Financial crises haven't disappeared since 2008 — they've simply gone quiet, and Deutsche Bank Research Institute's Henry Allen, Peter Sidorov, CFA, and Camilla Siazon set out to explain why in their November 3, 2026 report1. The Team opens with the uncomfortable premise economist Hyman Minsky left behind: "stability is destabilising." Stretches of calm breed the very complacency that eventually detonates. Yet nearly two decades on from the Global Financial Crisis, "we have yet to see another financial crisis happen on a scale that had a lasting market and macroeconomic impact." The Team's question is simple — why?
The Architecture of Calm
The Team identifies four forces behind the post-2008 resilience. First, "post-GFC deleveraging cleared out risks in the system." Second, reforms like Basel III and Dodd-Frank were "explicitly designed to avoid a repeat of 2008." Third, "the share of financial assets held by banks has fallen since the 2008 crisis," shifting risk toward non-bank financial institutions (NBFIs) that are "not as prone to these bank run dynamics." Fourth — and perhaps most consequential — "the lessons of 2008 have meant that when the financial system has had wobbles since, policymakers have acted with incredible speed to stop the turmoil," evident in both the 2020 Covid response and the 2023 regional banking panic, when the Fed, Treasury and FDIC moved within days to backstop SVB depositors.
But this resilience carries a cost. The Team notes that "the shift in debt towards sovereign balance sheets has cushioned private-sector shocks but narrowed the capacity for future intervention" — government debt has gone "from the smallest to the biggest category" of global debt since 2007, leaving less fiscal room for the next rescue.
Three Candidates for the Next Crisis
The Team then turns forward-looking, examining where today's risk is accumulating. On artificial intelligence, the analysis is notably measured: "a reversal in the investment cycle need not cause a crisis in itself," just as the dot-com bust produced a recession but not a systemic event. Applying the Greenwood-Hanson-Shleifer-Sørensen framework, the Team finds equity valuations already in "red zone" territory, but credit growth "has been considerably more subdued" — for now. Still, "approaching that territory by 2028 is conceivable if the AI cycle continues to accelerate." Their conclusion: "an AI correction is a clearer risk than an AI-led systemic crisis at this point."
On private credit, the Team draws a careful distinction: "parallels with the subprime credit boom have been periodically made. However, we think these should not be overstated," citing slower build-up and redemption restrictions. Still, warning signs are mounting — "weaker new inflows into BDCs, increasing payment-in-kind interest and prospective refinancing pressures." Crucially, the Team flags that "redemption gates can limit immediate cash outflows... but they do not restore a borrower's ability to service debt or absorb underlying credit losses."
On sovereign stress, the Team identifies a structural shift worth watching closely: "the changing structure of sovereign markets suggests that the sovereign–NBFI feedback loop may now warrant increasing attention," pointing to hedge funds' sovereign bond exposures, which have "more than doubl[ed] since 2022."
The Unresolved Core
The Team's conclusion resists false comfort: "the problem of financial instability has yet to be permanently solved... risks often shift rather than disappear." Most tellingly, "financial crises often start in areas where people are paying little attention, or risks are assumed to be contained."
Five Takeaways for Advisors and Investors
- Resilience is structural, not accidental. Post-GFC reforms, bank deleveraging, and faster policy intervention have genuinely reduced systemic fragility — this isn't luck.
- Watch non-bank finance, not just banks. The Team's own words: future stress "may originate outside banks," making traditional crisis indicators less complete.
- AI is a valuation risk before it's a systemic one. Clients should distinguish between a painful correction and a crisis — the Team sees the former as far more likely near-term.
- Private credit bears monitoring, not panic. Liquidity gates mask solvency problems; the Team's distinction here is a useful client-conversation framework.
- Fiscal space is the real constraint. With sovereign debt now the dominant category, "the margin for a future policy error is limited" — a theme advisors should weave into longer-horizon risk conversations.
Footnote:
1 Allen, Henry, Peter Sidorov, and Camilla Siazon. Why Hasn't There Been a Major Financial Crisis in Recent Years? Deutsche Bank Research Institute, 3 Nov. 2026.
