The First 24 Hours of a Blowup

A strategy that has run cleanly for years suddenly loses a large amount of money in a single day, and nobody is entirely sure why yet. This is one of the most stressful moments in trading, and it is also one of the most predictable in the sense that the same mistakes recur across nearly every well-documented blowup: doubling down because the position looks "too cheap to sell now," delaying an honest assessment because admitting the scale of the problem is uncomfortable, and treating the first day as a normal trading day when it should be treated as a discrete crisis with its own playbook.

What the first hours are actually for

The first 24 hours after an unexpected large loss are not for figuring out exactly what went wrong — that takes longer and shouldn't be rushed. They are for containing further damage, establishing the facts that are actually known versus assumed, and making sure the right people are aware before decisions get made under pressure by too few people with too little information.

Priority	What it involves
Confirm the numbers are real	Rule out a data feed error, a booking mistake, or a system bug before treating the loss as a genuine market event
Stop the bleeding	Reduce or hedge the exposure driving the loss to a size the desk can tolerate while the situation is understood, without necessarily exiting everything in a panic
Notify the right people immediately	Risk management, senior traders, and compliance should hear about a large unexpected loss from the desk directly, not discover it later
Preserve records	Trade logs, system states, and communications from the period should be preserved untouched, both for the eventual post-mortem and for any regulatory or investor reporting obligation
Separate known facts from working theories	"We lost money because of X" is often wrong in the first few hours; state clearly what is confirmed versus suspected
Avoid irreversible decisions made in isolation	A single trader deciding alone to significantly resize or unwind a complex book under stress is a common way an initial loss becomes a much larger one

Worked example

A statistical arbitrage book loses several times its typical daily volatility in a single session. Within the first hour, the team confirms the loss is real by cross-checking positions and prices against an independent source, ruling out a booking error. Within the next two hours, they reduce the book's largest, most correlated exposures by roughly half — not to zero, since a panicked full unwind into an already-stressed market can realise losses that a calmer, partial reduction would have avoided, but enough to cap further downside while the cause is investigated. Risk management and senior leadership are briefed the same day, with an explicit statement of what is known (several correlated positions moved together in a way the historical correlation structure didn't anticipate) versus what remains unclear (whether this reflects a temporary liquidity event or a genuine, lasting change in how those positions relate to each other). No final decision about the strategy's future is made on day one; that decision follows a proper diagnosis over the following days and weeks, informed by the kind of structured cause analysis covered in Diagnosing a Slump: Decay, Crowding or Costs.

In practice

• Have the playbook written before you need it. Deciding, under stress, who to notify and in what order is exactly the kind of decision that should already be a checklist, not improvised.
• Resist the urge to fully explain the loss on day one. An early, confident explanation is often revised once more information arrives, and communicating a wrong explanation loudly can be worse than admitting the cause isn't yet clear.
• Reduce exposure deliberately, not reflexively. Both "do nothing and hope it reverts" and "immediately liquidate everything" are common overreactions in opposite directions; a proportionate, partial reduction usually serves better than either extreme.
• Treat this as distinct from ordinary decay monitoring. A blowup is a sudden, discrete event calling for immediate containment, while the gradual monitoring covered in Detecting Decay in a Live Strategy and the deliberate exit triggers in Defining Kill Criteria Before Launch are for slower-moving situations — recognising which kind of event you're in is itself part of the first hour's job.
• Schedule the full post-mortem for after the immediate crisis, not during it. A rushed, incomplete investigation done under the same pressure as the initial response tends to miss the real cause.