Every Way a Trading Account Can Die

The risk management post was the weapon: sizing, stops, R-math, the daily fuse. This one is the map of what you are actually defending against, because "the market went against me" is only one of the ways a crypto account dies, and it is rarely the fastest. Most of the spectacular losses in this industry came from risks that never appear on a chart: an exchange that stopped processing withdrawals, a stablecoin that stopped being stable, a book too thin to exit, a phone number that got hijacked. Here is the full taxonomy, translated from the old forex textbooks into the market we actually trade, with what genuinely mitigates each one in 2026.

Market risk, the honest one

Price moves against your position. This is the risk everyone signs up for, the only one that is symmetric, and the only one the entire sizing playbook exists to manage. Volatility itself is not the enemy; it is the raw material, since a market that does not move pays nobody. The playbook post covers the management, so one sentence suffices here: market risk is survivable by construction when size comes from the stop, and unsurvivable by construction when it does not.

Leverage risk, the amplifier

Crypto exchanges hand retail traders 50x to 125x leverage the way bars hand out peanuts, and for the same reason: it makes you consume more. The old forex warning about 1:100 accounts transfers unchanged. Leverage multiplies market risk into liquidation risk, adds funding costs that quietly eat swing positions, and turns ordinary wicks into account events. The rule from the margin call post stands: size from the stop and leverage becomes an output; let the exchange's maximum tempt you into sizing from margin instead, and the liquidation engine becomes your real stop-loss, at the worst price available.

Liquidity risk, the exit that is not there

A stop-loss is an instruction, not a guarantee. It fills at whatever the book offers when your price trades, and on a thin low-cap perp at 4 a.m. on a Sunday, what the book offers can be several percent worse than what you planned; a planned 1R loss becomes 2R without asking permission. Liquidity risk lives in three places: coins whose books are shallow at the size you trade, hours when even deep books thin out, and event candles, when everyone wants the same exit at the same second. Mitigation is unglamorous: trade size appropriate to the book, respect the calendar, prefer the pairs where the walls in the order book are real, and remember that in a true cascade, the trader who leaves early leaves whole.

Counterparty risk, the biggest one in crypto

Every coin on an exchange is an IOU from that exchange. That sentence has cost this industry more money than every bad chart read combined: FTX vaporized customer balances in 2022, and even top-tier venues take hits, as Bybit's $1.5 billion cold-wallet theft in early 2025 showed, though that one also showed why tier matters, since withdrawals kept processing and users were made whole. The landscape did improve. Proof-of-reserves attestations are now standard at serious venues, with published ratios above 100% per asset, and regulation grew real teeth: since July 1, 2026, the EU's MiCA regime bars unlicensed exchanges from serving European users entirely, and the offshore-casino era is visibly shrinking. None of that changes the personal rules. Keep on the exchange what you actively trade, withdraw the rest to keys you control, spread size across venues if it is meaningful, and treat any withdrawal delay or "maintenance" excuse as a fire alarm, not an inconvenience. The traders who lost everything to FTX were not bad traders. They were good traders using the exchange as a savings account.

Regulatory and macro risk, the rules changing mid-game

The old textbooks called this political risk and meant elections. The crypto version is more concrete: a coin you hold gets delisted for compliance reasons, your jurisdiction gets geo-blocked, a stablecoin issuer freezes addresses, or a government decides an entire category is a security this quarter. You cannot hedge legislation, but you can avoid being concentrated in the coins and venues most exposed to it. Macro policy belongs here too, because crypto stopped being an island years ago: central bank rate decisions and liquidity cycles move this market the way they move every risk asset, which is why CPI and FOMC days appear in the event-risk rules of the sizing playbook. The interest-rate risk chapter from forex did not die; it just changed tickers.

Stablecoin risk, the cash that is not cash

Your "dry powder" is a token with an issuer, and depegs are not hypothetical; USDC spent a weekend at 88 cents in March 2023 because of a bank failure two steps away. The reserve rules that arrived with the GENIUS Act and its equivalents made the big issuers dramatically more solid, as the DeFi post covers, and they introduced the mirror risk: regulated issuers can freeze addresses. The mitigation is boring diversification: do not hold your entire cash position in one stablecoin on one chain on one venue, and know what actually backs the one you use.

Operational risk, the machine and the human

Exchanges go down, and they go down precisely when volume spikes, which is precisely when you need to close something. Apps hang, APIs reject orders, and none of the incident reports will reimburse your liquidation. Then there is the human layer: fat-fingered order sizes, market orders meant as limits, and above all account security, where SIM swaps and phishing kits harvest more retail money than any market crash. The mitigations are a checklist, not a philosophy: hardware or app-based 2FA and never SMS, withdrawal address whitelists, a password manager, size limits configured at the exchange where possible, and a healthy reflex to treat every urgent DM, airdrop claim and "support agent" as hostile until proven otherwise.

Fraud risk, the oldest one

The scams evolved from bucket shops to Telegram: rug pulls, fake exchanges with real-looking order books, "account managers" who turn deposits into screenshots, romance-flavored pig butchering that runs for months. One rule filters nearly all of it: nobody who can genuinely generate guaranteed returns needs your ten thousand dollars, and anyone promising them is charging you a fee for believing it. In a market where the real edge is small, slow and boring, "too good to be true" is not a cliche, it is a diagnostic.

The counterparty in the mirror

The last risk category holds the others' coats: you, tired, tilted, revenge-trading at 3 a.m., moving a stop because this one feels different. Every entry in this taxonomy gets worse when the operator degrades, and the operator degrades predictably, after losses, after wins, after too many hours awake. The sizing playbook's answer stands, because emotions are downstream of size, and the daily fuse exists exactly for this. The market has no memory of your last trade. Make sure you are not the only one in the room who does.

The audit that follows from the map takes ten minutes and most traders never run it: how much of my stack sits on exchanges, and which ones; what happens to my positions if my main venue goes down for an hour; which stablecoin am I actually holding and what backs it; what leverage am I really running book-wide; which of my coins dies if one regulator sneezes; and when did I last check my 2FA and withdrawal whitelist. Answer those honestly and you have done more risk management than most accounts that call themselves careful. The chart is where you make money. The rest of this list is where people lose it.