Claudio Borio, former Head of Monetary and Economic Department, Bank for International Settlements
Czech National Bank Workshop on Monetary Policy: Inflation Targeting Frameworks Under Review
Prague, 1 April 2025
Since it inception some 35 years ago, inflation targeting (IT) has spread across the world and become the standard monetary policy framework. IT has largely achieved what it was designed to achieve – help deliver an era of low inflation. The framework is also well on its way to overcome the challenge posed by the surprising post-Covid inflation surge. But is it fully fit for purpose?
What follows argues that in order to lengthen its lifespan, the framework needs to adjust in ways that do not figure prominently in the current debate. One is more systematic consideration of the longer-term damage that financial factors can cause to the economy. Another is more systematic consideration of the importance of safety margins in the conduct of policy. And all this grounded on a clear recognition of what monetary policy can and cannot deliver.
The core of the argument is that the framework inadvertently contributed to two challenges that are often taken as independent of it: the Great Financial Crisis (GFC) of 2007-09 and the long-term loss in the room for policy manouvre that intensified in its aftermath. Consider each in turn, setting the basis for the proposed adjustmets.
Challenge 1: The pre-GFC phase
At the root of the pre-GFC challenge was the failure to understand how the interaction between the monetary policy framework and tectonic shifts in the economic environment were springing chellenges from unexpected quarters by fundamentally changing the nature of economic fluctuations.
The two tectonic shifts are financial liberalisation and the globaisation of the real side of the world economy. Financial liberalisation, which was largely complete by the early 1990s, provided bigger scope for disruptive financial expansions and contractions – or financial cycles – in the form of self-reinforcing shifts in credit and asset prices, notably property prices. The globalisation of the real side of the world economy, which gathered pace in the 1990s, blowed powerful disinflationary tailwinds and provided fertile ground for financial booms – in Kindleberger’s terminology, acted as a “displacement”. Think, in particular, of the integration of former communist countries and China into world trade and of the opening up of emerging market economies.
In this context, the new IT regime could inadvertently accommodate the build-up of financial imbalances. There was no incentive to tighten monetary policy as long as inflation remained under control, especially since the new regime dispensed with credit and monetary aggregates.
Indeed, there is clear evidence that the business cycle changed from around the mid-1980s – from inflation-induced to financial-cycle-induced recessions (Graph 3). The main cause of downturns was no longer a monetary tightening to quell inflation but the spontaneous unwinding of financial imbalances. The GFC is just the most spectacular example. The downturns of the early 1990s had already exhibited similar patterns, not just in Japan and the Nordic countries, but also in the United States themselves.
Put differently, inflation became an unreliable signal of the unsustainability of economic expansions. That role was largely taken over by the build-up of financial imbalances.
Importantly, the unreliability of inflation as a signal was transmitted to the policy anchor that was taking over the role of monetary and credit aggregates – the natural rate of interest or r-star. R-star is typically defined as the short-term real interest rate consistent with output at potential and hence stable inflation – ideally at target – absent business cycle disturbances, ie in a steady state. Thus, it was common practice when inflation would turn out lower than expected to adjust estimates of potential output upwards and hence of r-star downwards. In turn, this would call for lower policy rates to keep the stance unchanged.
This pattern can promote a loss in the room for policy manoeuvre over successive business and financial cycles. Lower rates accommodate the build-up of financial imbalances. And when the boom turns to bust, even lower rates are needed to nurse the economy back to health. Moreover, they may be needed for longer, since financial-cycle-induced downturns tend to be deeper and recoveries from them lengthier. The economy struggles to absorb the debt overhang and the misallocation of capital. Low rates beget lower rates.
Challenge 2: The post-GFC phase
While it is necessary to ease monetary policy for long following financial recessions, arguably this response went too far. Even when economies had recovered, some key central banks strived hard to offset even moderate shortfalls of inflation from target. In nominal terms, interest rates reached historical troughs – drifting even into negative territory; in real terms, they remained negative for even longer than in the Great Inflation era; and central bank balance sheets reached historical peaks, hitherto approximated only during wars.
Here, the notion that r-star is exogenous to monetary policy played a key role. If the “equilibrium” real interest rate is exogenous, the only way to regain room for policy manoeuvre on a durable basis is to boost inflation. This means reducing interest rates so as to be able to raise them as higher inflation eventually takes hold. In other words, it means losing room for policy manoeuvre for sure today in the expectation of regaining it tomorrow. Thus, if inflation does not respond as envisaged, central banks end up losing, not regaining, policy headroom.
As it turned out, central banks did struggle to push inflation back to point targets even as economies recovered. For one, the lingering globalisation forces that had helped central banks keep inflation under control pre-GFC were now working against them; in addition, when inflation settles at a low level, it tends to be less responsive to policy. In a low-inflation regime, inflation exhibits significant self-stabilising properties, as its evolution tends to reflect largely sector-specific (idiosyncratic) price changes which, by their nature, leave little lasting imprint on inflation. And policy adjustments exhibit dinishing returns – they have less impact on aggregate demand when interest rates are already low or balance sheets are large and expanded further in normal market conditions. Empirical evidence supports these various propositions.[1]
The journey ahead: possible adjustments
The previous analysis suggests a number of possible adjustments to current IT frameworks. The proposed adjustments take into account at least four considerations.
First, the need for frameworks to be robust to alternative possible economic environments. The environment ahead is likely to be more inflation-prone, but one cannot rule out a return to the more disinflationary one of the past.
Second, the critical importance of long policy horizons, not least given that the financial cycle is much longer than the business cycle, as traditionally measured.
Third, the need to systematically incorporate the importance of safety margins (or buffers) in the conduct of policy. Economies that operate without safety margins are vulnerable to both the inevitable next recession and unexpected exogenous shocks, such as Covid.
Finally, and above all, the need for a realistic view of what monetary policy can and cannot deliver. As long as it is prepared to pay the price, monetary policy can prevent inflation from becoming entrenched. As long as it has fiscal backing, monetary policy can stabilise the financial system at times of stress even when solvency concerns mount. The central bank plays its traditional role of lender or, increasingly, market maker of last resort. And as long as prudential policy is supportive, monetary policy can help restrain financial imbalances and contain their damage. But monetary policy cannot fine-tune inflation within basis points, let alone economic activity within narrow ranges. Nor can it be relied on as a de facto engine of growth. The razzmatazz surrounding policy decisions suggests that we have long lost that sense of realism.
These broad considerations have implications for objectives and the strategies to pursue them.
First, hardwiring a low-inflation regime should remain a priority. The behavioural definition of price stability suggested by Volcker and Greenspan is especially apt – a condition in which inflation does not materially influence people’s behaviour.
This also means that raising current targets, as several observers have proposed, would be a bad idea. Quite apart from undermining the central bank’s credibility, this would endanger that condition and hence also the self-stabilising properties that inflation has in a low-inflation regime. Inflation would risk drifting out of the region of rational inattention into that of sharp focus. And it could then act as a more effective coordinating device for people’s behaviour. As the common factor behind price changes increases and these become more similar, the inflation rate becomes more representative of the price changes individual agents care about. Critically, this is one reason why transitions from low- to high-inflation regimes tend to be self-reinforcing.
Second, relative to how targets have been generally interpreted in the past, the key adjustment would be greater tolerance for moderate, even if persistent, shortfalls of inflation from narrowly defined targets. This would help retain precious room for manoeuvre, exploiting the self-stabilising properties of inflation in low-inflation regimes. And it would allow the central bank to take into account more systematically the longer-term damage that arises from the consequences of low interest rates, which weaken the financial side of the economy.
Implementing such a strategy would also require addressing what Felstein and Rajan have aptly termed the “deflation bogey man”. This is the view that deflation is a kind of red line that, once crossed, gives rise to a self-reinforcing downward spiral in economic activity or to a depression trap. Empirically, the evidence is not consistent with this view. There is no systematic link between falling price levels and weak economic activity. The Great Depression is very much an exception and reflected a broader set of forces. Conceptually, it is not clear why the link should exist in the first place if gradually falling prices reflect favourable supply factors and if productivity growth avoids the need for reductions in possibly rigid nominal wages.
Third, further thought should be given to how best to use the additional room for manoeuvre to tackle the financial cycle. This would call for lengthening horizons as far as possible. It would require upgrading the role of financial conditions, credit aggregates and property prices among the set of indicators that central banks follow. And, analytically, it would call for less reliance on macroeconomic models based on the standard “shock-propagation-return-to-steady-state” paradigm and more on approaches that allow for endogenous and possibly unstable fluctuations.
Fourth, it would be appropriate to rely less on forms of forward guidance that provide specific information about the future path of interest rates beyond the central bank’s reaction function. These can unduly constrain the central bank’s ability to respond to rapidly changing conditions. And they can also unduly compress risk premia and encourage risk-taking. Think, for instance, of the “measured pace of rate increases” ahead of the GFC. In the extreme, rather than guiding markets, the central bank may end up being taken down the wrong path, as the emergence of fragilities in the financial system can force its hand.
Finally,it would be important to put a premium on exit strategies whenever the central bank is called upon to take exceptional measures to stabilise the system and the economy. The difficulties in reducing the historically large and risky central bank balance sheets are testimony to the challenges involved. A reasonable principle is that central bank balance sheets should be as small and riskless as possible, subject to fulfilling mandates effectively. As argued elsewhere recently, except possibly for the need to hold foreign exchange reserves for precautionary purposes, balance sheets can be quite small. Central bank balance sheets should be elastic – ready to increase when circumstances require it – not large. Given the economic and political economy costs of larger balance sheets, the initial size is a hindrance, not a plus.
What would the resulting regime look like?
It would be a regime in which the central bank keeps a sharp focus on the medium term. The central bank seeks to ensure that the financial side of the economy, which it influences and through which it operates, does not end up derailing the economy, whether through inflation or financial instability, broadly defined to include the sizeable financial amplification of business flcutuations. That is, the central bank sets the monetary preconditions for sustainable growth but does not end up being relied on as the engine of growth.
It would be a regime in which the operational definition of the inflation target is consistent with that overarching objective. This means that the target is low enough so that inflation does not materially influence agents’ behaviour, but flexible enough to allow the central bank to take into account the financial forces that can generate damage down the road.
It is a regime in which the central bank’s reaction function calls for forceful responses when inflation threatens to get out of control but allows for greater tolerance for moderate, even if persistent, shortfalls from target.
Does this have implications for mandates? Not so much if by mandate we mean the general goals that may be set out in the central bank’s law or in agreements with the government. IT regimes have often operated in broadly similar ways despite different mandates. But it does have implications for the way in which mandates are interpreted and, above all, communicated. The word often missing here is “sustainable”. Once sustainability is added as an explicit consideration, whether in terms of inflation, output, employment or financial stability, all the pieces of the jigsaw puzzle fall into place. The lens though which we see the economy is more important than the ultimate goal, on which there is greater consensus.
Conclusion
Looking ahead, the toughest challenges for monetary policy regimes may well be still to come. For one, the political environment is becoming less conducive to a stability-oriented monetary policy. Over time, a dangerous expectations gap has been developing between what monetary policy can deliver and what it is expected to deliver. The pressure on central banks to take a short-term view is growing alongside the demands placed on them. In such an environment, institutional safeguards such as central bank independence become all the more precious – the option is “in the money”. But they can help only up to a point.
In this context, we should not take too much comfort from the recent rather painless reduction in inflation. Circumstances have been extraordinary. The typical future fight against inflation is likely to follow more familiar patterns and cause bigger costs. This would be especially the case if the fight took place in a less globalised world: firms and workers would regain pricing power and would offer greater resistance to compressions in real purchasing power and real profit margins.
In addition, and closely related, the unsustainability of fiscal trajectories represents another major threat. As discussed in detail in the BIS Annual Economic Report in 2023, it is probably the biggest longer-term threat to macroeconomic and financial stability in the years ahead – and hence, ultimately, a threat to the monetary policy regime itself. Monetary and fiscal policy are joined at the hip. And there are clear limits to what monetary policy can ultimately do if fiscal policy is out of kilter. Both policies need to operate firmly with a “region of stability”, consistent with sustainable growth.
These threats are also likely to make the adjustments proposed here even…less likely. But IT regimes cannot afford to stay still. IT does not represent the end of monetary policy history, as it were. It is just one, if exceptionally long, chapter in that history.
[1] What about the common view that the lng-term decline in real interest rates reflects factors that are exogenous to monetary policy. Three observations qualify this view. First, to th extent that monetary policy contributes to financial crises, it also contributes to changes in real interest rates that are for all intents and purposes, this is indistinguishable from an impact on r-star. Second, if the data are allowed more room to speak, the empirical evidence supporting a link between the usual real factors (eg, demographics, growth etc) and real interest rates does not survive if the data us extended back in time before the mid-1980s. Thrid, by contrast, a robust link emerges with monetary policy regimes.
This contribution summarises the main argument in C Borio “Whither inflation targeting as a monetary standard?”, BIS Working Papers no 1230, December 2024. The reader is referred to that piece for an extensive set of references and documentation of the various points. For a detailed documentation of the evolution of inflation targeting frameworks across countries, with special attention to the measurement of the degree of flexibility, see C Borio and M Chavaz, “Moving targets: Inflation targeting frameworks, 1990-2025”, BIS Quarterly Review, March 2025.
