Latest / Monetary Policy Explained with Fexingo: Central Banks, Money Supply, and Interest Rates / How Central Banks Hunt Financial Contagion
Transcript
- Lucas: So back in early February 2026, a mid-sized bank in Luxembourg starts seeing deposit outflows. Nothing massive — maybe 2 percent of deposits over a week. But the interbank market gets nervous. Banks start charging each other higher rates to lend overnight. Luna: And that's the moment where a localized problem turns into something bigger, right? If one bank can't roll over its short-term funding, suddenly everyone's looking at their counterparty risk. Lucas: Exactly. And this is where the central bank's contagion toolkit comes in. The ECB doesn't wait for a full-blown crisis. It watches a specific data point: the spread between unsecured interbank rates and overnight index swaps. When that spread widens more than 15 basis points in a day, they go into monitoring mode. Luna: Fifteen basis points — that's a pretty tight trigger. So they're looking for the earliest sign of distrust, not a bank run. Lucas: Right. The actual trigger for action is different. They look at the volume of emergency borrowing from their standing facilities. If more than three banks in the same region draw on the marginal lending facility in one week, they assume contagion may be underway. Luna: And that's exactly what happened in the Luxembourg case? Three banks tapping the facility? Lucas: Four, actually. Two from Luxembourg, one from Belgium, one from Germany — all with indirect exposure to that original bank through a common corporate borrower. So the ECB's financial stability team activates a pre-planned response: a targeted longer-term refinancing operation, or TLTRO, but with a twist. Luna: What's the twist? Lucas: Normally, TLTROs are cheap funding to encourage bank lending. This one carried a penalty rate — 50 basis points above the main refinancing rate. The idea is: you only come to this auction if you genuinely need liquidity, not because you want cheap money. It's a stigma-reducing mechanism, actually. Luna: Stigma-reducing? By making it more expensive? Lucas: Counterintuitive, I know. But here's the logic: in a crisis, healthy banks refuse to take emergency central bank funding because they're afraid it signals weakness. So they hoard liquidity instead, which worsens the crunch. By setting a penalty rate, the central bank says, 'This is not a bailout, it's a business decision. Only take it if you need it, and there's no shame.' It normalizes the borrowing. Luna: So the penalty rate actually makes banks more willing to participate, because it signals the facility is for everyone, not just the desperate. Lucas: That's exactly the theory. And in practice, it worked in February. Seventeen banks bid in the auction, total take-up was about 8 billion euros. The spread on interbank rates narrowed within two days. The ECB published the aggregate results but not individual bank names, so no one could be singled out. Luna: So the transparency is calibrated — enough to reassure markets, not enough to trigger a run on any specific bank. Lucas: Exactly. And this is the key difference from 2008. Back then, central banks had to create entirely new facilities from scratch — the Fed's Term Auction Facility, the ECB's Long-Term Refinancing Operations at fixed rates. Now they have pre-announced, standing tools with built-in triggers. The playbook is written before the fire starts. Luna: But what if the auction doesn't work? What if seventeen banks bid but the spread keeps widening? Lucas: Then you escalate. The ECB has a second layer: outright purchases of bank bonds — covered bonds, not sovereign debt — to directly support bank funding costs. That's what they did in 2023 during the Silicon Valley Bank aftershock, when European banks with large bond portfolios came under mark to market pressure. Luna: And that worked? The covered bond purchases? Lucas: They stabilized the market within a week. The key was the ECB announced it would buy bonds from any bank that met a minimum credit rating, not just the ones under stress. That way, the market couldn't infer which banks were in trouble from the purchase data. Luna: So the central bank is basically using its balance sheet to provide a floor for bank funding, but without naming names. Lucas: Correct. And the third layer — the nuclear option — is what the Bank of England did in September 2022 during the gilt crisis: they buy government bonds to stabilize the collateral that banks use for repo funding. That's even broader, and it's what central banks want to avoid, because it looks like they're bailing out fiscal policy. Luna: Let's talk about detection for a second. How do central banks actually spot contagion early? They can't just watch every interbank transaction in real time. Lucas: They do, actually. Most major central banks now have real-time monitoring systems that track payment flows between banks. The ECB has something called TARGET — the Trans-European Automated Real-time Gross settlement Express Transfer system. They can see if Bank A suddenly stops sending payments to Bank B, or if settlement times spike. Luna: And that's a leading indicator — if a bank starts hoarding liquidity by delaying outgoing payments, you catch it before deposits even start to leave. Lucas: Exactly. In the Luxembourg case, the ECB's TARGET monitors saw that one bank had delayed settling three large payments — each over 100 million euros — by more than four hours. That was the earliest signal, before the deposit outflows even appeared in the data. Luna: So they're looking at operational data, not just financial ratios. Lucas: Right. And they combine that with a network model that maps interbank exposures. Each large bank files a weekly report of its top 20 counterparties and the net exposure. The central bank runs a simulation: if Bank A defaults, which other banks take a hit, and how much capital do they lose? Luna: So they're stress-testing the network topology, not just individual balance sheets. Lucas: Exactly. And that's how they knew the Luxembourg bank's troubles could spread to four other institutions in a cascade. The pre-planned TLTRO was designed to inject liquidity at the nodes that were most vulnerable — not just the original bank. Luna: So the central bank is essentially re-routing liquidity through the network, like a traffic controller. Lucas: That's the metaphor. And the beauty is, the banks themselves don't realize they're being targeted. The auction is open to all, but the terms — the penalty rate, the collateral eligibility — are calibrated so that the vulnerable banks find it attractive, while healthy banks don't bother. Luna: So the central bank is using price to allocate liquidity to the right places without signaling. Lucas: Exactly. And this is a huge shift from the old approach, where central banks either did nothing and let the market sort it out — which often led to contagion — or they bailed out specific institutions, which created moral hazard. Now they use market mechanisms within a pre-set framework. Luna: There's one thing I'm still not clear on. How do central banks decide when to stop? When does the intervention end? Lucas: There's a specific exit trigger. The ECB uses the volume of the penalty-rate TLTRO. Once the total outstanding drops below 1 billion euros for two consecutive weeks, they stop offering new auctions. It's a self-liquidating facility — banks pay back the loans as their liquidity improves, and eventually the facility winds down. Luna: So the central bank doesn't have to make a subjective call about when the crisis is over. The market decides. Lucas: Right. And if the facility doesn't wind down — if banks keep borrowing at the penalty rate for more than six months — that's a signal of deeper problems that require a different tool, like a capital injection or a restructuring. Luna: Which is exactly what happened in a few cases, I think, in 2023 with some regional US banks. Lucas: Correct. The Fed's Bank Term Funding Program was designed similarly — penalty rate, collateral haircut, but it was meant to be temporary. When some banks stayed in the program for over a year, the Fed worked with the FDIC to facilitate a merger or wind-down. Luna: So the emergency tool is really a bridge to either recovery or resolution. Lucas: Exactly. And that's the key principle: central banks provide liquidity, not solvency. If a bank is fundamentally solvent but facing a run, the central bank lends. If it's insolvent, the central bank steps back and lets the resolution authority handle it. Luna: That's the theory. But in practice, the line between liquidity and solvency can blur fast. Especially when asset prices are falling and mark to market losses hit capital. Lucas: That's the hardest judgment call a central bank makes. And it's why they rely on that network model and the real-time data — to distinguish a liquidity problem from a solvency problem. In the Luxembourg case, the bank had a 12 percent capital ratio and its losses were from a single corporate loan that went bad. That's a liquidity problem. So the TLTRO was appropriate. Luna: And if the model had shown that capital would be wiped out after a 5 percent default rate on its corporate loan book, then they'd have known it was solvency and called in the resolution authority. Lucas: Exactly. And the resolution authority — for the euro area, that's the Single Resolution Board — has its own pre-funded mechanism, the Single Resolution Fund. So the central bank doesn't have to cross the line into fiscal policy. Luna: So the system is designed with firewalls between liquidity support and resolution. Lucas: It is. And that's the big institutional change since the financial crisis. Central banks now have a clear, limited mandate for contagion control, with specific triggers, specific tools, and specific exit criteria. It's almost like a machine. Luna: A machine that only activates when the right data inputs are hit. Lucas: Right. And that's what makes the February 2026 episode so interesting — it was the first real test of the post-2023 playbook. And it worked. The spread normalized, no bank failed, and the facility is already down to 2 billion euros as of last week. Luna: That's a pretty clean outcome. And honestly, understanding how that machine works is valuable — not just for finance professionals, but for anyone who wants to know why their bank doesn't collapse overnight. Lucas: Speaking of value — if today's breakdown was useful to you, the reason we can explore these details without any ads is listener support. It's buy me a coffee dot com slash fexingo. That's what keeps this show ad-free and allows us to go deep on stuff like TLTRO mechanics and network contagion models. Luna: Yeah, and it really does make a difference. Every contribution helps us keep the episodes focused on what actually matters, not on selling you something. Lucas: So, back to the playbook. One element we haven't touched on is international coordination. When a contagion spans multiple jurisdictions, central banks have to act together. The Luxembourg bank had a branch in London, so the ECB and the Bank of England had to align their TLTRO and the Bank of England's Indexed Long-Term Repo simultaneously. Luna: And they did that in February? Lucas: They did. The two central banks issued a joint press release — just three paragraphs — stating that they were offering liquidity facilities in coordination. No details on amounts, no bank names. The market interpreted it as a signal of preparedness, which itself reduced the contagion risk. Luna: So the communication itself is a tool. Lucas: Exactly. The joint statement was carefully crafted to be vague enough that no bank was identified, but specific enough that markets knew the central banks were watching. That's the art — and it's something central banks have gotten much better at since 2008. Luna: And what about the future? Are there any gaps in this playbook that worry you? Lucas: The biggest gap is non-bank financial intermediaries — money market funds, hedge funds, insurance companies. They don't have direct access to central bank lending facilities. In 2020, the Fed had to create new facilities for money markets. That area is still under construction in terms of contagion tools. The ECB is working on a standing facility for money market funds, but it's not in place yet. Luna: So the next crisis might come from the shadow banking sector, and the central bank's playbook is still incomplete. Lucas: That's the frontier. And it's where a lot of researchers are focused. But for traditional bank contagion, the toolkit is mature. And the February 2026 episode showed it works. Luna: Good to know the machine is ready. Lucas, thanks for walking through that. Lucas: Always a pleasure. And for those interested, we'll link the ECB's February press release and the TLTRO terms in the show notes.