I am pleased to welcome you once again to the Annual Research Conference of the National Bank of Ukraine and Narodowy Bank Polski. Let me begin by thanking our Polish partners and co-organisers. This conference has long been a shared project of our two institutions.
The circumstances in which we meet offer a useful starting point for a broader question about central banking.
Just two weeks ago, we expected many of you to be here with us in Kyiv. Many were ready to come, and I am sincerely grateful for that. In the end, however, we moved the conference fully online as Russia intensified its attacks on Ukrainian cities, people, businesses, and both critical and civilian infrastructure, including the energy system, railways, transport hubs and border crossings.
We could have ignored the risk, cancelled the conference, or postponed it until better times. Instead, we chose to adapt. This choice captures something fundamental about resilience: the ability to adjust to changing circumstances while preserving the function that matters.
Ukraine knows a great deal about this kind of resilience.
At the time of this conference, Russia’s full-scale war against Ukraine had entered its 1,672nd day — already 104 days longer than the First World War lasted. By the fourth anniversary of the full-scale invasion, Ukraine’s air defence had destroyed more than 140,000 aerial targets launched by Russia.
Behind these numbers are not abstract risks but human lives: destroyed homes and businesses, damaged power plants, nights spent in shelters, interrupted school lessons, and working days after only a few hours of sleep.
And yet the system continues to function. Payments go through, banks open, businesses produce, children learn, and institutions make decisions. The longest period of rapid credit growth in Ukraine’s history is taking place during the full-scale war.
For Ukraine, resilience is therefore not a concept from academic literature. It is daily practice.
The pandemic taught us how to work at a distance. War is teaching us something different: how to remain connected, continue performing our functions and operate under extreme conditions, even when the environment changes faster than institutions would normally be expected to adapt.
The format of this conference changed. Its purpose did not.
Beyond familiar shocks
This brings me to a broader challenge for central banks.
We increasingly operate in the face of shocks whose sources lie far beyond the boundaries of our mandates. We are used to speaking the language of demand shocks, supply shocks and financial shocks. That vocabulary remains valid, but it may no longer be sufficient.
I would suggest adding another category to our working vocabulary: existential, or system-altering, shocks — shocks capable of changing not merely the parameters of a system, but the system itself.
Central banks are accustomed to thinking in terms of marginal changes: a few tenths of a percentage point of inflation, a change in growth, a movement in an exchange rate, or a deviation from a baseline scenario. Some shocks, however, operate differently. They work through thresholds and discontinuities.
A system may absorb increasing pressure up to a point. Once a critical threshold is crossed, however, even a relatively small additional shock may push it into a different state. The relationships on which our models and expectations are built may cease to hold. Such a shock can do more than shift a forecast; it can change the coordinate system on which that forecast was based.
And this is where professional language meets human reality.
At the extreme, an existential shock does not take away a few tenths of a percentage point. It can take away everything: productive capacity, infrastructure, savings, institutional capacity and, most terribly, human lives.
War is precisely that kind of shock. Large-scale cyberattacks, the destruction of key infrastructure, geopolitical ruptures or technological disruptions may acquire similar characteristics when their effects cross critical thresholds.
Not every large risk is existential. What distinguishes an existential risk is that its consequences may cease to be marginal. This distinction matters for preparedness. Assessing the probability of an event is not enough. We also need to understand the potential scale of its consequences, the thresholds beyond which normal relationships may break down, and the functions that must remain operational even in extreme scenarios.
Ukraine encountered this reality earlier than many others and in a far more extreme form. We therefore have lessons to share.
After the start of the full-scale invasion, we ensured the uninterrupted operation of the banking system and payments, preserved financial stability, prevented a foreign exchange crisis and maintained an adequate level of international reserves. We continued our path towards European integration and restored the effectiveness of monetary policy, gradually moving from emergency measures back towards more market-based mechanisms.
We did all of this during the largest war in Europe since the Second World War.
But greater awareness of systemic risks must not become an argument for an unlimited central bank mandate.
The source of a shock may lie far beyond our direct remit while its consequences affect prices, financial stability, payments and confidence in money. The challenge is therefore not to respond to everything, but to understand how an evolving environment affects the central bank’s ability to fulfil its mandate.
Our tools may change. Our procedures may change. Even the boundaries of what society expects from a central bank may evolve. But our objectives must remain clear.
As Fabio Panetta argued at this conference, credibility sometimes requires institutions to stand firm, safeguard their independence and defend their policy framework. At other times, it requires them to adapt, rethink their tools and acquire new capabilities.
There is no contradiction between the two. Flexibility works only when it has an anchor. For a central bank, that anchor is clarity of objectives, institutional independence and accountability to society.
We can change a tool, a procedure or the way in which we achieve an objective. Even the role of a central bank can evolve. But we must not lose sight of the goal.
This idea also has a historical dimension for Ukraine.
This September marks thirty years since Ukraine’s monetary reform and the introduction of the modern hryvnia. Yet the first hryvnia banknotes had appeared several years earlier.
In the early 1990s, Ukraine had only just restored its independence. It had neither its own printing capacity nor sufficient resources to produce the new banknotes. The first hryvnias were printed abroad, including in Canada, effectively on credit.
In 1992, as part of the highly classified operation “Shield of Ukraine”, 105 containers of banknotes travelled by civilian cargo ship from Canada to Ukraine under the protection of a Ukrainian special unit. Inside those containers was part of the material foundation of our future monetary sovereignty.
For me, this story carries an important lesson. Financial sovereignty does not mean isolation. It means having the capacity to make your own decisions and to take responsibility for them. Reliable partnerships expand the space in which such decisions can be made.
Our decisions during the full-scale war have been made in Ukraine, and responsibility for them is ours. But international support has expanded the space in which those decisions could be taken. Financial assistance has made it possible to finance critical public expenditure without undermining macrofinancial stability, while cooperation with the IMF, European institutions, international financial institutions and fellow central banks has helped preserve reserves, predictability and trust.
Our Polish colleagues deserve particular recognition. Narodowy Bank Polski was among the first institutions to stand with us. Its foreign exchange swap decision was taken on 24 February 2022, the first day of the full-scale invasion, and already in March Ukrainians who had found refuge in Poland were able to exchange cash hryvnias for zloty.
But the first support came from people. Polish families opened their homes to millions of Ukrainians. We remember that.
These years have taught us how to operate in a world where uncertainty is persistent and shocks can no longer be treated as rare exceptions. They have also taught us how much resilience depends on the ability to act together.
This brings me to the three themes of this conference: resilience, credibility and innovation. I would like to frame them as three prompts for discussion.
I use the word deliberately. New technologies have reminded us of a simple truth: the quality of an answer often begins with the quality of the question.
Resilience: flexibility needs an anchor
The first prompt is this: resilience is impossible without flexibility, but flexibility needs an anchor.
A resilient institution must be capable of changing its instruments, procedures and operational frameworks quickly. A crisis will not wait for us to complete every discussion or construct a perfect model.
But flexibility does not mean the absence of rules, nor does it mean accepting chaos. The faster the environment changes, the more important it becomes to know what must remain stable.
For a central bank, this means above all clarity of objectives, independence and accountability. Institutional resilience depends on the ability to adapt without weakening those anchors.
The question, therefore, is not whether central banks should be flexible or anchored. It is how to achieve both at the same time.
Credibility: openness without false certainty
The second prompt is this: credibility requires openness, and openness requires accountability.
Without credibility, policies become more costly, signals weaker and transmission less effective. But credibility cannot be built on the principle: “Trust us, we know better.”
The work of Simon Johnson, Daron Acemoglu and James Robinson has shown how fundamentally the quality of institutions shapes countries’ long-term development. Ukraine knows the value of this insight. We did not have the luxury of inherited trust in institutions.
The credibility of a central bank does not arise simply because the law grants it a mandate. It accumulates, decision by decision and crisis by crisis, and it can be lost far faster than it can be earned.
For me, credibility is memory transformed into expectations. It is the bridge between how an institution behaved yesterday and what people expect from it tomorrow.
Credibility therefore does not mean the absence of doubt, nor does it require a central bank to pretend that it knows everything. In a world of radical uncertainty, a credible institution should be able to explain clearly what it knows, what it does not know, which risks it sees, and what developments could cause it to revise its assessment.
We can be certain about our objectives and principles of action without pretending to be certain about the future.
This also changes the nature of central bank communication. Communication is not simply an exercise in eloquence. It requires an institution not only to speak but also to listen; not only to explain a decision, but also to understand how that decision has been heard.
Innovation: capability without delegated responsibility
The third prompt is this: innovation expands our capabilities, but it does not relieve us of responsibility.
Artificial intelligence is already changing productivity, labour markets, demand, inflation and the work of central banks themselves. It also expands our capacity to know: to process data faster, identify relationships that humans may miss, construct more complex scenarios and detect change earlier.
At the same time, it can deepen our dependence on systems whose internal logic we may not fully understand. This creates a new dimension of institutional responsibility.
We can delegate computation to an algorithm. We may even delegate part of the forecasting process. But we cannot delegate accountability for our decisions.
When a model fails, the model is not the one that must answer for it. When an algorithm influences a decision, society has the right to know who defined its boundaries. When technology becomes critical to an institution’s functioning, we need to understand not only its capabilities but also the vulnerabilities created by dependence on it.
Who controls the algorithm? Who is responsible for its errors? What do we do with a model whose output cannot be adequately explained? And where is the line between using a technology and becoming dependent on it?
Regulators need the courage to experiment and to give innovation room to develop. At the same time, we need the capacity to identify risks early and to set boundaries when those risks begin to threaten stability or trust.
The task of a regulator is not to stop innovation. It is to ensure that society can capture its benefits without allowing unacceptable systemic risks to accumulate.
The Yellowstone lens
These three prompts lead to another question: do we see the systems we are trying to make resilient clearly enough?
The challenge here is different from identifying existential shocks. It is not only about how we classify risk. It is about how we observe the system in which risks emerge.
Imagine walking through Yellowstone National Park. Around you are geysers, hot springs, thermal basins, steam rising from the ground and colours that seem almost unreal.
Each phenomenon can be observed, measured and studied separately. Viewed locally, each may appear to be an isolated natural feature.
Change the scale of observation, however, and the picture changes. These phenomena are surface manifestations of a vast magmatic-hydrothermal system. A person studying an individual geyser is also standing inside the caldera of a much larger volcanic system.
I would call this the Yellowstone lens.
It is possible to measure what is directly in front of us with extraordinary precision and still misunderstand the scale of the system in which we are standing.
This distinction matters for central banks. We have built an extraordinarily powerful analytical toolkit. We measure inflation, expectations, credit, liquidity, exchange rates and financial imbalances. We continuously improve our models and increase the precision of our analysis.
But greater precision in measuring individual signals does not automatically produce greater understanding of the system as a whole.
A local disturbance may be only a local disturbance. Or it may be the visible manifestation of a broader change in the system. The analytical challenge is to know when to change the lens.
This does not mean that central banks must learn to predict everything. That is impossible. It means complementing precision with perspective: understanding interdependencies, identifying vulnerabilities, and exploring scenarios in which stresses propagate across sectors, markets or infrastructure.
We need to ask not only what a particular indicator is telling us, but also what may connect it to other parts of the system. What could stop functioning first? Which infrastructure must remain operational? Where are the hidden dependencies? And which decisions might need to be taken before all the information is available?
That is where resilience, credibility and innovation come together.
We need resilience so that the system can absorb shocks and continue to perform its essential functions; credibility so that institutions retain the space to act under uncertainty; and innovation so that we can see more, understand more and respond faster.
At the centre of all three stands the institution: strong, independent and accountable, capable of adapting its instruments and expanding its capabilities when reality demands it without losing clarity about its objectives.
The answer to the challenges of the future is therefore not to learn how to predict every shock that may come next. It is to build and preserve institutions capable of performing their functions even when forecasts become obsolete, traditional instruments prove insufficient, and there is no tested recipe to follow.
Four years ago, we formulated the Ukrainian version of that principle in the NBU’s mission: to ensure the fulfilment of the NBU’s mandate under any circumstances.
That brings me back to where I began.
We expected this conference to take place in a different format. The circumstances changed and we adapted. The purpose did not.
Perhaps this is the simplest definition of resilience: to change as much as reality requires without losing sight of the goal for which you exist.
We cannot know what the next major shock will be. But we can — and must — know what must endure when it comes, and what we need to do today to make sure that it does.
Thank you!