In the fall of 2026, the Bank of Canada and Government of Canada will complete their regular review of Canada’s Monetary Policy Framework. It will include an assessment of how well the inflation targeting framework, which aims to keep inflation close to two per cent, has served the Canadian people.
While the Bank of Canada has made welcome adaptations to its existing inflation targeting framework in recent years, my view is that these adaptations fall far short of what is needed. From my perspective, Canada — and other countries — need a fundamental review of the inflation targeting framework itself.
Inflation in Canada and almost all advanced countries rose well above the two per cent target during the pandemic. While it then fell sharply, it has remained above target levels in most countries. In February 2024, the Centre for Economic Policy Research (CEPR) published an ebook in which many prominent central bankers, including the Governor of the Bank of Canada, gave their frank assessment of what central banks did wrong during the pandemic, and the lessons to be learned.
The lessons included:
To the credit of the Bank of Canada, a careful reading of the post-pandemic speeches of the Governor, Deputy Governors and other staff indicates that these lessons have generally been well recognized. While the Bank has summoned little public attention to problems posed by fiscal policy, the Bank of Canada Act ensures that the Governor can impart his concerns privately to the Minister of Finance.
The Bank has also gone well beyond words of recognition to actual improvements to policy formulation since the pandemic. Recognizing the dangers associated with its swollen balance sheet, the Bank acted to reduce its size and has also resisted calls to return to pre-pandemic interest rate levels. New forecasting models have been introduced that put more emphasis on cascading supply shocks among diverse sectors and the possibility of non-linear outcomes.
Finally, recognizing the challenges posed by fundamental uncertainty about the future, the Bank is making greater use of scenarios and is incorporating geopolitical, climate and other shocks into its thinking. All these incremental improvements to the inflation targeting model are welcome.
But in my view, they do not go far enough.
The inflation targeting framework, used by the Bank of Canada and many other central banks, makes price stability (two per cent inflation in Canada) over a relatively short horizon (around two years) the primary objective of monetary policy. That objective is to be achieved by raising policy interest rates whenever inflation is forecast to overshoot the inflation target, and by lowering them when the opposite is the case.
This policy framework has been supported by a highly simplified model that assumes the economy is pushed away from equilibrium (full employment and low inflation) primarily by external (exogenous) demand shocks, such as an AI-related investment boom. While the possibility of negative supply side shocks, such as a rise in global oil prices, has been increasingly recognized since the pandemic, there has still been no explicit recognition that central banks failed in the past to identify positive supply shocks, like the return of China to the global trading system under Deng Xiaoping. This error led to excessively easy monetary policy over many decades, with policy interest rates staying too low for too long.
When it comes to the financial sector, the possibility of positive feedback processes leading to intermittent crises over time has been almost totally ignored. Positive shocks to profits almost always generate what Alan Greenspan, former chairman of the U.S. Federal Reserve, called “irrational exuberance” and prices that are bound to collapse. Given that the last three major economic downturns, in both Canada and other major developed countries, had their origins in financial markets, this omission seems increasingly hard to justify.
I have been a dissenter from orthodox monetary policy beliefs for some decades, preferring empirical reality to theoretical elegance. In reality, the economy is a complex adaptive system subject to tipping points (crises) where positive feedback dynamics, particularly within the financial sector, are the principal threat to macrofinancial stability.
The financial sector in advanced economies is by nature highly procyclical, magnifying booms and exacerbating busts. Inflation targeting regimes reinforce these procyclical tendencies. Indeed, I would contend that such regimes create a profound intertemporal inconsistency. Stimulating the economy through lower rates when inflation is deemed too low leads to more credit expansion and higher debt levels in the near term that significantly constrain demand over the medium to long term. This buildup of debt could eventually culminate in a profound economic crisis. I believe this threat still confronts us. In his autobiography, Paul Volcker seemed to agree when he said, “Ironically, the ‘easy money,’ striving for a ‘little inflation’ as a means of forestalling deflation, could, in the end, be what brings it about.”
While it is sometimes hard to visualize alternative policy frameworks, it is a simple fact that the Bank of Canada and other central banks have already rotated through a number of distinctive policy regimes since the end of the Second World War. In each case, a failure in the previous policy regime demanded a transition. In my view, that time has come again.
Since the end of the Second World War, macroeconomic policy has been mainly directed toward smoothing out the business cycle, which involves recurring cycles in production, employment and inflationary pressure. Inflation-targeting regimes are essentially of this nature. However, as post-1945 financial regulations dropped away, we witnessed the return of a pre-war phenomenon — the financial cycle. A financial cycle can be substantially longer than a business cycle and involves the gradual buildup of debt and asset prices to extreme levels, even though Consumer Price Index (CPI) inflation might remain quite stable.
The engine of the financial cycle is procyclicality, a self-reinforcing set of feedback loops involving funding conditions, the appetite for risk and asset prices. Some piece of good news (a reduction in policy rates, an increase in productivity, a new discovery, etc.) increases the demand for credit, which a deregulated financial system is increasingly willing to supply. The growth in credit then leads to asset price appreciation which, in turn, provides more collateral for further loans and a further increase in optimism. At some point, however, the optimism is seen as unwarranted by emerging facts and the whole process goes into reverse. The boom is replaced by bust.
Since roughly the late 1980s, we have had four financial cycles. Three of these ended in a crisis (1990, 2001 and 2008) while a fourth upturn was cut short by the COVID-19 pandemic that began in 2020. However, the optimism in financial markets then returned, buoyed by massive fiscal and monetary stimulus during the pandemic. This optimism led in turn to further increases in debt levels (particularly for governments) and stock prices, a massive extension of credit by non-regulated sources and a further easing of credit conditions. It seems increasingly likely that this process will end in crisis, as in all the previous cases.
Inflation targeting can alter people’s behaviour and support the procyclical process.
First, putting all the focus of monetary policy on near-term inflation invites policy-makers and others to ignore the grave possible harm arising from the financial cycle.
Second, inflation targeting can strengthen these processes, creating moral hazard. Financial speculators are encouraged to behave even more imprudently if they believe they will be bailed out by monetary easing, should disinflation result from a financial bust. Similarly, governments can become inured to the risks of rising debt levels after long periods of low interest rates and falling debt service requirements.
Third, and paradoxically, the more credibility a central bank has, and the more that inflation expectations are anchored around the target rate, the longer the financial cycle can expand without meeting any monetary resistance.
These concerns are not just theoretical but are supported by historical facts. Global monetary policy was eased significantly through most of the 1980s. This was first in response to the sharp global disinflation of the early 1980s, but later in response to developments in financial markets. The Louvre Accord of 1987 required global easing to support the U.S. dollar, and then the stock market crash later in that year elicited the so called ‘Greenspan put,’ whereby investors came to expect the U.S. Federal Reserve (Fed) to reduce policy rates any time material weakness in equity markets occurred. Debt levels rose sharply through the 1980s, as did speculative investments, culminating in a global financial crisis and a shallow recession in 1991.
Through the 1990s, positive supply shocks (faster productivity growth, favourable demographics, globalization, etc.) were increasing the potential growth rate and the neutral real interest rate. However, at the same time, they were also reducing inflationary pressures. Wrongly treating the latter as a symptom of inadequate demand, central banks failed to raise policy interest rates adequately and a debt-fuelled investment and stock market boom then followed.
When this boom finally culminated in higher inflation, monetary resistance through higher interest rates, global financial disruptions and the contraction of 2001, monetary policy was again eased aggressively to foster recovery. This policy worked, indeed supporting ‘The Great Moderation,’ but only at the expense of still more debt accumulation and rising financial imbalances. These culminated in the Great Financial Recession of 2008. After that, until inflation eventually rose during the pandemic, central banks were encouraged to keep rates “low for long” with a combination of the continuation of disinflationary supply side shocks, and the growing headwinds of rising debt that their own policies had encouraged.
Sadly, the undesirable effects of sustained periods of low real interest rates go beyond the reduced effectiveness of monetary stimulus and potential financial instability.
One effect is reduced potential growth as low rates prioritize financial speculation and the financialization of the economy over investment in real capital; massive increases in share buybacks in recent decades are a case in point. Low rates also encourage misallocations of capital and the survival of ‘zombie’ companies — heavily indebted companies kept alive by cheap finance — both of which reduce productivity growth.
A second effect is to increase wealth inequality as low rates increase asset prices, most of which are held by the already wealthy. Worse, slower growth and stagnant real wages, together with a perception of growing inequality, have been a recipe for political discontent. It is not surprising that we have seen the rise of populist movements in so many advanced economies.
Finally, when the policy rate controlled by the central bank reached zero, concerns arose that monetary policy had lost its effectiveness in combatting deflation. This is because increases in real rates could no longer be offset by reductions in nominal rates. Thus constrained, many central banks experimented with quantitative easing and forward guidance. These policies have had numerous undesirable consequences, including massive capital losses by central banks, which have done a great deal of reputational damage.
The pursuit of price stability over four decades has also left many economies struggling with record high debt levels, both private and public. This debt overhang now imposes a severe constraint on the behaviour of even ‘independent’ central banks. Raising real rates to fight inflation could threaten the debt service capacity and even solvency of private sector companies and heavily mortgaged households. This threatens recession and perhaps deflation. Moreover, easing policy in such a crisis could reignite the debt accumulation process making future crises even worse. This private sector problem is often referred to as the problem of financial dominance.
In contrast, the problem of fiscal dominance occurs when it is governments that cannot bear the burden of higher debt service. In this case, the solution is likely to be much higher inflation as governments put pressure on central banks to buy newly issued debt at acceptable rates. To forestall private holders of old government debt from selling it to the central bank, a process likely to end in hyperinflation, administrative and regulatory means will be used to increase the private demand for such debt. This combination of higher inflation and forced holdings is called financial repression. It successfully reduced the government debt overhanging several countries after 1945, at the expense of bond-holders who received negative real returns.
These problems of financial and fiscal dominance will be exacerbated in the future by structural shifts that reduce aggregate supply (fewer workers in an aging population, adaptation to the effects of a changing climate, deglobalization, etc.) and increase aggregate demand (defence spending, increased health-care expenditure, climate change initiatives, AI investments, etc.).
This combination of factors would seem inherently inflationary, implying that real interest rates should rise and nominal rates should rise even more. If this cannot be allowed to happen, then the likelihood of higher inflation and eventual financial repression would be significantly increased. Arguably, markets are becoming more aware of these dangers since, very unusually, recent reductions in short-term interest rates have been met by higher rates for bonds issued for longer maturities. This could imply that we are closer to a global bond market crisis than we currently think.
Some hold out hope that an artificial intelligence revolution will raise productivity levels and GDP enough to avoid such problems. However, for a crisis to be avoided, we also need an appropriate response from monetary policy. Somewhat counterintuitively, this means real interest rates rising to prevent current spending from increasing excessively in response to anticipated income gains. If one accepts this logic, then it was an error for Greenspan to lower rates in the late 1990s as productivity surged.
Moreover, lowering rates today in the face of a similar shock might be even more dangerous. First, the productivity gains today are presumed, rather than actual as they were in the 1990s. What if they do not materialize? Second, the debt trap must still be reckoned with. With debt stocks having risen so much further in this century, the encouragement of still more debt accumulation would hardly seem prudent.
As a guiding philosophical principle, central banks should recognize that the economy is not a linear and deterministic system, but rather complex and adaptive. Such systems require a totally different governance framework.
Consistent with this belief, central banks in association with other government institutions should adopt a macrofinancial stability framework.
Such a framework would still pursue price stability but over a significantly longer horizon than the two-year horizon used currently. This would allow monetary tightening to mitigate excessively rapid credit and debt creation, even if inflation were to be temporarily pushed below short-term targets.
This framework would also be more tolerant of economic downturns, recognizing that they serve a useful purpose (creative destruction) in capitalist societies. Moreover, successive small recessions have fewer ramifications than less frequent, but potentially much larger, economic downturns. In the former case, households and businesses can adapt. In the latter case, problems seem insurmountable and political unrest often follows.
A macrofinancial stability framework would require central banks to “lean” against credit excesses (booms) rather than attempt to “clean” up after the boom collapsed. The current strategy, focused on minimizing downturns, builds moral hazard into the system. It encourages excessive risk-taking and larger bubbles, which then lead to larger downturns. It also results in cumulative debt increases, both private and public, which can create sudden crises.
A strategy based on leaning against the upturn would reduce the amplitude of both the upturn and the associated downturn. The indicators for tightening would include a broad range of variables, financial as well as real, rather than being restricted to variables linked to forecasts of the CPI. This could include money supply growth, credit growth, credit spreads and indicators that financial assets are overpriced. If such tightening implied that inflation would fall below putative short-term targets, this shortfall would be expected to reverse over time.
A macrofinancial stability framework would pay more attention to the symmetry of policy over the cycle. Under current policies, interest rates have been reduced more in downturns than they have been raised in upturns. This has caused policy rates to ratchet downward over time, eventually to the effective rate of zero. Similarly, debt levels and debt ratios (e.g., to GNP) have ratcheted upward. A macrofinancial monetary framework would be usefully complemented by a government pledge to hit a target for the ratio of government debt to GDP over time. This would recognize that fiscal dominance poses a real threat to inflation.
Finally, a macrofinancial stability framework would ensure that monetary and regulatory tightening were used as complements rather than substitutes. Under the current system, macroprudential regulatory measures are used to prevent the materialization of the worst implications of excessive credit creation. Thus, they support lower-for-longer interest rate policy, which allows the bubble to build for a longer time than would otherwise be the case.
In Canada, that could mean greater collaboration between the Office of the Superintendent of Financial Institutions (OSFI) and the Bank of Canada. The Bank of England, for example, has a Monetary Policy Committee and a Financial Policy Committee with joint members.
Moving to a macrofinancial stability framework would certainly help in reducing the procyclicality problem and could be implemented without major difficulty. Changes to financial regulation and other structural domestic reforms (to be discussed in a subsequent commentary) might also be useful. However, if none of these changes were deemed adequate to reduce the incidence of financial crises, more radical monetary solutions might eventually have to be considered.
Advocates of ‘free banking’ focus on the moral hazard created by continuously expanding government safety nets. If these were swept away, the stability of the system would be ensured by enhanced self-discipline as well as market-discipline.
Another suggestion, first made by the Chicago School in the 1930s, would be to introduce a system of ‘narrow banking,’ wherein money creation would be solely in the hands of the central bank. The key idea is that the capacity of commercial banks to create money out of nothing is at the heart of procyclicality.
Deposits with narrow banks would have to be backed up 100 per cent with central bank reserves or short-term government securities. In this way, governments would take away the capacity of banks to make loans by simply writing up both sides of their balance sheet.
The review should begin by recognizing that debt accumulation, whether private or public, can lead to serious economic crises. Such dangers, which can appear suddenly after long periods of tranquility, are totally ignored by the current inflation targeting framework. Because of the problem of prospective fiscal dominance, monetary policy cannot continue to be thought of as independent of fiscal policy.
The next step would be to assess whether, given current debt levels in Canada, introduction of a new macrofinancial monetary framework would be helpful or harmful. Would it reduce the probability of future crises to acceptable levels without unacceptable implications for economic growth?
A further complication might be the implications for the Canadian dollar from the unilateral introduction of a new monetary framework. Arguably, the failure of central banks in smaller countries (like Canada, Sweden and New Zealand) to lean against worrisome increases in house prices and household debt was due to concerns that the exchange rate might appreciate and reduce competitiveness. The defeat in 2018 of a Swiss referendum on narrow banking might also have reflected fears of going it alone.
If the introduction of a new macrofinancial framework is not helpful, perhaps because of initially high levels of existing debt, the review should then investigate alternative measures to reduce debt ratios to less dangerous levels. It is likely that most of these suggestions will meet serious political opposition, because any reduction in one party’s debt is the reduction of someone else’s asset. The Bank of Canada must then strongly make the case that unpalatable outcomes are better than disastrous ones, and must champion the least unpalatable of the alternatives.
The summaries and links below trace the development of my concerns about monetary policy risks while working at the Bank for International Settlements (BIS), then the Organisation for Economic Co-operation and Development (OECD), and in subsequent presentations, speeches and writings. The intent is to set out the evolution of my critique and proposed better paradigm for central bank policy frameworks, including demonstrating both the consistency in my central themes and empirical evidence in support of my arguments over the past two decades.
The consensus among central banks, financial markets and governments during the Great Moderation of the mid 1980s through the mid 2000s was that price stability, measured over a relatively short horizon (typically two years), is sufficient to ensure economic and financial stability. In contrast, my paper suggested that the attainment of sustained low inflation over a short horizon might be inadequate to avoid serious macroeconomic disturbances over the long term.
History is replete with examples of major economic and financial crises that were not preceded by inflationary pressures. Conversely, history shows that many periods of deflation, based on rising productivity, were simultaneously characterised by rapid growth. Structural changes in the global economy implied that this past experience could have more relevance than commonly thought. If so, the implication in the 2000s and going forward was that policies focused on achieving price stability require more flexibility and a longer-run focus to avoid major macroeconomic disruptions. The subsequent occurrence of the Great Financial Crisis (GFC) underscored this need to look beyond short-term price stability.
My paper assessed the desirability of unprecedented and sustained monetary policy easing by all the major central banks in the years following the GFC. It weighed up the balance of the desirable short-run effects and the undesirable longer-run effects — the unintended consequences. I concluded that there are limits to what central banks can do. One reason is that monetary stimulus, which encourages borrowing (debt accumulation) to stimulate spending, might now be less effective than previously because of higher debt levels. It is also the case that ultra easy monetary policies can eventually risk the health of financial institutions and the functioning of financial markets, threaten the “independence” of central banks, and encourage imprudent behaviour, especially rising debt, on the part of governments. None of these unintended consequences is desirable.
This paper documented that great fluctuations have repeatedly taken place in the goals of monetary policy, the optimal exchange rate framework, beliefs about the monetary transmission mechanism, the system of political oversight, and many other aspects of domestic monetary frameworks over the five decades through the early 2010s. In each case, a failure in the previous policy regime demanded a transition. I suggested that time might have come again. I recommended that the role played by money and credit, the interactions between price stability and financial stability, the possible medium-term risks generated by “ultra easy” monetary policies, and the facilitating role played by the international monetary (non) system all needed urgent attention. My conclusion was that monetary policy after the GFC was relied on too heavily in the pursuit of “strong, balanced and sustainable growth.”
This paper averred that the way monetary policy is conducted needs to change fundamentally. Past practice, based on the assumption that the structure of the economy is both knowable and controllable, is simply wrong. In reality, the economy is a complex and adaptive system, like many others in nature and society, and cannot be well understood or closely controlled.
Highly expansionary monetary policy provided strong growth for a while, but it was neither sustainable nor adequately inclusive. This paper contended that unduly easy monetary policy actually contributed materially to both economic and financial instability (unsustainability) and to the fact that the gains from stronger growth and international trade have been unfairly shared among the social classes (non-inclusiveness). This perception of unfairness has generated anger and contributed to political divisions.
A particular shortcoming of prevailing analytical models was how little emphasis they put on supply side developments. In pursuing price stability prior to the GFC, central banks failed to recognize that low inflation in the in the 1990s, and subsequently, was due to positive supply side shocks (e.g., technological advances and the re-entry of China into the world trading system).
On the demand side of the economy, central banks failed to see the dangerous implications of the cumulative increase in credit and debt also associated with easy monetary policies. Over time, these headwinds have threatened less demand, not more, constituting another vicious and downward spiralling circle.
I argue in this paper that monetary policy should be guided much more by financial sector developments (credit and debt) and much less by near-term targets for inflation.
This argument is first supported by an empirical review of the negative outcomes produced by the current policy framework’s objective of a low level of inflation with its primary focus on labour markets and output gaps in the “real” economy when setting monetary policy. The resulting financial bubbles from too little focus on credit and debt have created ever larger bubbles which continue to threaten future growth prospects.
A second level of support is provided through identifying a number of false beliefs that underlie the current approach to policy.
First, the need to use monetary policy to avoid disinflation was much exaggerated. Historically, most disinflations have been associated with positive growth in productivity and did not cause broader problems.
Second, the effectiveness of easy money in stimulating demand was overestimated. Easy monetary policies encourage debt accumulation which acts as a headwind to spending over time.
Third, it was a false belief to assume that the many unintended and dangerous consequences of easy money could simply be ignored, given the boosts to inequality, financial instability and fiscal unsustainability. In addition, it threatens central bank independence and reduces the potential growth of the economy over time.
An alternative monetary policy framework would recognize an economy as a complex, adaptive system like many others in nature and society. Embracing this view has important implications for the objectives of monetary policy, assessing and reacting to deviations from objectives, and preventing and managing crises. From this different perspective, arguments for introducing a “narrow money” regime also need significantly more attention.
