## Paul Fisher: Current issues in monetary policy
Speech by Mr Paul Fisher, Executive Director for Markets and Member of the Monetary Policy Committee of the Bank of England, at the Global Borrowers and Investors Forum, London, 21 June 2011.
* * *
I would like to thank Ronnie Driver for his help in preparing this speech.
Monetary policy is currently a subject of great debate. More so than at any time since the start of the inflation targeting era nearly twenty years ago. That almost certainly reflects the hardships associated with the recession on the one hand and higher inflation on the other, with the resultant squeeze on real incomes affecting the vast majority of us. In this keynote address today, I want to assess some of the reasons why the current conjuncture is proving so difficult for achieving the inflation target. In summary, the nature of the shocks that have hit the economy mean that there is no easy path for monetary policy and the outlook for the economy is especially uncertain. The main message I want to get across, is that there is no magic solution to these challenges. The MPC are trying to set the best path back to the inflation target, but even the best path is an extremely uncomfortable one.
## Why has CPI inflation been so high recently?
CPI inflation has been above its 2.0% target since December 2009, reaching 4½% in April and May 2011. And, as discussed in the Monetary Policy Committee's (MPC) May Inflation Report , it is likely to rise further in the near-term before beginning to fall back ( Chart 1 ). Moreover, the rate of inflation may well stay above target for the remainder of 2011 and 2012.
Chart 1 CPI inflation projection based on market interest rate expectations and £200 billion asset purchases
As I and many of my colleagues have already discussed in various speeches, MPC Minutes and the Inflation Report , the MPC attribute the rise in CPI inflation to three main shocks - the rise in VAT, increases in global commodity prices, and the impact of the fall in sterling since mid-2007.
First, let me consider the increase in the standard rate of VAT to 20%. Mechanically this tax increase could account for around 1½ percentage points of the current inflation rate
although, allowing for the behaviour of retailers in absorbing some of the increase, the marginal impact is probably nearer 1 percentage point. 1
Second, there have been large increases in a number of commodity prices over the past two years or so. Those increases reflected the combination of renewed strength of the world economy, especially in emerging market economies, and the limited ability of the world's commodity producers to increase supply - at least at the same pace as demand. Commodity prices are traded as an asset - they are inherently forward looking. Information about future demand and supply levels ought to be reflected in the current price and in the prices of commodity futures. But the increases from Autumn 2010 onwards were largely unexpected, and have been a genuine surprise for our forecasts. For example, spot oil and gas prices have risen very substantially since their low points in 2009. 2 Spot oil prices are well above the levels implied by futures prices at the time of the February Inflation Report ( Chart 2 ). Gas futures prices have also moved markedly higher since then ( Chart 3 ).
Chart 2 Sterling oil prices (a)
Chart 3 Sterling gas prices (a)
per
Sources: Bloomberg, Thomson Reuters Datastream and Bank calculations.
(a) Futures prices for February and May are averages during the fifteen working days to 9 February and 4 May respectively. Each futures curve assumes that the sterling-dollar exchange rate remains constant at its average during those periods.
Sources: Bloomberg, Thomson Reuters Datastream and Bank calculations.
(a) Futures prices for February and May are averages during the fifteen working days to 9 February and 4 May respectively.
(b) One-day forward price of UK natural gas.
(b) Brent forward prices for delivery in 10-21 days' time converted into sterling.
Third, we have been experiencing the ongoing effects of sterling's fall during 2007 and 2008, when the nominal effective exchange rate fell by around 25% ( Chart 4 ). Estimating the impact of such a shock on CPI inflation, and the time it takes for that impact to come through, is very difficult, not least because the precise scale depends in large part on the reasons for the shock. It is fair to say that the MPC originally underestimated the degree and timing of
1 Since the change was only announced in July 2010, this was a surprise addition to our inflation projections made before that date.
2 These prices impact directly on the retail prices of petrol, gas and electricity, which together account for just under 10% of the CPI basket.
exchange rate pass through into consumer prices. One possible reason for that is that it would appear that the sensitivity of CPI to the exchange rate has been rising over the past decade or so.
Chart 4 Sterling exchange rates
A one-off, autonomous shock to the exchange rate would ultimately change the price level not the inflation rate. But this particular shock probably reflected a reappraisal of real UK economic performance in the light of the financial crisis rather than, for example, interest rate differentials between the UK and its major trading partners. 3 During the decade prior to the crisis - the NICE or non-inflationary, consistently expansionary decade - the UK experienced an unparalleled run of sustained growth, falling unemployment and low, stable inflation. But the economy also became badly unbalanced - overly dependent on domestic demand and with a trade deficit that was steadily expanding relative to domestic output. The real exchange rate change brought about by the depreciation of sterling (and indeed the increase in VAT) reflected a need to rebalance the economy away from domestic expenditure and towards net exports. That real exchange rate shock will cause a change in relative prices, with import prices rising relative to domestic wages and prices. It may take several years for that price level effect to work through fully, generating higher inflation rates for the duration. It is possible, but by no means certain, that this effect has now worked through.
A key and recurring theme in setting monetary policy is that one has to look at the source of the shocks to understand their inflationary consequences and judge the appropriate policy prescriptions. The first question I want to address today is, given our inflation target, what is the right monetary policy response to these shocks? I then want to talk a little about how we make such decisions and, in particular, the role played by models and forecasts.
## How should monetary policy have responded to the recent shocks to inflation?
The economics profession has long accepted that different shocks warrant different policy responses. Assume for a moment that one can perfectly identify the source and nature of a shock to the economy (which of course, one never can). Faced with a demand shock - such as an exogenous increase in consumer spending (for example driven by a change in preferences) - inflation and output would be pushed in the same direction. The monetary
3 For example, policy rates in the United States were being cut well before those in the UK.
policy prescription in such cases is relatively simple - monetary policy just has to lean against the demand shock sufficiently that inflation returns towards the target. Tightening policy helps to reduce the deviation of both inflation (from target) and output (from its sustainable path) albeit with somewhat different time lags. For such a shock the challenge for the policy-maker is one of recognition and timing (because demand will usually react more quickly than inflation to a policy change) rather than direction.
Things are much, much more complex when faced with real, relative price shocks, such as a rise in world energy prices. Such a cost shock pushes up on inflation but simultaneously pushes down output relative to its previous level. Technically it is a negative supply-side shock. That makes policy-making much more difficult - one can only counter the resulting inflation by exacerbating the fall in output. Moreover, the pressure of demand on inflation becomes much harder to estimate when both actual output and the level of sustainable supply are moving around at the same time.
The three price shocks share a common feature - they all represent real shocks to relative prices. Each has manifested itself in the form of an upwards shock to large parts of the CPI basket: VAT-able elements, energy, and imported goods and services respectively. 4 But these do not reflect increased pressure of demand. In fact, because each of them will push prices up relative to wages, it is likely that each shock will have pushed down on domestic demand, at least temporarily 5 . In addition, one can argue that the financial crisis also had elements of a negative supply-side shock: forcing up the cost of credit and depressing activity levels.
How should monetary policy have responded? In principle, the best policy path is probably to allow the one-off price level effects of such shocks to flow through to final prices but to make sure that there are no second-round effects - such as compensating wage increases - that would leave inflation above the target in the medium term. That is essentially what the MPC has been trying to do. Yes, we could have tightened policy to keep inflation at target when the shocks first hit - but it would have needed to be a very material tightening. I believe that would have engendered a worse outcome on all counts.
First, shocks such as the VAT rate increase and the step up in the level of commodity prices were not anticipated, and their impact on inflation is likely to be temporary in nature. By the time Bank Rate had exerted its full influence on the price level, the effect of the shocks was likely to have already dissipated, if not reversed. Second, Bank Rate would have dampened inflation by adding even more downwards pressure on demand on top of the existing recessionary forces. The consequent fall in output and rise in unemployment would have been greater. These two considerations mean that a monetary policy response would most likely have only injected extra volatility to inflation: by the time the price shocks wore off, we would have been facing a severe undershoot of the inflation target and deflation would have been a distinct possibility. The path of output would have been unnecessarily volatile and the deviation of inflation from target might have been different but no more acceptable.
The possibility of such shocks, and indeed the ability to respond as we have done, was foreseen when the inflation target regime was conceived. Indeed our Remit from the Government is clear on the consequences. The Remit says that (with my emphasis):
'The framework takes into account that any economy at some point can suffer from external events or temporary difficulties, often beyond its control . The framework is based on the recognition that the actual inflation rate will on occasions depart from its target as a result of shocks and
4 The latter would affect CPI directly though its impact on imported consumer goods and services, and indirectly by affecting the cost of domestic production reliant on imported intermediate inputs.
5 The real exchange rate shock is likely to push down on domestic demand but should increase export supply.
disturbances. Attempts to keep inflation at the inflation target in these circumstances may cause undesirable volatility in output'. 6
In other words, if shocks drive inflation sufficiently away from the target, the MPC are allowed to choose how quickly to return inflation to 2% and should explain such policy decisions via the open letter system.
I have often been asked whether or not changes to the target, or the remit, or to CPI itself would help us to set policy. The answer is a clear no. Of course, the Government can decide to change these things if it wants to, but that wouldn't magic away the problem. The unfortunate and difficult truth is that the shocks that have hit the economy recently have not been caused by monetary conditions. They are real economy shocks which make us individually and collectively worse off - we have to pay more tax, we have to pay more, in real terms, for petrol, gas, electricity and imported goods or services. The choice that monetary policymakers face is how much of these relative price shocks should be accommodated by a higher level of prices, and how much should be accommodated by a squeeze on nominal wages (and higher unemployment). Even if we had had perfect foresight at the time, the MPC would have still been faced with the same choice - the adjustment in real wages was unavoidable. Of course if nominal wages start to rise in an attempt to offset that inevitable fall in standards of living, risking a wage-price spiral as the UK had in the seventies, then the MPC would be duty-bound to raise Bank Rate sooner to bring inflation back to target, regardless of any short-run pressures on output.
Given the nature and timing of the shocks, I would argue that the best we can do with monetary policy is accept the initial impact and then to gently steer inflation back to target in the medium term. In the remainder of this speech, I want to consider how we come to make our judgments about that, month-by-month, and in particular, discuss the relative role of economic forecasts and human judgment.
## Monetary policy in practice
The Bank has come in for a lot of criticism recently about its forecasting record. For a period now, CPI inflation has been significantly higher than the Bank had been previously expecting. And in 2010 Q4 and 2011 Q1, output growth disappointed relative to our central expectations. In this section, I want to dig a little deeper into the issues around models and forecasts.
Step back for a second and consider what an economic model actually is and why it might be useful. At its heart, a model - any model - is just a framework for thinking about an issue, be it how the economy reacts to different shocks, how the setup of a car might affect the way it handles a corner, or how a disease might spread through a population. As all modellers would concede, they can only provide an approximation of reality, which is always far more complex.
That isn't to say that models aren't useful - far from it. Economic models, like any other, are essential. But they are not a panacea. For example, it would be a mistake to think one can just keep learning and adding to a single economic model until it gets ever more complete. What tends to happen is that one just ends up with a very large model which nobody can understand or use for any practical purpose. Instead, the best way forward in understanding the economy seems to be to use a variety of different models, remembering the relative strengths and weaknesses of each, and applying judgment accordingly. That is exactly what the MPC does when it considers the appropriate stance of monetary policy. The Bank's approach to modelling strives to be eclectic.
6 The remit can be found at http://www.bankofengland.co.uk/monetarypolicy/pdf/chancellorletter110323.pdf
What about forecasts? They too are essential, given that the lags between setting policy and its impact on inflation require us to be forward-looking. To come up with its forecast, the MPC has to reach a view based on all the information available to us: data, theories, a variety of models, business and market intelligence, our own experiences and judgement and so on. But there are huge uncertainties within the information set and so there must be equally great uncertainties about the resulting judgements. That is why the one, indeed the only, thing we know for certain about any forecast of the economy is that it will be precisely wrong. That is also why the MPC puts little weight on its central projection, and chooses to communicate its judgments to the public by focusing on the range of possible outcomes and the balance of risks, as summarised in the fan chart. 7
What this means is that setting monetary policy is about decision making under uncertainty. It isn't clean. In fact it's very messy. But nobody has a working crystal ball. The important question for us is not whether we got the central projection right but whether we set the right policy, given what we knew at the time. Crucially, we are accountable to the public, including via the Treasury Select Committee, for our policy decision. I do not think I have heard many argue that we should have done something materially different with policy during the depths of the recession in 2009 (i.e. to the extent necessary) in order to generate a different outcome for inflation now.
## Conclusion
What I hope this discussion highlights is that, in practice, the job of monetary policy makers comes down to making difficult and complex decisions in the face of uncertainty. Over the past couple of years the challenge has been dealing with a succession of real changes in relative prices (via negative supply side shocks) which have pushed up on prices whilst depressing demand and output. That is extremely uncomfortable for everybody. But there was, and is, no easy way for monetary policy to deal with the impact of such shocks. In our current projections there are very major risks to either side of the central case. On one side, higher inflation expectations could become entrenched making it very costly for the MPC to subsequently bring inflation back to target. On the other side, the economy could be much weaker than we expect pushing down on inflation and risking deflation. Recovering to the target from that could be even harder (at least in my personal view). MPC members place different weight on these possibilities and reach different judgements accordingly. But it is clear to all of us that both risks exist.
I believe the MPC is charting the right course through these difficult times. Despite temporarily higher inflation for a period, and notwithstanding all the uncertainty, I want to assure you that we remain determined to bring inflation back to target in the medium term, consistent with our remit.
7 For more information, see Britton, E, Fisher P, and Whitley J, 'The Inflation Report projections: understanding the fan chart', Bank of England Quarterly Bulletin, February 1998, pages 30-37.