Finn Brooks: Hey — you look like you've been doing the same thing I've been doing, which is staring at one number and feeling slightly unwell.
Clara Bennett: Mm, the thirty-to-one figure. Yes. I've written it down three times this week as if writing it again would make it feel less strange.
Finn Brooks: Thirty-to-one leverage — that's where we're starting today, that's the thing we're trying to actually understand. Not just name it, but like — understand how a number that clean causes a crisis that large.
Clara Bennett: So let's name the mechanism first. A leverage ratio is total asset exposure relative to equity. At thirty-to-one, you control thirty dollars of assets for every dollar you own. The consequence: a three-point-three percent fall in asset value wipes out equity completely.
Finn Brooks: Three point three percent. Pre-2008 investment banks. That's — hold on, that's what actually happened, right? That's not a hypothetical stress test.
Clara Bennett: That's what happened. Mortgage-backed assets declined by just over three percent. The institutions running at those ratios crossed their own wipeout threshold, forced liquidations started, and because those institutions were interconnected — distress spread.
Finn Brooks: No but here's the part that keeps tripping me up — leverage is structurally symmetric, right? The same math that wipes you out on the downside is supposed to make you extraordinarily rich on the upside. So why does one direction feel so different from the other?
Clara Bennett: That's the puzzle. And importantly, it's the same arithmetic as a mortgage — someone puts ten percent down on a house, they're at roughly ten-to-one on that position. We don't treat that as catastrophic.
Finn Brooks: Which is genuinely wild to me, that we've sorted leverage into 'normal life thing' and 'system-destroying thing' and the math doesn't actually explain the sorting.
Clara Bennett: Right — and that's the question the whole episode is really turning on. Not 'what is leverage' but 'why does identical math produce outcomes this far apart.'
Finn Brooks: Because three point three percent — I want to just sit with that for a second. That is a Tuesday. Markets move that much. And the entire equity cushion at these investment banks was... gone.
Clara Bennett: Yes. And the crisis didn't wait for a dramatic number. It just waited for three point three.
Finn Brooks: And that's — okay, that's what I keep bouncing against. Because the math sounds like it should work both ways equally, right? Like if a 3.3% drop erases you, a 3.3% gain should double you. Perfect symmetry on paper.
Clara Bennett: It does. That's actually not a flaw in the framing — the structural symmetry of leverage is real. A ten-to-one ratio multiplies a one percent gain by ten, and a one percent loss by ten. The math doesn't care which direction.
Finn Brooks: So then why does losing feel so — wait, actually I think I know where I'm going wrong.
Clara Bennett: Go.
Finn Brooks: Let's say I have ten dollars. I borrow ninety. Now I'm holding a hundred dollars of something. It goes up ten percent — I've made ten dollars on my ten dollars. I've doubled my money. That's wild but fine. Now it goes down ten percent — I've lost ten dollars. My ten dollars is completely gone. Not halved. Gone.
Clara Bennett: That's the click. And now add one thing: where does the borrowing actually happen? In a margin account — a broker holds your existing securities as collateral and extends you the borrowed portion. The moment your equity dips below a threshold they set — the maintenance margin — the arithmetic stops being theoretical.
Finn Brooks: Meaning they call you.
Clara Bennett: They issue a margin call. Deposit more funds immediately, or we liquidate your position. And — in practice, the thing people don't sit with — that call doesn't arrive when markets are calm. It arrives when prices are already falling. Which means you're forced to sell into a falling market.
Finn Brooks: Which pushes prices down more. Which triggers the next person's margin call.
Clara Bennett: Exactly — that's the forced liquidation loop. Pro-cyclical selling. And that's where the symmetry actually breaks down, not in the math, but in the timing. The upside is patient. The downside has a deadline.
Finn Brooks: Okay that — the upside is patient, the downside has a deadline — that's the sentence I needed. Because gains don't force your hand. Losses literally do.
Clara Bennett: And that's the layer the symmetry argument misses. The math is symmetric. The market's response to your losses is not.
Finn Brooks: And that timing thing — I keep pulling at it because, okay, imagine you're a retail investor in October 1929. You bought on margin. Not a bank, just a person with a margin account. The price dips — not crashes, just dips — and the broker doesn't wait. The call is instant. You have to sell. Right now. Into the same market where everyone else just got the same call.
Clara Bennett: That's the 1929 collapse in miniature. Retail investors, margin accounts, simultaneous calls. Not one person making a bad decision — thousands of forced liquidations hitting at the same moment, all selling into the same falling prices.
Finn Brooks: All at once.
Clara Bennett: All at once. Now scale that to 2008. The same mechanism, but now the entities running at thirty-to-one are investment banks, and they're not holding independent positions — they're holding the same mortgage-backed assets, cross-collateralized across institutions. One bank's forced liquidation depresses the price of those assets. Which means the collateral posted at the next bank is now worth less. Which triggers their margin call.
Finn Brooks: So the contagion isn't — wait, it's not that they all made the same bad bet. It's that the forced selling by one of them *changes the math* for all of them.
Clara Bennett: That's exactly the distinction. The collateral loses value because of the selling, not because the underlying asset fundamentally changed overnight. The liquidation creates the price drop that justifies the next liquidation. That's the pro-cyclical loop.
Finn Brooks: No but seriously — that's almost like a self-fulfilling crisis. The mechanism *produces* the conditions that make the mechanism fire again.
Clara Bennett: It is. And the Financial Stability Board has flagged this specifically — not just banks, but hedge funds and other non-bank financial intermediaries carrying interconnected leverage. Because if the channel is the same, the size of the institution doesn't actually protect you from contagion.
Finn Brooks: The FSB is looking at hedge funds as a systemic risk vector right now? I mean — I knew there were concerns, but that's... that's not historical, that's live.
Clara Bennett: Ongoing concern, yes. The interconnectedness is the variable. A hedge fund at high leverage, in isolation, is a problem for that fund. A hedge fund at high leverage, whose collateral overlaps with three prime brokers, is a contagion channel. The maintenance margin threshold is the same; the blast radius is different.
Finn Brooks: Which brings me to the question I can't shake — if this mechanism is that reliable, if we've seen it in 1929, in 2008, if the FSB is still watching it unfold in non-bank institutions — why did we let investment banks run at thirty-to-one in the first place? Like, that wasn't a secret ratio.
Clara Bennett: It wasn't hidden, no. The leverage ratios were published. Regulators saw them. So the question you're really asking is whether the regulatory response after 2008 actually addressed the mechanism — or just moved where it accumulates. And that, I think, is where things get genuinely uncomfortable.
Finn Brooks: Because — okay I have a suspicion, and it involves the EU's 1:30 retail cap and the fact that capping the tool in one place might just be pushing the pressure somewhere darker. But we should probably — that part gets worse before it gets better.
Clara Bennett: Much worse. The short version is: you can cap leverage ratios. You cannot cap the incentive to find leverage somewhere the cap doesn't reach.
Finn Brooks: And the EU cap is the clearest proof of that, right? Like — they set 1:30 for retail forex CFDs, a hard ceiling, and 70 to 85 percent of retail CFD clients still lose money. The cap didn't fix the problem. It just bounded the catastrophe.
Clara Bennett: That's the distinction worth holding. The 1:30 cap addresses magnitude — how much you can lose in a single event. It does nothing about the underlying structural problem, which is that forced liquidation happens at the worst possible moment regardless of the ratio.
Finn Brooks: So you're capping the blast radius, not defusing the bomb.
Clara Bennett: And then — now genuinely uncomfortable — the risk doesn't disappear. It migrates. Cap retail leverage, and the capital that wants higher leverage moves to hedge funds. Cap the investment banks post-2008, and the leverage accumulates in what the FSB calls non-bank financial intermediaries. Shadow banking, essentially.
Finn Brooks: Wait — the FSB is saying post-2008 reforms redistributed the risk, not removed it? That's — that's a pretty damning thing for a regulatory body to admit about its own era of reform.
Clara Bennett: They're still monitoring highly leveraged institutions as an ongoing systemic concern, yes. Which tells you everything about what the reforms actually accomplished. And SEBI — India's regulator — has its own margin trading facility rules, its own leverage compliance architecture. Completely separate regime. So capital that finds the EU cap too tight, or post-2008 bank requirements too restrictive, can seek jurisdictions where the patchwork has gaps.
Finn Brooks: And VaR — Value at Risk — that's the tool regulators are actually using to set those margin requirements, right? The statistical model?
Clara Bennett: It is. And here's — I mean, this is the part that should make anyone who works in risk management uncomfortable. VaR calculates potential worst-case losses based on historical correlations. But in a forced-liquidation event, correlations spike. Everything sells off together. The model's estimate of tail risk is constructed from data that doesn't include the very conditions that make tail risk most dangerous.
Finn Brooks: So the tool fails exactly when you need it.
Clara Bennett: Precisely when you need it. Because VaR was calibrated on periods of normal market function. When correlated forced selling begins — which is the margin call cascade — the actual losses exceed what VaR predicted, so the collateral requirements that were set using VaR turn out to be insufficient.
Finn Brooks: That's — okay I genuinely did not clock that before. The margin requirements are built on a model that systematically underestimates the scenario where margin requirements actually get triggered. That's not a quirk, that's structurally broken.
Clara Bennett: Which brings you to the hardest wall in this whole conversation. A mortgage is leverage. Corporate debt is leverage. Operating leverage — the way fixed costs amplify profit swings before you borrow a single dollar — that's leverage too. You cannot eliminate leverage without eliminating credit itself. The policy question was never 'can we remove it.' It was always 'where, at what ratio, and for whom.'
Finn Brooks: And the answer keeps being — we cap it here, it appears over there. We regulate banks, shadow banking grows. We protect retail with 1:30, hedge funds are operating on something else entirely. So what actually constrains it? Like, is there a constraint that isn't just a pressure valve pointed somewhere new?
Clara Bennett: That's the question this regulatory architecture hasn't answered. The margin call mechanism is structural — it's built into collateral and maintenance margin, into how leverage functions at any ratio. Caps move the pressure. They don't change the mechanism. And until someone answers that honestly, the shell game continues.
Finn Brooks: The thought I can't shake is — we keep asking where the leverage went after each crisis, right? Moved from banks to shadow banking, from retail to hedge funds. But I think maybe that's the wrong question. I think the actual question is who gets to fail.
Clara Bennett: 2008 answered that implicitly, didn't it. Not through policy — through the crisis itself. The system accumulated leverage until it broke, and then we decided which institutions to save. That decision wasn't made in advance. It was made under duress.
Finn Brooks: So the real constraint was never a ratio. Not the thirty-to-one, not the EU's cap, not VaR. The actual constraint is just — how much pain the system absorbs before it acts.
Clara Bennett: And that tolerance is revealed, not declared. No regulator sets it in advance. The Financial Stability Board monitors, SEBI enforces margin requirements, the EU caps at 1:30 — all of that is real. But the systemic tolerance for contagion? We discover it by crossing it.
Finn Brooks: Which means — and this is the part that sits really uneasily with me — we started this conversation staring at that thirty-to-one number like it was the problem. And now it feels like the number was almost beside the point.
Clara Bennett: The number was always transparent. Three point three percent, published ratios, visible to regulators. The fragility was structural and the tolerance for it was social. And we chose, each time, to find out where the limit was rather than set it. That's where I keep landing.
