (rubbishing another of my forecasts).  Other variants could be contemplated, (large scale purchases of corporate debt, for example) but these are untested and bring their own risks (financial and reputational).

By contrast, if MPC find that they have not raised rates quick enough, there are better options.  Fiscal policy can tighten and assuage market worries about the longer term trajectory of the debt/GDP ratio to boot.  Short rates can be raised quickly and by as much as the MPC choose.  The MPC could also engage in ‘tightening forward guidance’.   The BoE can close the FLS.  And the FPC can tighten its macroprudential instruments (granted those also untested as yet).

The speech is interesting is in its adding to the accumulation of words about the current forward guidance.  As with Mark Carney’s speech, this one carries the same ambiguity.  Is he speaking for himself, or the Committee?  Where do these words leave the framework for monitoring the commitment not to raise against the 17 indicators?  Unpicking the cricket analogy, where are the front and back foot positions to be judged relative to?  Would a batsman following the market trajectory for short rates be someone playing off the middle foot?!  One meaning in my daft cricket pun ‘spin’ refers to the trickery of a bowler who bowls with a flick of the wrist to make the ball spin to make the batsman uncertain what will happen when it bounces.  A spin bowler injects uncertainty into the confrontation.  This meaning doesn’t really work for me.  AGH doesn’t inject uncertainty, but I don’t think he exactly resolves it either.

Another reason I found the speech interesting, is in wondering how AGH would address the task of cranking the monetary policy handle.  His speeches as Executive Director for Financial Stability often touched on how the crisis had shown that our mainstream DSGE models and the economics canon more generally were misguided.  His authoring of the introduction to the report by the Post Crash Economics group on the state of the economics curriculum, for example, has been interpreted this way.  The MPC’s current model of choice is as plain vanilla a DSGE model as you can get.  It doesn’t even have a financial sector in it.  MPC forecasts are anchored, therefore, by a model that one must presume AGH thinks is not fit for purpose.   The staff and MPC are not fools of course.  The financial sector is left out, one infers, because they think existing ways of modelling it cause more problems than they solve.  Too much complexity eventually makes a model unmanageable.  Financial frictions models don’t fit everyone’s stories about the crisis.  The stories they tell that we do like can be incorporated in the forecast in other ways:  each forecast is overlaid with healthy doses of judgement.  But, if you think that your core model is fundamentally flawed, how do you know how to get from its predictions to predictions that are not?

Haldane alludes to a large program of investment in technology and people drawing on the analogy of the Met Office, which upped its game after the mistakes forecasting the 1987 storms.  This is to be heartily welcomed.   But good luck with producing something usable in the next twenty years.  A lot of people have been working in the top 30 economics departments in the world on the same task, absorbing funding many multiples of the increment that the BoE will commit, and paying much higher salaries too, and not only since the crisis, but for many years before it, and the profession is where it is.  Meanwhile, what should the MPC use to figure out the effects of its forward plan for interest rates and the Asset Purchase Facility balances?  The analogy with the Met Office is provocative, but I don’t think it works.  Senior staff there told me that the small-scale physics was pretty well understood.  What was needed was greatly expanded computing power so that their understanding of the small-scale physical laws could be used to confront the difficulties of large-scale dynamic chaos.  In economics, we don’t yet understand the small-scale economics.  We don’t have settled models of how people comprehend the problems they face, and how they go about solving them.

The second meaning in my daft cricket pun ‘spin’ is the one associated with modern politics:  putting a favourable gloss on a policy.   I applaud the generously funded modelling-revisionism in progress at the Bank, but I bet that we will have to be more patient for usable results than AGHs words suggest.  In the meantime, the MPC could be more openly pluralist in ways that exploit existing technologies.  But that is for another post.