On October 10th, 2025, crypto markets experienced their largest liquidation event in history—$16-19 billion wiped out in 24 hours. Trump announcing 100% tariffs on China was the match. But the gasoline was how crypto market infrastructure actually works, and how it fails under stress.
This isn't about crypto being volatile. This is about understanding the specific systems that broke.
The most important thing that happened wasn't Bitcoin falling. It was a stablecoin called USDe—which is supposed to always trade at $1—crashing to $0.65 on Binance while staying at $0.97 everywhere else. Same asset. Same backing. 35% price difference based purely on which exchange you used.
Binance lets you use various assets as collateral when trading with borrowed money (margin trading). You post something valuable, they let you borrow against it. Standard banking logic.
The problem: Binance was using its own internal market to determine what your collateral was worth, instead of checking what the asset was actually trading for everywhere else.
Think about it like this: imagine your bank determines your house's value by only looking at houses for sale on their own website, ignoring Zillow, Redfin, and every other house in your neighborhood. That's what Binance did.
When Bitcoin started falling and some traders panic-sold their USDe on Binance, Binance's internal price dropped. That triggered margin calls—basically the exchange saying "your collateral isn't worth enough anymore, we're liquidating you." Those forced sales pushed Binance's internal price even lower. More margin calls. More forced selling. The feedback loop spiraled until USDe hit $0.65 on Binance's system.
Meanwhile, Binance also restricted deposits and withdrawals during this period. So traders couldn't bring in USDe from outside to buy the discount and fix the price. The system had no way to self-correct.
On Curve Finance (a decentralized exchange), USDe never fell below $0.90. It stayed within 0.3% of $1 throughout the entire crash.
The difference: Curve looks at multiple price sources and lets anyone arbitrage price differences instantly. If USDe is cheap on Curve, you can mint fresh USDe at $1 from the issuer, sell it on Curve, profit from the difference, and that activity pushes the price back to $1. During the 24 hours, approximately $2 billion in USDe was redeemed across these open platforms with barely any price impact.
This isn't about "decentralized good, centralized bad." It's about system design. Open systems with multiple price checks and arbitrage proved more robust than closed systems with single price sources and centralized failed here.
On October 6th, Binance publicly announced they'd switch from using their internal prices to using external price feeds by October 14th for two other assets (BNSOL and WBETH). They knew this pricing system was vulnerable. The crash happened October 10th—right in the middle of that window.
WBETH dropped 88.7% on Binance while holding steady elsewhere. BNSOL crashed to $0.13 on Binance. These weren't real—they were artifacts of Binance's pricing system running out of liquidity.
The second failure was about liquidity—specifically, where it went when things got bad.
Market makers are firms that continuously offer to buy and sell, making money from the difference (the "spread"). In normal times, they're invisible. In crashes, they determine whether you get an orderly 10% decline or a chaotic 40% flash crash.
When Bitcoin started falling, initial liquidations hit the big markets - Bitcoin and Ethereum trading pairs. Market makers stayed there because those markets are deep enough to handle the volume and they could hedge their risk.
But market makers were suddenly holding massive amounts of crypto from buying all those liquidation orders. They needed to protect themselves. So they pulled entirely from smaller markets—altcoins, less popular tokens, anything without deep liquidity.
With market makers gone, smaller coin markets became paper-thin. A sell order that should move the price 2% suddenly moved it 20%. Results:
Bitcoin fell 10-12%. Altcoins got destroyed. That gap shows what happens when liquidity disappears.
Some coins briefly showed $0 on Binance, not because they were worthless, but because there were literally no buy orders left in the system, a different problem altogether..
The third problem was how modern crypto exchanges let you use one asset as collateral to trade another, creating chains that all break together.
Binance's "Unified Account" system lets you:
Your altcoin position depends on your Ethereum position, which depends on your Bitcoin collateral, which depends on Bitcoin's price. One long chain of dependencies.
The worst period was 21:36 to 22:16 UTC on October 10th—40 minutes when USDe, BNSOL, and WBETH all crashed on Binance. Nearly 1.7 million traders were liquidated.
Trump's tariff announcement came at approximately 20:50 UTC.
Bitcoin fell from $122,000 to briefly below $102,000. Ethereum dropped 21%. In the first hour, $7 billion in positions were liquidated.
Then the chain reactions:
During that 40-minute window, thousands of accounts using USDe, BNSOL, or WBETH as collateral got liquidated—not because those assets actually lost value, but because Binance's pricing system said they did.
Here's where it gets particularly brutal. Even traders who were profitable—who correctly bet prices would fall—got forcibly liquidated.
Auto-Deleveraging (ADL) is the exchange's nuclear option. When an account goes negative (they owe the exchange money), and the exchange's insurance fund can't cover it, the system forcibly closes out profitable traders on the other side to make up the difference.
You made money? Too bad. The exchange needs your profits to cover someone else's losses, so your winning trade gets closed whether you like it or not.
During October 10th, as one trader described it: "Leverage long? Liquidated. Leverage short? Forced deleveraging through exchange auto deleveraging."
Hyperliquid saw $10.3 billion in liquidations, with $9.3 billion on the long side. But shorts didn't escape either. The HLP vault (Hyperliquid's liquidity provider mechanism) earned approximately $40 million in fees by taking over distressed positions, but ADL was triggered aggressively across the platform. Over 1,000 wallets were completely wiped out, with more than 6,300 wallets showing losses totaling over $1.23 billion.
Market makers with hedged portfolios were particularly exposed. ADL closed out their profitable short positions while leaving them with underwater longs, destroying their carefully balanced books. As Spencer Hallarn from GSR noted, "This mechanism can create complex problems, especially for participants with more complex portfolios."
The irony: traders who correctly predicted the crash and profited from shorts had those winning positions forcibly closed by the exchange to cover other people's losses. Your reward for being right was getting your profits confiscated by the platform's risk management system.
The crash happened Friday evening US time. Traditional markets were closed. Bitcoin ETFs weren't trading. Weekend liquidity void made everything worse because there was nowhere to get fresh capital.
In a normal crash, selling pressure hits one asset. With cross-leverage and ADL:
The system designed to prevent bankruptcy instead amplified the crash.
The question isn't whether markets bounce back—they do. The question is whether this exposed design flaws or solvency problems.
Liquidity crisis: Forced selling creates a temporary price collapse, but everyone is actually solvent. Once leverage resets, markets stabilize. This appears to be what happened.
Solvency crisis: Entities that looked healthy are revealed to be broke. 2022's Terra/Luna → Three Arrows Capital → Celsius → Voyager → BlockFi → Genesis → FTX was a solvency crisis. Each failure revealed someone else's assets were gone.
Solvency crises announce themselves slowly. Firms don't immediately say "we're insolvent." They surface days or weeks later when they can't hide it anymore.
Binance knew their pricing system was vulnerable—they announced they were fixing it. The crash happened in the four-day gap between announcement and implementation. Whether deliberate exploitation or bad timing, the lesson is clear: using your own thin market to price billions in collateral creates catastrophic fragility.
The broader question: how many other exchanges use similar systems? Binance's problem was visible because of their size. Smaller exchanges with the same architecture might have proportional problems that don't make headlines but still wipe out users.
The split between Curve (0.3% deviation) and Binance (35% crash) for the same asset matters for anyone managing serious money in crypto. If your risk model says "USDe trades at $1," you were right—but only on certain platforms.
This means understanding not just what you hold, but where you hold it and how that platform's infrastructure works under stress.
Market makers who absorbed liquidation orders are sitting on underwater positions. Lending firms with Binance exposure are checking their counterparties. Smaller exchanges are hopefully auditing their collateral systems.
In 2022, the cascade took months. In 2025, with bigger players and tighter connections, it could be faster. Firms announcing problems this week were broken on October 10th. Firms waiting weeks are hoping to trade their way out—those are the ones that might trigger the next wave.
Binance announced compensation and accelerated their fix. The question is whether other exchanges learn from this, or whether we're just waiting for the next exchange-specific crisis.
October 10th exposed how crypto infrastructure handles stress. Internal price oracles, market maker withdrawal, cross-leverage chains, and auto-deleveraging aren't bugs—they're features of how these systems work. Understanding them matters if you have capital in crypto.