Don't give up on inflation targeting just yet

The Money, Macro and Finance Research Group (MMF) annual policy conference is one of the best places to gain an overview of current issues in monetary policy and I was fortunate to attend this year’s event to hear presentations from the likes of St Louis Fed President Jim Bullard, former Fed Board member Frederic Mishkin and Riksbank governor Stefan Ingves (and not forgetting a useful contribution from the MPC’s ever-thought provoking Gertjan Vlieghe). Mishkin’s presentation won the prize for the most entertaining. In addition to being a very accomplished speaker, his was the first presentation I have heard which incorporated a reggae track, produced by the Bank of Jamaica to highlight its new inflation targeting regime (here – and it’s well worth a look).

There were a significant number of issues raised and I will undoubtedly come back to many of them in the course of future posts. If there was an overarching theme from the whole event, it was the general agreement that monetary policy has reached the limits of what it can do to support the economy without some additional fiscal support. Not that central bankers would ever admit they are out of bullets, but it seems obvious they cannot go on doing what they are doing in the expectation that things are going to change. There was also a lot of head-scratching as to why inflation continues to undershoot central banks’ 2% target and Mishkin’s presentation on inflation targeting is a good starting point to think about the inflation framework adopted by most central banks.

In my view, Mishkin started from the wrong place. He started by pointing out that inflation in the Anglo Saxon world started falling in the early-1990s at the same time as central banks began to directly target inflation. Although he did not say so in as many words (although it was explicitly stated by some in the audience), the underlying message was that central banks’ focus on reducing inflation was the key factor in quelling the rampant inflation of the 1970s and 1980s. This idea has been bandied around by numerous central bankers over the years, although it is heard less frequently today. And with good reason: it is far from the whole truth.

Inflation expectations took a battering following the early-1990s recession, the second in a decade, whilst intensified competition – itself the product of the 1980s deregulation regime – also acted to depress expectations. I heard nothing about the impact of the end of the Cold War or the rise of China, both of which increased the global economy’s productive capacity with consequent dampening effects on inflation. The reason why such denial may be a problem is that if central bankers really believe they have curbed inflation, when in fact it is largely the product of exogenous forces, they may not be best equipped to force it back towards target and may understand the inflation process less well than they think.

This raises another question. If inflation is persistently below target, what is the point of maintaining the target? In the case of the ECB, for example, CPI inflation has averaged 1.0% since 2013 which is well below the target rate of below, but close to, 2%. In order that the ECB’s inflation rate averages 1.9% over the ten-year period 2013 to 2022, it would have to average 3.9% between now and December 2022. Such arithmetic has prompted luminaries such as Olivier Blanchard, the IMF’s former chief economist, to suggest raising central bank inflation targets to 4%. But there are good arguments against such a strategy.

In the first place, it is more difficult to stabilise inflation at 4% than 2% because the former figure is not consistent with the definition of price stability in which inflation is not a big factor in economic decision-making. An inflation rate of 4% implies that the price level doubles every 17.5 years versus 35 years in the case of 2% inflation. If you are making long-term calculations, such as investment or pension planning, there is a big difference between these two figures. Moreover, once we start to deviate from a 2% inflation target, why stop at 4%? Why not raise the target to 6% or 8%? Before too long, we could very easily get back towards 1970s territory.

But there is an argument suggesting that central banks could allow temporary deviations from the 2% target. Rather than target the inflation rate, an alternative would be to target the price level on (say) a two-year horizon. One of the disadvantages of targeting inflation (i.e. the rate of change of prices) is that shocks which impact on prices – and therefore temporarily raise the inflation rate – become fully embedded in the price level. But in a price level targeting regime, they are not.

We can illustrate this graphically (below). Consider the case where there is a shock to prices which causes them to fall by 1.5%. If the central bank does not attempt to compensate for this, in the knowledge that this is a one off impact, the price level is permanently lower and the rate of change eventually rises back to the 2% target (grey line). But if the central bank attempts to restore the price level to its steady state path on a 2-year horizon, it can tolerate much higher inflation and it is only three years after the initial shock that the inflation rate returns to the steady state rate of 2% (black line). One clear advantage of this regime is that it acts as an anchor for long-term price levels, and thus provides more certainty for inflation rates used as a basis for long-term wage and pricing contracts.

This sounds good in theory but does it work in practice? Much of the literature is based on the assumption that inflation expectations adjust relatively quickly and in a pre-defined manner. However, the experience of the last decade suggests that expectations adjust much more slowly than is often assumed. Nor do we know the functional form of the expectations formation process. This means that central banks may have to allow inflation to run further ahead of target for longer than they would like in order to give time for expectations to normalise. This in turn runs the risk that central banks may be perceived as having made a permanent adjustment to their inflation target.

Much as we may be critical of central banks which bang on about their target even though inflation continues to deviate from it, the constant repetition serves a purpose of reminding us that this target exists. If they stop talking about it, we might start to pay less attention to it and eventually forget about it altogether. And if we get to this point, there is a real fear that expectations could become unanchored and move around wildly as economic conditions change.

You might argue that in a world of exceptionally low inflation, there is no need to worry about any pick up in the pace of price growth. But central bankers can counter, with some justification, that just as the exceptionally high inflation of the 1980s quickly gave way to lower inflation within the space of five years, so the process could just as quickly run the other way. It is thus important to try and ensure that expectations remain anchored and it may be too soon to call for radical changes to the inflation framework.

The underperformance gathers pace

One of the exercises I have been conducting over the past three years is to assess how the UK economy has performed relative to pre-Brexit referendum expectations. At the end of 2017 I looked at the economy’s performance and motivated by this set of tweets by FT journalist Chris Giles, I thought it worthwhile providing an update. The analysis is complicated by historical revisions to data which make it extremely difficult to compare GDP levels. But based on data for growth rates we can make some observations.

Rather than use one single forecaster as a reference, I have compared outturns against a panel of long-term GDP growth forecasts made in May 2016  which are reported in the Treasury’s monthly publication ‘Forecasts for the UK economy’. No less than 14 institutions made projections covering the period 2016 to 2020 (with a higher number providing near-term projections) providing a decent sample size. Using the current 2019 consensus forecast as a basis for this year’s outturn (we already have half of this year’s GDP figures so it is a reasonable assumption), real GDP in 2019 is likely to end up around 2 percentage points lower than predicted prior to the referendum (chart 1). That is equivalent to a year’s worth of pre-referendum expected GDP growth.

Measured in terms of constant 2016 prices this amounts to £41.5 bn. With the GDP deflator estimated to have increased by a cumulated 5.7% since 2016, this amounts to a loss of GDP in nominal terms of around £43.9 bn. For the record, that averages out at around £281 million per week. Recall Boris Johnson claimed that the UK would be better off by £350 million per week. Even on the basis of his vastly inflated figures, 80% of the gains have been wiped out by weaker growth. But since the UK’s net contribution to the EU is actually between £150 million and £212 million per week, depending on how we measure it, (here for a useful ONS calculator) the hit to growth now far exceeds any savings from ending contributions to the EU budget. The further ahead we roll the figures, the worse it looks. Comparing the current 2020 consensus forecasts with the May 2016 projections, the loss of output translates into a figure of £328 million per week.

To understand why GDP has underperformed so dramatically it is worthwhile taking a dig into some of the components. We do not have long-term consensus projections for all the components so I have resorted to using the OBR’s pre-referendum forecasts to give a feel for the underlying detail (which can be found here). This dataset does, however, allow us to look at the data on a quarterly basis so we can compare the figures from the aftermath of the referendum right up until the latest data available (Q2 2019). 

After adjusting for the change in the national accounts base year, we discover that by mid-2019 household consumption was around 1% below the OBR’s 2016 projection. Government consumption is around 3% above the OBR’s estimate, largely thanks to a huge increase in outlays over the last nine months as the government has opened the taps. But the biggest single shortfall comes from fixed investment which is 10% below the 2016 projection, with business investment falling 15% below the forecast (chart 2). Indeed, all the non-investment components broadly net out against one another leaving the collapse in fixed investment to account for pretty much all the slowdown in GDP. And why has investment proved so weak? Because companies have lined up to tell us that Brexit-related uncertainty has forced them to put their capital spending plans on hold.

We cannot say with any certainty that the outturn in 2019 would have come anywhere close to the predictions made in 2016. But the comparison does show that there was a structural break in the way the economy was expected to behave and it is too much of a coincidence to say that it was not associated with the Brexit referendum. Brexit supporters will point to the continued strong performance of the job market and argue that the UK has continued to create jobs. It is certainly true that the unemployment rate will turn out around 1 percentage point lower than the 4.9% rate predicted by the consensus in May 2016.

But in 2016, the OBR expected a cumulative increase in productivity of 5.2% between 2016 and 2019. Data released this morning suggest that output per hour worked has risen by a measly 0.9% in the last three years. Productivity is the key driving force of our living standards, and it is simply not rising fast enough. This may not all be due to Brexit – after all, productivity growth has underperformed for the last decade. However, the increase in output continues to be driven by labour input in a way which is reminiscent of less well developed economies and without any capital input to supplement it (i.e. investment), it is likely that GDP per head will continue to stagnate.

Finally, just to illustrate the UK’s economic underperformance in an international context, the ONS has compiled quarterly data on GDP across a range of countries. Using a base period of Q4 2016, the UK has underperformed against all major industrialised economies bar Italy and Japan, up to and including Q2 2019 (chart 3). You can argue as much as you like about the poor performance of Germany in recent months, but the data still show that it has outperformed the UK in growth terms since the start of 2017.

Politicians (and their advisers) focused on “getting Brexit done” are oblivious to the risks which Brexit poses to the economy. To be frank, most of them are spectacularly ignorant of even basic economic concepts so those ordinary voters who still support Brexit can be excused for their lack of understanding. But what the overall picture shows is that there was a structural break in the performance of the UK economy in the second half of 2016. It cannot be denied; it cannot be dismissed as Project Fear. It is there in the economic numbers in black and white. And the growth shortfall is broadly in line with estimates made prior to the referendum. Moreover, the UK has not yet left the EU. Things could get even worse, depending on the nature of the departure.

As I noted in a post last week, economics will ultimately determine whether Brexit is a success. The evidence so far suggests it is not going particularly well and those who continue to push this ruinously stupid economic policy will ultimately be forced by the evidence to account for their actions. Nobody has yet come up with any good economic arguments in favour of Brexit, and there is a good reason for that – there aren’t any.

Beware the ultra-long trap

The bond market world has moved into strange territory of late. Around one quarter of the debt issued by governments and companies is currently trading at negative yields and in Germany government bonds are yielding negative rates of interest out to a maturity of 30 years. In effect, investors are paying the government for the privilege of owning their debt, quite a long way out across the maturity spectrum. This does not mean that investors periodically hand over a sum of money to the government, but it does mean that the stream of income that the bond pays is less than the amount the investor paid for the bond.

Why would anyone be prepared to do that? The simple answer is that investors need the reliability and liquidity that high quality bonds can provide when they have to allocate their portfolio over a wide range of assets. Think of it this way: Even though investors would have maximised their returns over the past decade if they had been fully invested in equities, at no point could they ever be sure that the bandwagon would keep on rolling. Theory suggests that long-term returns are maximised if the portfolio is spread across a range of assets, and the likes of pension funds are required to allocate a minimum portion of their portfolio to bonds in order to meet their payout obligations. Even now pension fund providers will not hold less than 40% of their assets in bonds, and although the returns may be miserable the buyers of high quality government (and corporate) debt are highly likely to get their money back. The bond market is thus a safe place to store wealth – a hedge in an uncertain world.

On the basis that demand for bonds is unlikely to dry up, despite low interest rates, governments arguably have a strong incentive to issue debt at current low rates, either to finance additional spending or refinance existing debt. Moreover, they have an incentive to issue much longer maturity debt than previously in order to lock in these rates for as long as possible. One would think that no rational investor worth their salt would buy bonds issued with a negative coupon rate. Think again. In August, the German Federal Finance Agency sold a 30-year zero coupon bond (i.e. it pays no interest) but because the bond was sold at 3.61% above par (i.e. investors pay 103.61 but receive only 100 at maturity) this amounts to a negative interest rate of 0.11% per annum if the bond is held to maturity. Demand fell short of target, with sales of €824 million versus a target of €2 billion, but it was generally perceived as a useful trial of the extent to which the market was prepared to accept low rates.

In 2017, before yields fell to current levels, two countries – Austria and Argentina – decided to test the demand for ultra-long issuance by selling 100 year bonds. Their experience has been rather different, with the price of the long-term Argentinean bond since halving in value whereas the Austrian bond has doubled. It is understandable that Argentina tried to lengthen the maturity of its debt profile – it has been a serial debt defaulter over the last 70 years with five episodes of default or rescheduling since 1950, so it made sense to reduce disruptions caused by debt rollovers. But this history also weighs on investors. Can Argentina really be trusted not to default on its debt in the next 100 years? In order to get investors onboard, Argentina had to issue at a coupon rate of 7.9%. That might seem high in the context of the US or Europe, but with the central bank benchmark rate at 74.98% and 10-year yields currently trading at almost 28%, it does not sound quite so bad.

One country which could conceivably get away with issuing longer dated bonds at low coupons is the US where the Treasury, which only issues as far ahead as 30 years, is mulling the possibility of going even further out along the curve to maturities of 50 and even 100 years. Treasury Secretary Steven Mnuchin said last month that “we are looking at potentially extending the portfolio. If there is proper demand, we will issue 50-year bonds.” He went on to suggest that if these bonds prove to be a success, the US would consider the possibility of 100-year bonds. Historically, the US has issued debt at longer maturities and between 1955 and 1963, it sold bonds at maturities up to 40 years (here). In 1911, it even issued a 50 year note to fund the construction of the Panama Canal.

But as attractive as long-term issuance sounds, there are a number of factors to consider. History cautions that issuers have to strike a balance between offering yields which are sufficiently attractive to investors but which minimise the costs to the issuer, and as the US found in the 1950s and 1960s that is a difficult balance to get right. One of the reasons for sticking to the current maturity schedule is that the US Treasury has tried to avoid tactical or opportunistic offerings of debt, and has focused instead on maintaining a regular and predictable schedule. This has helped the US Treasury market to become amongst the most liquid financial markets in the world. This in turn allows the government to offer relatively low coupons in return for the privilege of liquidity and helps to keep down Federal debt servicing costs. Issuing ultra-long debt threatens to reduce market liquidity which might in the long-run push up US rates.

Odd as it may sound, almost 20 years ago there were fears that rapid US growth and declining budget deficits would lead to a shortage of Treasury securities. Obviously that never happened, but in a world where there is a move to increasing the duration of debt we could get to a situation where there are temporary issuance droughts which could distort the shape of the yield curve. If there is an increase in the proportion of debt issued at longer maturities, a prolonged period of reduced issuance – perhaps because of rapid growth and smaller fiscal deficits – could mean a shortage of supply at the short end of the curve which would push down yields (i.e. raise prices) and result in a significant steepening of the yield curve. The point may be hypothetical but it demonstrates that changing the duration of debt issuance could have significant market consequences.

It is a mark of desperation that investors would even consider buying such long-dated bonds, particularly at a time when global uncertainty appears so high. Although the US continues to issue the world’s reserve currency, in which case it makes sense to buy dollar bonds, we cannot say that will be the case in 100 years’ time. Nor can we be sure that even the US will be able to pay its debt. After all, those British and French creditors hoping that the Russians would finally pay up on their pre-1918 debt are still waiting. And if we cannot be sure that even the US will pay up, who can we trust?

Border disorder

It has been a tumultuous week in the world of British politics as the political establishment tears itself apart over the issue of Brexit. With just 29 days to go, Boris Johnson reiterated at today’s keynote speech to the Conservative Party conference that the UK is "coming out of the EU on October 31, come what may." There is only one way that this can be achieved and it requires the UK and EU to agree a deal that will enable the transitional arrangement to come into force which will prevent a no-deal Brexit. For all the ongoing excitable talk in the press, no doubt whipped up by Conservative sources, that the government is prepared to leave without a deal, do not forget that to do so would be a breach of the law. Parliament has already ruled out this option and having lost one legal battle, Johnson would not relish the prospect of MPs once again resorting to the courts to force him to comply.

If ever there was a time for ignoring press chatter and speculation, now is that time. Ahead of Johnson’s speech, there were stories suggesting that the government would unveil its strategy to deal with the Irish border problem, and present it to the EU as a “take it or leave it” option. That was not what we got. The rhetoric was far more conciliatory as the government presented a plan which allows for regulatory alignment between Northern Ireland and the Republic for a limited period of time. But it is what happens once the transition period ends that is far more problematic.

At that point, Northern Ireland will leave the EU customs union and revert to being part of the UK customs union. However, the North will be fully aligned with the EU for the purposes of the agri-food business and “in addition, Northern Ireland would also align with all relevant EU rules relating to the placing on the market of manufactured goods.” This effectively means the creation of two borders: A regulatory border between Britain and Northern Ireland for manufactured goods and agri-food products but a customs border between the North and the Republic.

Moreover, the British government has reiterated its position that “no customs controls necessary to ensure compliance with the UK and EU customs regimes will take place at or near the border.” Instead, they will rely on “continuing close cooperation between UK and Irish authorities” with movements across borders notified using a declaration process, utilising processes such as the trusted trader scheme which the EU has already rejected as unworkable. None of this sounds to me like a solution that the EU can buy into – it is all a bit vague. And from the EU’s perspective it gets worse. The UK has proposed that every four years, the Northern Irish electorate will get a chance to have its say on the regime to ensure that it is still satisfied with the arrangements. But “if consent is withheld, the arrangements will not enter into force or will lapse … after one year, and arrangements will default to existing rules.” This is clearly not the permanent solution to the border problem that the EU is seeking. In its current form, I would be highly surprised if the EU accepts it. Indeed, chief EU negotiator Michel Barnier was reportedly scathing of Johnson's Irish border plan, describing it as "a trap".

The best way to view the proposal is as a starting point for discussion. But it all seems a bit late although that seems to be the only way that anything gets done in these negotiations. Clearly the EU does not want to reject it outright since this would effectively scupper any chance of a deal at the summit in two weeks’ time. The question is then how much are the British prepared to bend to accommodate the EU’s requirements (and vice versa)? And where does it leave the Democratic Unionist Party? After all, they were the ones who objected to a customs border between Northern Ireland and the mainland yet that is exactly what they have now signed up to.

But Johnson is now entering the most critical phrase. His mantra “let’s get Brexit done”, for which read "leave the EU on 31 October", rings increasingly hollow. It is 100% certain that Brexit will not be “done” by the end of this month even if the UK leaves the EU. The October deadline represents a staging post which would at best mark the start of negotiations regarding the nature of the UK’s long-term relationship with the EU. As a third country with much less bargaining power than the EU, the UK will be in a much weaker position than it is now and negotiations will be a lot tougher than anything seen so far. As a consequence, the economic uncertainty that has held back business investment over the past three years is likely to persist – a point made by MPC member Michael Saunders in a speech last week.

Although the Conservative government believes that it will be possible to reunite the country once it has dragged Brexit over the line, I continue to have my doubts. A no-deal Brexit will exacerbate, rather than heal, divisions and the damage to the UK as a business location may already have been done – a view I heard expressed at a business forum I attended yesterday. Brexit long ago stopped being about the economics and simply became the toughest battlefront in what now has to be seen as a culture war. But the economics matters – ultimately it will determine whether Brexit is deemed a success. I am not holding my breath.