Breakthroughs – Cleaned Transcript Episode interview with Quirin Fleckenstein (HEC Paris) Have Banks Outsourced Financial Fragility? When credit markets seize up in crisis periods, most people still blame the banks. But new research suggests the bigger pullback may come from nonbanks like the syndicated loan market. These have grown fast but remain fragile, cutting lending harder in times of stress – with negative consequences on employment. That shift could reshape how we think about crises, jobs and financial resilience. HEC Assistant Professor in Finance Quirin Fleckenstein explores these issues as part of his research into macro-finance, financial intermediation, and corporate finance. Original research paper: https://academic.oup.com/rfs in The Review of Financial Studies, or https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3629232 (without a paywall). Sig tune Hello and welcome to this month’s Breakthroughs podcast, presented and produced by yours truly Daniel Brown. Today, we plunge into a sector that’s has a slightly surrealistic name, nonbanks. We’ll be finding out what THAT is thanks to our HEC guest from the finance department… and his name is: Pl. 1: Okay, my name is Quirin Fleckenstein. I'm an assistant professor in the finance department at HEC Paris. And I'm originally from Germany and I joined HEC in September 2023. Quirin earned his Ph.D. in finance from The Stern School of Business. For a number of years, he’s been researching Macro-Finance, Financial Inter-mediation, and Corporate Finance. Recently, he published a major study on the contribution of banks and nonbanks to fluctuations in the supply of syndicated loans. But let’s have him clarify the difference between banks and nonbanks before we discuss how the latter is responsible for most or much of the reduction in syndicated credit and employment losses during financial crises. Nonbanks? Quirin Fleckenstein: Non-banks are best defined by what they're not. They're not banks. They don't take deposits and therefore they're not regulated like banks. But they perform the same function as banks: they take money from savers and lend it out to borrowers. Daniel Brown: I've found examples like insurance firms, currency exchanges, some microloan organizations and pawn shops, which are very popular in New York City. But your study focuses on the syndicated loan market. Quirin, how would you describe it, and why this choice? Quirin Fleckenstein: The syndicated loan market is a market in which a group of lenders provides a loan. These loans tend to be too large for banks to keep on their balance sheet. This group of lenders can be banks, but it can also include non-banks. And these non-banks are not really pawn shops; they would be too small. They are investment funds, hedge funds, credit hedge funds, mutual funds and ETFs. Retail investors can invest in these funds. The largest non-banks in the syndicated loan market are collateralized loan obligations, which are also a type of fund that often raises money from insurance companies, pension funds and endowments, and then uses those funds to buy these syndicated loans. YOUTUBE EXTRACT 00: Hey everyone, welcome back to the channel. Nonbank mortgage lending now accounts for the majority of home loans in the United States, marking a major shift in the structure of the housing finance system. According to a recent review by the U.S. Government Accountability Office, nonbank lenders have steadily gained market share over the past decade. Unlike traditional banks, these firms do not accept deposits… MIKE: Lior Lustig, the New York-based the Founder of Nadlan Capital Group. After the 2008 financial crisis, stricter capital requirements made mortgage lending less attractive for banks. And consulting firms like Nadlan latched onto nonbanks to fill this gap, especially in federally backed loan programs. Between 2014 and 2024, the share of federally backed mortgages serviced by nonbanks rose from 27 to 66%. But Quirin Fleckenstein’s research shows that most of the contraction in syndicated credit and the associated employment losses is caused by nonbanks reducing their loans, especially during the Global Financial Crisis – and not the traditional banks. Why is this important? Well, this could change how you listeners interpret financial fragility in moments of crisis… like the one we’re experiencing just now. So, let’s go back to Quirin to better understand these neglected dynamics in finance and their impact on real world issues like unemployment… JINGLE… Interview Daniel Brown: In your paper, which is co-authored by researchers from Harvard Business School, Georgia Tech and the University of Washington, you say that the share of outstanding loans by non-banks rose from 22% in 2001 to 46% in 2022. How do you explain this increase? Quirin Fleckenstein: This increase is specific to the syndicated loan market, but you see very similar trends in credit markets overall. I think the main reason is probably regulation. There is good research on that. Since the global financial crisis, bank regulation has tightened a lot in response, and rightfully so. As a result, banks reduced their lending to corporations, especially riskier corporations, meaning non-investment-grade-rated corporations. The void that was left was then filled by non-banks. I think that's the main explanation for this increase, although not the only one. Quirin Fleckenstein: You can already see this rise of non-banks before the global financial crisis, when bank regulation had not really tightened. So I think another explanation that has been proposed, and for which there is some initial research, is that the decline in interest rates since the 1980s may also have contributed to the rise of non-banks. It's a bit technical because it involves a shift in banks' liability structure, which then leads to a shift in their assets. But that could be a second explanation. Daniel Brown: Researchers are always cautious about projecting beyond their data, but your paper stops in 2022 and we're now in 2026. Can you give an idea of what the trend looks like now? Quirin Fleckenstein: The trend right now looks very similar. In the last couple of years, there has been a lot of discussion about private credit, which is also provided by non-banks, although not in the syndicated loan market. There has been a very strong increase there. Bank regulation has also tightened in the last couple of years. The question is whether this trend will continue. If you ask large asset managers who are basically managing these non-banks, they argue very strongly that it will continue, but of course they may not be unbiased. Right now, though, it doesn't seem as if bank regulation is going to tighten much further, especially in the US. There isn't really the political or regulatory mood for that at the moment. So that may indicate that the trend could start to fade. Daniel Brown: For listeners who don't follow credit markets closely, what is the core question your paper asks, and what did you find? Quirin Fleckenstein: Maybe I should provide some context, there is very strong evidence in the literature of a link between the financial sector and the real economy, especially the fact that financial crises lead to recessions. One channel through which that happens is that financial institutions reduce their lending to corporations during crises. They cut back their credit supply to firms. As a result, firms reduce investment, reduce employment, and that then trickles into the real economy. There is research showing that very clearly. The question we ask is what type of institutions actually reduce their credit supply and cut back lending during these periods of crisis. More specifically, we want to understand whether banks or non-banks are the ones cutting back lending the most. To answer that question, we look at the syndicated loan market. The big advantage is that in this market, banks and non-banks are often lending to the same firm at the same point in time, which allows us to control for firms' credit demand. Otherwise, if you only compared aggregate lending by banks and non-banks, the difference might simply reflect the fact that they lend to different kinds of firms whose credit demand varies across the cycle. Daniel Brown: So, you have a control group. Quirin Fleckenstein: Exactly, with a large sample of firms. We can really compare bank and non-bank lending to the same firm. Presumably, banks and non-banks then face the same demand for funding from that firm, which allows us to isolate the difference in credit supply between them. We do that during the global financial crisis specifically, but also more generally over the credit cycle, which are fluctuations in lending to firms. What we find is that whenever credit conditions tighten and lenders cut back funding to firms, non-banks reduce lending substantially more than banks do. Daniel Brown: Why? Quirin Fleckenstein: We try to answer that question as well. What we think is the main driver is the government support that traditional banks enjoy relative to non-banks. You can see it very clearly during the global financial crisis: banks were bailed out. There are explicit guarantees in the form of deposit insurance, so retail depositors know their funds are safe. But beyond that, there are also a lot of implicit guarantees and, ultimately, bailouts of uninsured depositors. During the global financial crisis, in both the US and Europe, and again during the 2023 banking crisis with Silicon Valley Bank, uninsured deposits were protected. Banks therefore have an important advantage relative to non-banks. YOUTUBE EXTRACT https://www.youtube.com/watch?v=iUWSBAKXHeo 4 mars 2026 #StraitOfHormuz #MiddleEastConflict #Geopolitics The Strait of Hormuz, just 33 kilometers wide at its narrowest point, is a vital artery for global energy trade. Now, escalating U.S.–Israel–Iran tensions have disrupted traffic through the strait, with tankers damaged and vessels stranded. Nearly one-fifth of the world’s crude oil and a significant share of LNG supplies pass through these waters. For India, the stakes are high: about 50% of its crude imports and most LPG shipments transit Hormuz. While short-term reserves offer a buffer, a prolonged disruption could drive oil toward $100 per barrel, stoke inflation, strain the rupee and raise fuel and cooking gas costs nationwide. https://www.youtube.com/watch?v=zAmMXsfT_LY Fears of global economic consequences if Iran closes Strait of Hormuz • FRANCE 24 English … on the Islamic republic's nuclear facilities, with closing the Strait of Hormuz, a chokepoint for one-fifth of the world's oil supply, one possible option for the Iranian government. Doing so could mean major consequences for the global economy. Daniel Brown: Could we draw lessons from the crises you focused on for current disruptions, for example around the Strait of Hormuz, which may have consequences for the real economy? Quirin Fleckenstein: I think the main takeaway is that in crisis periods regulators usually want to boost lending, so they provide facilities to banks to support lending. One implication of our research is that non-banks are now very large. As you said, they provide more than 40% of the loans, at least in the syndicated loan market, and also play a major role in other markets. They are also very cyclical. So if you really want to stimulate lending, you should definitely look at non-banks as well. Regulators have started doing that. As for the current war with Iran, I think there is perhaps less concern for now because many investors still expect it to end relatively soon, so the impact on oil prices may be temporary. If you look at oil futures, for example, they are not exceptionally high, and spot prices have also eased over the last few days. But that could change if the war drags on. In the last two months there has actually been more concern about AI and the leveraged loan market. Leveraged loans are part of the syndicated loan market, and there has been concern in particular about software companies in that market. Leveraged loans are syndicated loans made to non-investment-grade borrowers, so to firms that are relatively risky. This market is very similar to the high-yield bond market, which is its equivalent on the public bond side. It is a very large market in the US, around $1.5 trillion. These loans are traded on a secondary market, which is helpful for us because we can observe in real time whether there are concerns about these borrowers. Over the last couple of months, loan prices for software companies in particular have been declining. Daniel Brown: Many people still assume that when credit dries up, banks are the main story. How important is it for you and your co-authors that people take the growing role of non-banks into account? Quirin Fleckenstein: I think it's very important. These non-banks are becoming more and more important as lenders. As we show in our study, they are also the institutions whose lending is most cyclical. That may even have become more pronounced since the global financial crisis because banks became safer. So if there is stress in credit markets, non-banks are likely to be more exposed than banks, and therefore more likely to cut back lending. Daniel Brown: Why do firms increasingly turn to non-banks? You've explained the dynamics, but what attracts them? Quirin Fleckenstein: It's a good question. Firms are often attracted by the possibility of levering up, and there can also be tax advantages. But once firms have high leverage and become riskier, it is harder for them to obtain funding from banks. Regulation really limits banks' ability to lend to such firms, so naturally these firms go to non-banks. Daniel Brown: Who are more prepared to take the risk. Quirin Fleckenstein: Exactly. But naturally, they are also the ones that cut back lending in bad times. That's the trade-off these firms face. In good times, they can lever up your balance sheets, and obtain funding from non-banks, but that funding may disappear when a crisis hits. And during a crisis, switching back to banks is very difficult because the firm's balance sheet is still too risky, and banks may also worry that they are only getting the weaker borrowers. Daniel Brown: Geographically, where do these dynamics play out most strongly? Only in the United States, or also in other Western economies? Quirin Fleckenstein: You see similar dynamics in Europe as well. Europe has a large syndicated loan market and many of the same non-banks are present as in the US. I haven't personally replicated the same analysis for Europe, but a group of Master students I supervised wrote their thesis on replicating our study with European data, and they found very similar results. Daniel Brown: Your paper shows that the role of non-banks has grown sharply over time. Does that make the structure of credit markets more fragile over the cycle? Quirin Fleckenstein: It's possible, but we try to be very careful in the paper not to make that claim directly. One reason for caution is that if the rise of non-banks is largely driven by bank regulation, and if that regulation has made banks safer, then the overall system may actually have become safer even though a greater share of lending is now done by more fragile non-banks. If safer banks can step in during crises when non-banks cut back their lending, the system may in fact be more stable. So we can't really answer that question conclusively, although it is certainly something to worry about. Daniel Brown: Our listeners often like concrete examples. Do you have an example of the main empirical finding in your paper, namely that non-banks are much more cyclical than banks? Quirin Fleckenstein: There are numerous examples, but I don't have one specific firm in mind because it's hard for us to say that a given firm definitely had credit demand and then did not get funding from a non-bank. We have to use statistical techniques and zoom out from individual cases. But yes, there were certainly companies that had trouble getting funds from non-banks. I think all kinds of sectors are affected by that. JINGLE Daniel Brown: Quirin, one striking number in the paper is that non-bank credit supply is roughly three times as cyclical as banks'. In practical terms, what does that mean? Quirin Fleckenstein: It means that when credit conditions tighten, bank lending goes down by a certain amount, and we find that non-bank lending in the same period goes down by about three times as much. So for every dollar by which bank lending falls, non-bank lending would fall by roughly an additional three dollars. Daniel Brown: Another striking result is the aggregate comparison. A one-standard-deviation increase in credit stress is associated with roughly a 70% reduction in non-bank lending, against just 17% for banks. How should non-specialists understand that gap? It's huge. Quirin Fleckenstein: Yes, it's very large. What we mean is that if credit conditions are one standard deviation tighter than in normal times, then on average bank lending goes down by about 17%, while non-bank lending goes down by about 70%. YOUTUBE N°2 extract Daniel Brown: You also show that during the global financial crisis (GFC), most of the contraction in syndicated credit was explained by non-banks. Why is that result so important? Quirin Fleckenstein: I think it's important because there is a strong narrative, and partly a true one, that banks were the main institutions cutting back credit supply during that period. We saw bank failures such as Lehman Brothers, Wachovia and Washington Mutual. It's true that banks cut back lending, but what we find is that non-banks cut back much more, and they were the main drivers of the overall reduction in credit supply in the syndicated loan market during that period. Daniel Brown: Your paper also links this to employment. You write that the majority of employment losses during the global financial crisis are the result of a reduction in non-bank rather than bank credit supply. What does that tell us about the real-economy consequences of who provides credit? Quirin Fleckenstein: What we do there is take estimates from the literature on the relationship between credit supply and employment, and on how that relationship differs across firms. We then apply those estimates to our measured reductions in credit supply by banks and non-banks. There is already ample evidence that credit supply matters for employment. So, if non-banks are the ones cutting back credit supply in crisis periods, it follows naturally that they play a large role in the employment losses that result from those contractions in the syndicated loan market. Daniel Brown: Your model emphasizes government support as a key reason banks are more stable over the cycle. Can you explain that mechanism simply? Quirin Fleckenstein: Basically, governments often guarantee banks' liabilities. That means banks can continue operating more or less as before, because depositors and other providers of funding know that the government may step in. They are therefore willing to keep lending to banks even when banks' balance sheets look weak. Non-banks do not enjoy support to the same extent, so their own lenders become more reluctant to fund them, which forces non-banks to cut back lending more. Daniel Brown: How did you test whether the result was really about non-bank credit supply rather than bank health, information asymmetries or changing borrower characteristics? Quirin Fleckenstein: To give a bit more context, in the syndicated loan market both bank and non-bank loans are originated by banks. A borrower may go to its relationship bank, for example Goldman Sachs, and Goldman Sachs arranges the loan deal. Some of that loan is then sold off to non-banks, while some stays on Goldman's own balance sheet or on the balance sheet of other banks. One concern is that our result might be driven by a selection of more fragile banks originating more of the non-bank loans. When a bank arranges a non-bank loan, it still has to keep it on its own books for at least a few days, sometimes weeks. If those arranging banks are themselves less healthy, they might create a bottleneck in the non-bank loan market. In that case, the decline in non-bank lending would not reflect non-banks being unwilling to lend, but rather the fragility of the banks that originate those loans. To rule out that explanation, we compare bank and non-bank loan originations within the same bank. In other words, we look at Goldman Sachs and ask whether, in a crisis period, Goldman reduces non-bank originations relative to bank-loan originations. That is exactly what we find, which allows us to rule out bank health as the main explanation. Daniel Brown: Is the key message of your paper that the system has not necessarily become safer as more lending moves outside traditional banks, or would you phrase it more carefully? Quirin Fleckenstein: I would phrase it more carefully. As I mentioned earlier, it is hard for us to rule out the possibility that banks became so much safer through regulation, and that this shift of credit intermediation to the non-bank sector still left the overall system safer. If that is true, then the rise of non-banks may not be too worrying. But if non-banks rose Wachovia for other reasons and that rise was not offset by banks becoming much safer, then I do think the increase in non-banks is worrying. Daniel Brown: Looking ahead, what is the single most important takeaway for policymakers from this paper? Quirin Fleckenstein: The key takeaway is that if policymakers want to stimulate lending during crisis periods, they really have to look at the non-bank sector. And if they think it is appropriate, they should also try to stimulate lending there. Daniel Brown: And for listeners in business who are not policymakers but do care about financing conditions and economic resilience, what should they take from your paper? Quirin Fleckenstein: Especially for firms, the key takeaway is that borrowing from non-banks may give you very attractive credit conditions in good times, but that money can disappear very quickly in downturns. Firms need to be aware of that. And if you lever up too much and become no longer eligible for bank loans because you are too risky, you may face problems in bad times because you won't find any lender willing to lend to you. Daniel Brown: You focus mainly on the United States, and also on Europe, as we discussed earlier. What about emerging economies such as China, Brazil or India? Is the issue of non-banks important there too? Quirin Fleckenstein: It's a good question. Certainly there are non-banks that are becoming more active in those markets as well. The difficulty is that we often lack the detailed data needed to study banks and non-banks precisely in emerging markets. Daniel Brown: Quirin Fleckenstein, thank you very much. Quirin Fleckenstein: Thank you. MIKE: Quirin Fleckenstein, who cosigned his paper with fellow academics Manasa Gopal, German Gutierrez, and Sebastian Hillenbrand. It’s called “Nonbank Lending and Credit Cyclicality” and is available online for free. The research really underlines how nonbanks pull back harder in times of crisis with human and economic costs that policymakers often underestimate. And it poses the question, have we shifted risk out of banks without making the system less fragile overall? Well, that rounds up this month’s Breakthroughs. Next time round, we’ll be inviting Olivier Darmouni. He’s Associate Professor and the Pierre Andurand Chair in Sustainability at HEC. And Olivier will be sharing his research on the financing challenges and opportunities in the energy transition. This rising star in sustainable finance was recently invited to France’s National Assembly to analyze the impact of hedge fund activity on industry’s production capacity. So plenty to chew over in the next Breakthroughs podcast. Meanwhile, keep your comments coming in to brownd@hec.FR. That’s… Until next time, good-bye. Quirin Fleckenstein: The project with Sebastian is still a work in progress. We don't have a draft yet. But there is a puzzling finding in the literature: the spread, or basically the interest rate, that firms have to pay on loans is substantially higher than what they have to pay on bonds, once you correct for differences in risk. Quirin Fleckenstein: That's quite puzzling, especially because the study partly concerns loans that are traded on a secondary market, including leveraged loans of the kind I mentioned earlier. We can observe prices for these instruments, and retail investors can invest in them through ETFs and mutual funds. The interest rates on these loans are then being compared with bonds in which retail investors can also invest. So these two markets should not be as segmented as this fact in the literature seems to suggest. Quirin Fleckenstein: So we are trying to solve that puzzle. That's basically what the project is about. Daniel Brown: So in general terms I can say that you and Sebastian are studying that, without jeopardizing anything confidential. Quirin Fleckenstein: Sure. Daniel Brown: Great. Ambiguities to check against the audio Time code Note 00:05:26 Original transcript reads 'in this conjecture and research'. I cleaned this to 'in the literature', but the source wording should be checked. 00:06:53 The raw transcript gives '70% reduction in non-bank lending' and '17% for banks' in the question, but the reply repeats '70%' twice. I normalized the answer to '17%' for banks and '70%' for non-banks; worth confirming against the audio. 00:09:21 Original transcript reads 'these nomics' and 'circular loan market'. Cleaned to 'non-banks' and 'syndicated loan market'. 00:10:34 Original transcript reads 'significant loan market'. Cleaned to 'syndicated loan market'. 00:13:12 Original transcript reads 'fight in non-banks'. Cleaned to 'rise in non-banks'. Background 00:00:17 Original transcript reads 'there's a positive finding in the literature'. I cleaned this to 'there is a puzzling finding in the literature', but that adjective should be checked against the audio.