All summer long, the Grand Continent will remain in motion. Each day, wherever you are, we will bring you ideas you won’t find anywhere else (yet they will be everywhere come fall), texts that are hard to find or refreshingly new with our Sundays. To receive them straight in your inbox and to support this momentum, please consider subscribing to the review
Financial crises cannot always be prevented. While regulation and supervision can help contain risks, shocks remain inevitable. Yet, it is possible to influence what comes next. A crisis’ impact depends less on the initial disruption than on how it is answered. That is the distinction between robustness and resilience, and in the case of global financial crises, resilience requires international cooperation. It stabilizes the system and prevents national policy responses from spiraling into mutually destructive loops.
The 2008 financial crisis serves as an example. It inflicted substantial damage. Yet, unlike the interwar era, it did not unleash a cycle of retaliation among major economies. There was no broad swing toward protectionism nor a currency war. The United States and China preserved the stability of their commercial and monetary relations. Public interventions were closely coordinated within the G20. Central bank swap lines, which allowed authorities to exchange liquidity, played a crucial role in stabilizing global markets.
The current context could hardly be more different. Protectionist sentiment has largely re-emerged. So far, trade flows have shown remarkable resilience in the face of rising barriers. Yet the interwar lesson remains: once protectionism is perceived as an acceptable policy response, a financial shock can swiftly translate into contractions in trade, capital movements, and confidence that reinforce one another. There are few guardrails left. Such a shock could thus become far more dangerous than the original disruption.
Moreover, the transactional mindset that dominates today’s international relations is poorly suited to crisis management. International cooperation rarely rests on pure altruism. However, as Charles Kindleberger argued, the stability of the international system requires at least one country willing to play a stabilizing role—namely, to keep markets open and to provide liquidity when circumstances demand it. Yet today, no country—America least of all—appears prepared to shoulder that responsibility. On the contrary, essential instruments of cooperation, such as central bank swap lines, risk becoming politicized, thereby weakening one of the most effective crisis-management tools.
Against this backdrop, several fresh sources of vulnerability have emerged, capable of triggering crises. Rather than ranking them by importance, we present them in a logical order, since many are interdependent. First, risk has shifted rather than disappeared: pushed out of the regulated banking system, it has resurfaced in the less transparent recesses of finance. Second, a set of vulnerabilities envelops public finances (debt sustainability, functioning of key government bond markets, and the supply of safe assets), with fiscal and financial dominance threatening central banks’ ability to focus on price stability. Third, technological advances offer many benefits but also cause serious disruptions. Taken together, these forces point to a world where shocks may be more frequent and harder to forecast. But the principal danger remains the erosion of the international cooperation required to contain these shocks.
1 — The Mutation of Risk: Non-Bank Finance and Private Credit
Over the past decade, banks have retreated from certain segments of financial intermediation, leaving room for non-bank financial institutions (NBFIs). This sector now accounts for roughly half of global financial assets and has grown at twice the pace of the banking industry. The term covers a broad spectrum of investors, from long-horizon players like insurers and pension funds to more fragile actors, such as leveraged hedge funds backed by pension-related financing with banks, money market funds, and open-ended bond funds.
NBFIs operate as an integrated balance-sheet network. They transform maturities and rely on leverage, often embedded in derivatives, synthetic exposures, and guaranteed financing structures. Their risks resemble those of banks. Yet the complexity of their interconnections creates liquidity dependencies across institutions and markets. In a crisis, this can provoke a sudden market malfunction and a rapid loss of intermediation capacity, often beyond the reach of central-bank liquidity facilities.
The system is also more opaque. Exposures are complex, data are incomplete, and supervisory authorities frequently have limited visibility into leverage and concentrations. Private credit—the non-bank loans to firms that are issued, traded, and held by private funds—is the most striking example. The market now stands at roughly $2 trillion in size, while assets under management total around $2.5 trillion, with the bulk of that coming from the United States. The sector has been jolted by a succession of bad news, and investors in private-credit funds have tried to retrieve some of their capital. In the second quarter of 2026, redemption requests reached $22 billion, according to the Financial Times. Blue Owl found itself at the center of this retreat from private credit due to its relatively high exposure to software firms, amid the so‑called “SaaS Apocalypse.”
The ties to private equity are close: major alternative-asset managers increasingly operate both buyout platforms and private-credit arms. Insurance constitutes a second major axis. The growing integration of insurers, private-credit funds, asset managers, and reinsurers provides a stable funding base, but it also raises insurers’ exposure to illiquid and hard-to-evaluate assets. These assets are often rated by smaller agencies whose valuations can be overstated, as the Bank for International Settlements has noted.
The underlying stability risks are well known: leverage, maturity and liquidity transformation, opacity. Private credit, lacking a deposit base, is less exposed to mass withdrawals than banks; losses should, in principle, be borne by the investors. Yet systemic risk rises because of correlated strains among over-indebted borrowers unable to refinance the companies in their portfolios. The recent pushback against liquid private-credit vehicles—such as caps and withdrawal restrictions—serves as an early warning sign of forthcoming difficulties.
The regulatory response to this evolution is paradoxical. Rather than broadening oversight to cover private credit and other non-bank intermediaries, policymakers are relaxing constraints on banks to keep them competitive with their less-regulated rivals. The United States has led by easing the supplementary leverage ratio and loosening the final Basel provisions; the European Union, citing the competitiveness of its banks, is weighing a similar loosening of capital rules. The rules of the game are being harmonized downward, not upward. This episode reveals a deeper political-economic reality: resilience costs—capital buffers, forgone lending, lower returns on equity—are visible every day, while their benefits materialize only in a crisis, which, thanks to these buffers, might never occur. The payoff of resilience has natural support; resilience itself does not. As the memory of the last crisis fades, the insurance premium ends up being viewed as waste.
2 — Unprecedented Public Debt in Peacetime
By the end of 2024, OECD public indebtedness stood at 112% of GDP on a weighted basis. This is roughly 40 percentage points higher than in 2007, on the eve of the global financial crisis. Such a level has never been observed in peacetime. It is expected to rise further as aging populations push up pension and health expenditures, while new public investment needs surface in infrastructure, defense, and the energy transition.
At this level, public debt already places substantial pressure on fiscal policy. Interest payments represent about 3.3% of GDP across the OECD, a share that now exceeds defense spending in many cases and is likely to become one of the major budget lines in numerous advanced and emerging economies.
A self-reinforcing dynamic is at work. Higher debt raises risk premia and long-term interest rates. To mitigate financing costs, governments increasingly rely on shorter instruments: Treasury bills now account for roughly 15% of the debt stock, and since 2023, bill issuance has surpassed fixed-rate bonds. Shorter maturities render government debt more vulnerable. Nearly 45% of OECD sovereign debt is due to mature by 2027, and refinancing needs of this magnitude expose states to swings in funding conditions.
The trajectory of public debt is unsustainable. Higher interest rates aggravate the problem because debt sustainability hinges on the gap between real growth and real interest rates. In most countries, solvency is not at immediate risk. Yet markets have become more volatile and susceptible to abrupt shocks. The government bond markets of the world’s largest economies—especially the United States—anchor the entire financial system; any destabilization in these markets would reverberate through the monetary order, complicate monetary policy, and put the independence of central banks under pressure.
3 — Central Bank Independence Against the Risk of Fiscal Dominance
In the aftermath of the financial crisis and the pandemic, public debt levels are not only extraordinarily high but also on an unsustainable trajectory, while central-bank balance sheets remain bloated. Their growth stems from purchases of government bonds, and central banks now hold a substantial share of public debt. Within the OECD, their holdings of national sovereign bonds peaked at 29% of the total in 2021, and fell to 19% in 2024 as quantitative tightening took hold. The balance between public bond markets and central banks thus hinges on how central banks manage their portfolios: whether they shrink, stabilize, or expand their balance sheets, and how that affects interest rates and the sustainability of public finances.
This greatly alters the link between monetary policy and fiscal policy. It creates a risk of “fiscal dominance,” a regime in which monetary policy is subordinated to financing needs and debt stabilization. The risk of monetary financing has clearly increased. In this setting, the role of expectations becomes decisive: as theory and intuition suggest, merely anticipating future monetary financing can trigger inflation today.
The problem is compounded by the fact that central banks could be drawn to buy bonds in the future for two very different reasons. The first is monetary easing: lowering long-term rates to stimulate the economy when other tools are exhausted—such as when rates hit the zero lower bound. If these purchases are interpreted as monetary financing, they may fuel inflation. The second reason is market stabilization: buying government bonds can remedy a malfunction or paralysis in the sovereign-bond market, as was necessary in March 2020. A single instrument, therefore, serves two aims that may not coincide. In periods of high inflation, a central bank might be called upon to stabilize sovereign markets while maintaining a restrictive monetary stance.
The risk of ambiguity and doubt is enormous. Only a clear, unequivocal commitment to central-bank independence will allow monetary authorities to act decisively when needed, without arousing suspicions about their true motives. Independence has become a prerequisite not only for monetary stability but also for financial stability. As the Trump administration’s approach toward the Fed showed, independence is not guaranteed today.
4 — The New Fragility of Benchmark Bond Markets and the Risk of Financial Dominance
One of the major surprises of the last decade has been the fragility of benchmark sovereign bond markets, including those long regarded as the most developed and liquid. The U.S. Treasury market, a core benchmark for the global financial system, has experienced episodes of serious dysfunction, notably the March 2020 “dash for cash.” Liquidity vanished, selling broadened, bid-ask spreads widened, price gaps appeared between spot Treasuries and futures, and yields rose as macro prospects deteriorated.
Beyond the immediate trigger—the pandemic-induced uncertainty—the causes are structural. The investor base has evolved. The share of official foreign‑exchange reserve holders, once a stable foundation, has declined. Banks retreated in part because regulation and evolving business models raised the costs of trading and risk management, while debt markets expanded rapidly. Central banks themselves pulled back by engaging in quantitative tightening. Markets have thus become dependent on price-sensitive investors, such as hedge funds, households, and certain foreign buyers whose funding and risk-management models differ markedly from those of banks. Leveraged hedge funds, in particular, are active in both spot and futures markets. They rely on short-term repo financing and are exposed to liquidity shocks and margin calls.
The result is a vicious circle, what the Bank for International Settlements calls the new linkage between fiscal and financial stability. Budgetary shocks are amplified via market channels, creating structural fragility even in the most liquid markets. When faced with margin calls, leveraged investors are forced to unwind positions, amplifying price swings and tensions. This pattern was evident during the March 2020 U.S. Treasury tensions and in the autumn 2022 gilt crisis in the United Kingdom.
These episodes required large-scale central-bank interventions to restore market functioning. In the future, central banks may again be called upon to preserve the functioning of sovereign-bond markets that underpin the entire monetary and financial architecture, even if doing so runs counter to inflation-fighting aims. This is the phenomenon of financial dominance, which also blunts the stability of prices. The scenario is no longer merely hypothetical. When tensions resurfaced in the U.S. money markets in late 2025, the Federal Reserve halted its rate hikes earlier than expected, announcing in October that tightening would pause in December, effectively freezing its balance at about $6.6 trillion. Whatever the technical logic, the takeaway is clear: markets have become so dependent on public liquidity that the central bank’s balance sheet size is functionally dictated by market needs. Balance-sheet policy is thus no longer a freely deployable monetary-policy instrument. If inflation fears resurface, the room to tighten would be constrained by the need to maintain liquidity in key markets.
5 — The Erosion of Safe-Haven Status
A safe-haven asset is like a loyal friend: it stands by you when you need it most. Unlike other financial instruments, its nominal value remains stable across nearly all plausible scenarios. It serves as a benchmark and a reliable store of value. It does not depreciate, and can even appreciate during crises.
For seven decades, the U.S. dollar has been the world’s sole global safe asset. Treasuries are accessible to all, highly liquid, and regarded as virtually risk-free. The U.S. financial markets are the singular arena where foreign central banks and international investors can place large sums in a near-exclusive manner.
Recent shifts in U.S. policy and stance, however, have raised questions about the dollar’s international role going forward. They compel us to consider a scenario once deemed unthinkable: a partial, or even total, erosion of the dollar’s reserve status.
Such a shift would profoundly alter the global financial architecture. Interest rates would likely be higher and more volatile, and liquidity would become scarcer. In the absence of a credible alternative to the dollar, countries would seek protection through capital controls and other measures. The world would become more unstable, more fragmented, and consequently less prosperous. This shift would be particularly damaging for low-income economies and developing nations, which depend on access to external capital to fund investment and growth. Their borrowing costs would rise, while anti-cyclical financial support would wane.
Would alternatives eventually emerge? Europe is debating the creation of a regional safe asset through euro-denominated debt instruments, but progress remains slow. The problem is even more acute in Southeast Asia, where, in the absence of a regional safe asset, a large portion of regional savings continues to flow into the U.S. financial system.
6 — Emerging Economies and Global Stability
For many decades toward the end of the 20th century, emerging economies were a major source of global financial instability. Recurrent balance-of-payments crises repeatedly tested the world’s financial system. That is far less the case today. Crises still occur, but they are more localized and affect mainly the poorest countries.
Emerging markets are now better prepared to absorb external financial shocks for several reasons. A larger share of public debt is issued in domestic currency, reducing the “original sin” that historically translated currency depreciation into sovereign distress. Exchange rates are more flexible, helping to absorb external pressures without letting them build up. Foreign-exchange reserves are larger. Most importantly, emerging economies now benefit from stronger macroeconomic and crisis-management frameworks, with independent central banks and robust financial regulation.
But significant challenges remain. Global capital markets have evolved and can be more volatile. The same forces that reshaped national financial systems create global vulnerabilities: a growing dominance of non-bank intermediaries, heightened price sensitivity, and an increased reliance on liquidity. The system remains markedly asymmetrical: the U.S. dollar and U.S. monetary policy continue to shape global financial conditions. Major shocks and market disruptions cannot be ruled out.
7 — Artificial Intelligence and Financial Stability
AI is set to transform the finance sector. Yet its impact on financial stability remains largely theoretical at this stage. It helps to distinguish two categories of risk: those typically associated with automation in finance, which AI could amplify, and risks that are intrinsic to AI itself.
Information gathering, processing, and trading are already largely algorithm-driven. Systems rely on machine-learning models to identify arbitrage opportunities, forecast short‑term price fluctuations, and execute high-frequency trades. AI could boost the efficiency of these systems, but it could also magnify their flaws: model mis-specification, selection bias, overreliance on historical correlations, and tacit collusion between algorithms and inadequate responses to rare events. New episodes of liquidity evaporation and “flash crashes” should be anticipated. A subtler danger is the joint failure problem: if many institutions rely on the same models and data, their reactions will be highly synchronized exactly when diversity of behavior is most needed.
Information asymmetries, already pervasive in financial markets, could widen further. Large institutions—with superior IT, exclusive data sets, and advanced modeling capabilities—may gain an extra edge over smaller players. This could create entry barriers, strengthen concentration, and enhance the market power of dominant players, raising concerns about front-running, latency advantages, and large-scale information extraction. The network effects produced by data aggregation ultimately raise questions about market integrity.
Two additional risks stem from the fact that finance mirrors broader economic and societal challenges posed by AI. The first is explainability: AI systems uncover patterns humans cannot perceive, but using them to guide allocation decisions can make outcomes hard to interpret, challenge, or control. The second is alignment: AI systems pursue operational objectives that may not fully coincide with human intents. Agency problems are a constant in finance: asset managers act on behalf of investors, and mandates and incentives are designed to ensure alignment. When AI directs decisions, mandates and incentives translate into instructions, constraints, and reward schemes. Specification problems arise, especially since determining in advance what level of financial instability is acceptable, at either the micro or macro level, is nearly impossible. The machine could end up with too much discretion. It is possible to delegate thinking to the machine, but never understanding to it.
8 — Stock Valuations Tied to the Promises of AI
A sharp correction in stock markets could trigger a crisis, and there are many reasons to stay vigilant. Valuations are elevated, particularly for AI-related firms, implying earnings growth well above recent historical norms. Furthermore, market dynamics are increasingly led by a handful of AI-centered giants, especially the hyperscalers, whose outsized valuations reflect their perceived growth potential. These companies have enormous capital needs: in 2025, nine major players raised $122 billion on bond markets, roughly half of all technology-sector issuance, and their investment plans for 2026–2030 exceed $4 trillion. If revenue growth fails to meet expectations, a brutal correction could follow.
Two amplification channels deserve close attention. First, circular financing arrangements in which AI developers, chip manufacturers, and big tech buy from one another create an opaque, potentially unhealthy interconnection. A disruption in any segment of this network could trigger a chain reaction with wider market consequences. Second, investors—including households—are increasingly encouraged to take on risk through leveraged exchange-traded funds, which magnify losses and gains and can generate pronounced procyclical market moves under stress.
Moreover, AI enthusiasm masks other risk sources. The stand-off around Iran and the Strait of Hormuz is a stark example. Asset prices do not fully reflect these risks; they still presuppose that the conflict will be temporary and that AI benefits will continue to propel markets higher. The apparent resilience to geopolitical shocks may thus be illusory. If AI-driven gains prove lower than expected, the bullish dynamic that currently absorbs geopolitical risk could falter, triggering multiple concurrent corrections.
That said, technological booms—and even bubbles—do not automatically undermine welfare, provided that enthusiasm helps to overcome investment-coordination frictions and is funded by equity rather than debt. In modern financial history, equity-driven corrections have generally been less damaging than credit-driven collapses, except when stock purchases were funded by debt and margin calls forced rapid selling. Valuation corrections affect the real economy mainly through wealth effects: poorer households cut consumption, dampening overall demand. This mechanism may be more potent today due to higher household exposure to equities. Yet stock markets carry lower intrinsic contagion risk and less structural fragility than debt‑financed activities that entail significant maturity transformation. The real danger emerges when stock risk silently migrates into debt, as is currently happening with data-center expansions financed by bond debt and private-credit vehicles.
9 — Tokenization of Money and Finance
Tokenization of money and financial assets could well be the most significant financial innovation of recent decades. It means money and securities can be represented by digital tokens that circulate on the internet. Digital tokens are easy to create; they are fungible, divisible, transferable, and can be tailored to specific needs. Not surprisingly, tokenization has unleashed a wave of private-money initiatives.
The digitization of money offers major opportunities. For the general public, a digital token stored on a mobile device can faithfully replicate cash’s characteristics, without the physical constraints of weight or distance. It could spur competition in fast and cross-border payments, promote financial inclusion for the unbanked, and unlock substantial gains in the custody, trading, and settlement of securities.
Yet tokenization can also disrupt and destabilize. It threatens banks’ business models by introducing new competition in payments and deposits. It could destabilize the economy if unstable forms of private money—especially certain “stablecoins” backed by conventional assets—proliferate, and if citizens lose access to public money with the disappearance of cash. Digital tokens cross borders with ease, potentially ushering in a new era of monetary competition—not only among states, but also among private networks run by major tech firms. Digitization could therefore reshape the international monetary system itself.
To date, private investors’ exposure remains limited but is rising, and immediate stability risks to the financial system appear manageable. Nevertheless, tokenization raises fundamental challenges to currency integrity and stability. Incidents in unregulated segments, where underinformed investors could suffer large losses, could have serious political repercussions and trigger public outcry. For stability and sovereignty reasons, some central banks, notably the European Central Bank with its digital euro initiative, have decided or are seriously considering issuing their own central-bank digital currencies.
10 — Geopolitics: The Risk That Tears the Safety Net
We return to the starting point: the most significant threat to the resilience of the global economy is the erosion, or collapse, of international cooperation. Crisis triggers will arise; the crucial question is how they will be managed. The 2008 crisis did not degenerate into a spiral of protectionism and recession thanks to close G20 collaboration and the central banks’ swap lines that supplied dollar liquidity where it was most needed. Echoing Kindleberger, cooperation rests on the willingness of at least one country to stabilise the system by keeping markets open and providing unconditional financial support.
Today, that could be lacking. International relations have grown transactional and short-term, not conducive to policy coordination or prompt mutual support before financial dynamics spiral out of control. There are openly expressed concerns about the risk that swap lines become politicized and used for geopolitical ends rather than for financial stability. If global trade has thus far absorbed the rise of protectionism, it might not withstand a severe financial shock added to it.
Resilience is built before the shock, not during it. The first priority, for willing countries, is to design a shared framework for financial cooperation and to prepare for potential future contingencies, so that when the next trigger arrives, the system flexes like a reed rather than breaking like an oak.