Brave new world (of monetary policy) –
Is the era of inflation-rate targets coming to an end?
How the Fed intends to stimulate inflation (expectations)
In the current environment, policy makers at major central banks, who manage the ups and downs in their country’s – and, in some cases, the global – economy from a monetary-policy perspective, have to solve two key conundrums. Firstly: Why is inflation still so low, even after many years of ultra-expansionary monetary policy? And secondly: What tools can be used to respond to the next economic downturn, given that interest rates are already very low?
In the US in particular, these questions have sparked a fundamental debate about the general focus of the Fed’s monetary policy. The bank wants to achieve three objectives as part of its brief: a high level of employment, moderate interest rates in the long term, and stability of prices. Its primary instrument in this endeavour is its ability to adjust the key interest rate. In the past, the Fed typically responded to economic downturns by lowering interest rates – on average by 3–6 percentage points per recession, as can be seen in the chart above. At present, the key interest rate stands at 2.5 per cent. This means that if the economy slipped into a recession now, a comparable reduction would take the Fed’s key interest rate well into negative territory. What to do?
Many central bank experts now believe that interest rates bumping along at around zero will not be an exception in the monetary policy landscape of the future but rather a relatively regular occurrence. Against this backdrop, discussions have long been ongoing about what the most effective monetary policy manoeuvres would be in such an environment. In short, there is no clear and simple answer.
In October 2017, this question – already a long-standing matter of debate among experts – came to the attention of a wider audience when Ben Bernanke, then chair of the Federal Reserve, publicly challenged the concept of inflation targeting. Since then, economists have been busy calculating and analysing whether it would be better to use a target price level instead of an inflation target. What might look to outsiders like two ways of looking at broadly the same thing would constitute nothing short of a paradigm shift in monetary policy.
The Fed conference in Chicago on 4–5 June 2019 is going to revisit this topic. Many market participants expect that a decision about abandoning inflation targeting will be made soon – maybe already at this conference – and that the implementation could follow in the first half of 2020.
In order to achieve the greatest possible degree of macroeconomic stability, monetary policy makers have been putting price stability at the heart of their work. As early as 1990, the central banks of New Zealand and Sweden began to define price stability as a quantitative target for the medium-term rate of inflation – i.e. a percentage rate of increase in price levels in the economy during a specified period. All major central banks around the world subsequently adopted this approach.
Every year in January, the Fed publishes a ‘statement on longer-run goals and monetary policy strategy’. In particular, this statement includes the definition of the explicit inflation target of 2 per cent (as measured by the annual change in the price index for personal consumption expenditures), which was adopted by the bank in 2012.
A key objective for the Fed is to stabilise inflation expectations, because in standard economic models, this forecast of the future rate of price increases is the main driver for inflation. If inflation expectations fall below the central bank’s target rate, actual inflation is likely to fall as well in the medium term. In such a scenario, the Fed faces two major challenges:
One is that lowering the nominal interest rate will no longer be effective enough as a means of mitigating a potential economic downturn. For example: If the average level of inflation is 2 per cent and the equilibrium real interest rate is 1 per cent, a central bank (in this case the Fed) can lower interest rates by 3 per cent before they – or, more specifically, money market interest rates – drop into negative territory. But if average inflation stands at just 1 per cent, the scope for adjustments before hitting zero is only 2 per cent.
The other is that inflation also serves as a means of lowering otherwise downwardly rigid wages in real terms in response to a sudden fall in demand in order to stabilise employment. And the lower the rate of inflation, the less effective this mechanism becomes.
Inflation targets have no memory…
When central bankers attempt to control inflation, they compare the actual level of inflation with their target rate and seek to correct any deviations from their (self-defined) benchmark. In very simple terms, this means that they mainly focus on a single data point – the rate of inflation reflects whether they have achieved their objective or not. It is important to note that previous inflation levels are – at least superficially – irrelevant in this context.
… but target price levels do
If a central bank works with a target price level rather than with a numerical target data point such as the rate of inflation, it is managing a target path. This means that it has to mitigate deviations from this target path within a specified time period.
The following stylised model analysis demonstrates the difference between the two approaches. For the purposes of this example it is assumed that a random central bank is aiming for a 2 per cent rate of increase in price levels as a target. During periods nil and one, there are no unusual occurrences. In period two, the system is hit by a price shock and inflation rises from the target level of 2 per cent to an elevated level of 3 per cent. The economy’s aggregate price index rises from 102 to 105.
Inflation-rate targeting (left): In the model scenario, the central bank brings the inflation rate back down to the desired target rate of 2 per cent in the following period. Problem solved. But while the price shock has only had a temporary impact on the rate of inflation, it has had a lasting effect on the path of price levels. Over time, the shifts in price levels resulting from successive price shocks will accumulate. The longer the period reviewed in an economic analysis, the harder it becomes to predict the path of price levels.
Price-level targeting (right): The central bank subsequently corrects the effects of the price shock. In the model scenario, the price level realigns with the target path in period three. How can this be achieved? For a limited period, the central bank has to lower the rate of inflation below the target rate of 2 per cent. In period three, inflation drops to just 1 per cent.
If a central bank uses price-level targeting, it cannot ‘forget’ about periods in the past where inflation was higher or lower than the economically optimum rate of inflation. Instead, the bank will try to offset periods of lower-than-desired inflation by subsequently allowing inflation to rise above the target rate of 2 per cent. A statistical expert would say that the inflation rate and the price level are both ‘stationary’ under price-level targeting, whereas only the inflation rate is stationary in an inflation-rate targeting regime.
What has really happening so far:
The Fed has been working with a numerical data point as a target since 2012: It aims to maintain the private consumption deflator at a level of 2 per cent. Has it been successful in meeting this objective? The answer is ‘yes and no’: During two months of the overall period – namely in January and February 2017, the rate of inflation was exactly 2 per cent. But if we define a slightly wider band around this specific target data point, we can see that inflation always moved within a range of one standard deviation from the target rate, except in the period between December 2014 and December 2015. If the Fed was pursuing a target band rather than a specific target rate, it would be well on track.
A hypothetical view: If the US central bank had already been using a price-level target in the past, actual inflation would now have taken the price level quite far off this target. Depending on the selected start time (either July or December 2008 in the illustration above on the right), the price-level path would be either 6 per cent or 4 per cent above the actual rate of inflation. If the Fed wanted to offset this difference now, it would have to let medium-term inflation rise to 6 per cent or 4 per cent respectively.
Critics of price-level targeting argue that in order to let the rate of inflation ‘catch up’, central banks would have to allow the economy to overheat. If the economy was hit by a negative supply-side shock such as a drastic rise in oil prices during this period, inflation could easily rise to unacceptable levels. But if the central bank was pursuing a price-level targeting strategy, it would then not be able to ignore such a ‘spike’ in inflation. In theory, it would have to respond by raising the key interest rate substantially in order to offset the high actual inflation level by forcing inflation down below the target rate. This would be far from ideal.
Advocates of price-level targeting dismiss these arguments as mainly theoretical and argue that no central bank would actually raise interest rates while the economy was struggling to cope with the shock of a spike in oil prices.
The best of both worlds?
In reality, the Fed is obviously free to adopt a monetary policy strategy that combines elements of both approaches. Inflation averaging is considered to be the middle ground between inflation targeting and price-level targeting. Under this approach, the central bank pursues an inflation target which is not based on one point in time but rather on an average across several years.
Experts are also discussing temporary price-level targeting as an alternative to pure and permanent price-level targeting, meaning that the approach would only come into effect if and when interest rates drop to zero. If inflation is below the target rate at this point, the central bank should keep the key interest rate at a low level until the rate of inflation has risen and remained above the target rate for a prolonged period of time (to be defined) and no longer. This would serve to offset the previous phase of below-target inflation. Interest rates would remain at zero for much longer in this case than under an inflation-targeting regime.
Those involved in the public debate about a paradigm shift in monetary policy and its implementation – in whatever shape or form – hope that it will give those sections of the economy that closely follow the actions of central banks confidence that the Fed will keep interest rates low for a prolonged period so that its monetary policy will have a (more) expansionary impact – even though it will not, in contrast with previous downturns, be able to implement sweeping interest-rate cuts of 6 per cent or so in the event of a recession.
This is precisely the measure that the parties to the debate regard as the solution to the monetary policy dilemma: Central banks should commit to keeping interest rates very low for a very long period of time. In monetary-policy circles this is referred to as ‘lower for longer’ or L4L, for short. Many economic models do, in fact, demonstrate that a convincing L4L strategy generates prosperity. The promise of keeping interest rates low for a long time lowers returns and drives up inflation expectations. Both of these effects create downward pressure on real interest rates, incentivise market participants to spend their (borrowed) money and thus boost the economy. What a brave new world!
A glance at the eurozone
The European Central Bank (ECB) is facing the same problem as the Fed, but its situation is even more precarious, because its key interest rate has been at zero since March 2016. In previous recessions, the ECB lowered its base rate by 3 percentage points and 1 percentage point respectively. A key distinction from the Fed is that price stability is the sole primary objective pursued by the ECB. In order to achieve this objective, the bank aims for a target rate of inflation of just under 2 per cent. Has it been able to meet this target so far? Not really.
The chart below on the right shows that the ECB would also have to let inflation rise well above its target rate if it were to adopt price-level targeting with retroactive effect instead of its inflation target. Irrespective of whether or not a price-level target would be set with retroactive effect or not, Mario Draghi indicated in April 2019 that he would be willing to engage in a debate about strategic approaches to monetary policy in Europe. Challenging and exciting tasks will await his eventual successor in November 2019.
Impact on the capital markets
The Fed adopting such a paradigm shift in its monetary policy in the near term, e.g. next year, would be seen as a clearly dovish signal with far-reaching ramifications for the capital markets. A (more) expansionary approach to monetary policy typically leads to falling yields, which translates into persistently low, or even negative, interest rates and – most importantly for central banks – ideally also rising inflation expectations.
Under an inflation-targeting regime, medium-term forecasting is a rather uncertain exercise for all market participants, because central banks do not correct deviations from the inflation target in individual years. On the contrary: The game of hitting 2 per cent inflation starts anew every year. This results in a growing gap between the (economically desirable) path of the price level and the actual rate of inflation.
A central bank that pursues a price-level target, on the other hand, is more likely to allow inflation to rise above the target rate on occasion. Based on current discussions by the Fed, this could happen even in an environment of relatively low interest rates. Households and companies would not need to change their consumption and investment habits, because – unlike at present – they would not need to worry about potential interest-rate hikes. This, in turn, would typically (further) drive up prices.
Recent statements by the Fed suggest that it is likely to adopt some form of inflation averaging as a moderated version of price-level targeting in the wake of its conference in Chicago. This would create an asymmetric outlook on the Fed’s future monetary policy. If the economy shows signs of slowing and/or inflation starts to dip, the central bank would lower interest rates. But in the event of rising inflation, the Fed would allow the rate to rise above the current target level of 2 per cent under the new monetary policy paradigm in order to let the price level catch up.
And in the long term? The case of Japan poignantly reminds us to tread carefully. Its economic trajectory shows that the mere willingness of a central bank to tolerate – or even actively promote – elevated levels of inflation is no guarantee of an actual rise in inflation. Which neatly takes us back to our first conundrum: Why is inflation still so low, even after many years of ultra-expansionary monetary policy? Insider rumours suggest that even the central banks don’t have an answer to this one.
-  Ben Bernanke (2017): Monetary Policy in a New Era. Paper presented at the conference of Rethinking Macroeconomic Policy Peterson Institute, Washington, DC, 2 October 2017.
 Fed (2019): Conference on Monetary Policy Strategy, Tools, and Communication Practices. 4–5 June 2019, Federal Reserve Bank of Chicago.
 Deutsche Bundesbank (2010): Monthly Report – January 2010.
 Many theories exist about the etymological origin of this expression. However, nobody seems to know exactly where it came from. Some say that doves are simply peaceful animals, and that they fly lower than hawks. The German Bundesbank issued the following statement: “It is likely that these terms were originally coined in the context of security and defence policy and were later adopted into monetary policy jargon.”
 Thomas Mertens and John Williams (2019): Monetary Policy Frameworks and the Effective Lower Bound on Interest Rates. AEA Papers and Proceedings 2019, 109: 427-432.
Unless otherwise noted, all Information and illustrations are as at 03 June 2019