CHAPTER 11 · When Bad Things Happen to Rich People
A commitment survives the market mood
Trace underwriting exposure through a market shock and evaluate an exit without hindsight.
Michael Lewis, Liar’s Poker · Chapter 11. Original paraphrase and analysis; not the book text.
The idea beneath the story
The final chapter brings financing commitments, corporate setbacks and the 1987 market crash into the same frame. Lewis describes Salomon’s exposure to transactions whose economics change as markets deteriorate, including bridge financing and the BP share sale. The issue is not simply that prices fall. A firm may already have promised to supply capital or buy securities, and that promise can remain binding after the market becomes much less welcoming.
An underwriter can earn a fee for helping an issuer raise money. Depending on the agreement, it may also commit its own balance sheet. If demand weakens before distribution, the underwriter can be left holding securities at an unattractive acquisition cost. A commitment that looked like a short interval between purchase and resale can become a material position. The exact obligation depends on the contract; the word underwriting alone does not specify every risk.
Hypothetical worked example: an underwriter commits to buy one million shares at £5 each and expects to distribute them at £5.10. The expected gross spread is £100,000 before costs. If the market price falls to £4 before sale and the commitment still binds, selling the shares produces a £1 million loss against the £5 million purchase cost. A small expected fee was attached to a much larger principal exposure. This is a teaching example, not the numerical history of the BP transaction.
The stock-market connection is the distinction between fee income and contingent balance-sheet risk. A financial firm may describe a business as advisory or distribution-led while some contracts still expose it to unsold inventory or financing obligations. Read commitments alongside current holdings. What is not yet on the balance sheet may matter when a stress scenario turns an expected sale into a retained asset. Off-balance-sheet wording does not mean outside economic risk.
Lewis’s account also shows how a crisis can redirect attention toward blame, loyalty and personal rewards. Those responses are understandable but not necessarily useful for reducing exposure. A practical incident review begins with obligations, cash needs and available actions. Who can stop new risk? What has already been promised? Which markets are still executable? Who needs accurate information now? Establishing the sequence is more useful than selecting a villain before the facts are assembled.
Knight Capital is a later equity-market comparison about exposure accumulating faster than controls can stop it. A software deployment failure differs from a market crash and an underwriting commitment. The common analytical question is how a firm knows its aggregate position, how quickly the position can change and what authority can limit the damage. The comparison does not imply that a manual 1980s trading floor and a modern automated router should use identical controls.
The epilogue adds a different kind of exit. Lewis explains leaving the firm to tell the story rather than continuing the career. Do not rewrite that decision as a perfect market-timing signal. Leaving a job can reflect a person’s interests and values rather than a prediction of imminent collapse. Likewise, an investor’s sale can be appropriate because the position no longer fits the mandate or financing capacity, even if the asset later rises. Evaluate the information and constraints at the time.
The book closes a useful circle. At the beginning, status tempts people to take a spectacular wager; at the end, obligations and incentives reveal the cost of commitments. For a stock-market reader, the lasting practice is to inspect the full structure behind a confident story: cash flows, capital, contracts, counterparties and the route out. You do not need to predict every crash to avoid assuming that yesterday’s ability to sell will always be available tomorrow.
Knight Capital: when orders outrun controls
Knight is revisited to examine the speed of exposure accumulation and the authority to stop it. The historical mechanism differs from underwriting during the 1987 crash. This case is shared with related chapters.
Documented sequence. On 1 August 2012, a faulty code deployment activated defective functionality in Knight Capital’s equity order router. The SEC described millions of erroneous orders and a loss exceeding $460 million. It found weaknesses in deployment, exposure controls and incident procedures. In 2013, Knight agreed to a $12 million settlement without admitting or denying the findings. The episode concerns operational exposure in equity markets, rather than an investment forecast that happened to be wrong.
SEC · Knight Capital market-access enforcement, 2013 · Checked 10 October 2026
Original analysis: a chain, not a typo. A software error becomes a financial disaster when it reaches a live market, generates exposure and continues long enough to overwhelm the firm’s capacity. A useful explanation therefore includes deployment, detection, authority and containment. Focusing only on the defective code leaves unanswered why the resulting orders were allowed to accumulate and why the response did not stop the financial consequence sooner.
Think about the difference between a message and an effective alert. A message is information emitted by a system. An effective alert has an owner, a meaning, a deadline and an action. If the recipient cannot tell whether it indicates a harmless anomaly or an accumulating position, more messages may simply increase noise. This is an original design inference, not a claim that the precise communication choices of every employee are known from the source.
Hypothetical worked example. A system mistakenly buys £100,000 of stock each second. In one minute it has acquired £6 million; in ten minutes, £60 million, assuming the orders execute at those amounts. A person who takes five minutes to understand the issue may be conscientious and still far too slow for the system’s risk-creation rate. A pre-trade exposure cap or automatic stop can operate on a different timescale. Real controls also need to handle false positives and safe recovery.
The shareholder’s question is whether the business can create obligations faster than its governance can observe them. This applies beyond high-frequency trading. Credit approvals, payment systems and automated pricing can all scale a mistake. Growth in throughput is not automatically growth in resilience. If management celebrates transaction speed, ask how quickly it can detect a discrepancy and how much capital could be consumed before an authorised intervention.
The connection to Lewis’s training culture is specific: policies and expertise require an operating environment that turns them into action. The connection to his final chapter is different: commitments can accumulate before people fully understand their economic effect. These comparisons do not make software failure equivalent to an underwriting loss. They supply two separate questions about a documented later event.
A common misunderstanding is that more technology automatically means less human risk. Automation removes some manual errors and introduces new dependencies. Another is that one employee’s mistake fully explains the event. Individual mistakes are part of the chain; robust systems assume that mistakes can occur and limit their reach. The relevant standard is not perfection but bounded consequences.
Takeaway. A risk limit should be able to prevent or contain the unwanted position, not merely report it after the fact. Ask what is measured, where the measurement occurs, who can halt activity and how the system is restored safely. A successful test should include an abnormal path, because the ordinary path is precisely where a fragile design can look most convincing.
Workshop: review the trade before the commitment
A transaction has more than one price. There may be a last traded price, an indicative quote, a firm executable bid, an offer and a model estimate. They can differ without contradiction because they refer to different sides, sizes or conditions. A useful trade review labels each one. If someone says a position is worth £1 million, ask whether that is a model value or the cash an identified buyer would pay for the entire position now. The answer determines which risks remain between the estimate and the realised result.
Size changes execution. A screen may show an attractive price for a small quantity while a larger order must trade through several price levels. A hypothetical buyer wants 1,000 shares. There are 200 offered at £10 and 800 at £10.20. Buying the full amount costs £10,160, or £10.16 per share, before fees. The best displayed offer of £10 was not the average acquisition price. Real order books can change during execution, so even this arithmetic assumes the displayed quantities remain available while the order is filled.
A limit order and a market order solve different problems. A limit can constrain the price but may leave the order unfilled. An instruction prioritising immediate execution can expose the buyer to price uncertainty. Neither eliminates the need to understand liquidity. This is a conceptual distinction, not a current broker-specific instruction or a recommendation about a particular order. The trade review should state which uncertainty the participant is accepting: price, completion, time or some combination. An apparently cheap opportunity is incomplete if its assumed execution cannot occur.
Now trace the seller’s role. A dealer can sell inventory as principal, arrange a trade as an intermediary or provide analysis under another arrangement. These roles carry different economic incentives. Ask how the participant is paid and what it retains after the trade. A disclosed spread can compensate inventory risk and service. An undisclosed material conflict can distort the buyer’s understanding. Do not collapse these into the claim that all compensation is improper; the purpose is to identify what service is being purchased and what information is needed to evaluate it.
The commitment date may precede the cash movement. An underwriter can become obligated before final distribution; a lender can promise funds before a borrower draws them. A balance sheet showing current holdings may therefore understate the exposure that appears if a planned sale or refinancing fails. Build a commitment register containing amounts, dates, conditions and the party that must act. This is especially useful for a shareholder in a financial intermediary, because fee-generating activities can carry contingent principal risk that is not obvious from the revenue line.
Consider a hypothetical bridge loan of £50 million intended to be sold to investors after an acquisition. If distribution succeeds promptly, the lender may receive fees while releasing capital. If investors demand a lower price, the lender must choose among selling at a loss, retaining the loan or renegotiating where possible. Each choice has costs and contractual limits. The phrase bridge loan describes an intended transition, not a guarantee that the bridge will have an open exit. A stress test should specify who funds the position if the expected buyers disappear.
Execution risk also includes operations. Confirming a trade, settling it, valuing collateral and reconciling positions are parts of the economic activity, not clerical decoration. A mismatch between internal records and actual obligations can produce wrong decisions even if the original investment view was sound. The right control depends on speed and scale. A daily reconciliation may help a slow-moving portfolio and be insufficient for a system that creates large exposures in seconds. Control frequency should be related to the rate at which harm can accumulate.
Plan escalation before the stressful event. Name the role that can halt new activity, the information required and the route for resolving uncertainty. A vague instruction to contact management can fail when several managers assume someone else is responsible. A practical plan also distinguishes stopping new risk from safely managing existing commitments. Switching a system off may prevent further orders without closing the positions already acquired. The incident response must account for both, including the risk that a hurried correction creates a second error.
After the event, compare the decision with the information available at the time. A profitable trade does not prove that the process was sound if it depended on an unrecognised exposure. A loss does not prove negligence if the risk was understood, bounded and consistent with the objective. The review should nevertheless search for errors rather than use uncertainty as a blanket excuse. Record what was expected, what occurred, where the mechanism differed and which change would reduce a repeat. Avoid explanations that name only market conditions while ignoring controllable choices.
For a stock-market reader, the final output is an execution-aware thesis. It describes the business or security, the price basis, the relevant liquidity and the obligations created by participation. It also recognises that a decision not to trade can be rational when the evidence or mechanics are unclear. This workshop does not promise a safe order type or a profitable transaction. It connects a persuasive idea to the actual process by which cash, ownership and risk change hands. That connection is where many apparently complete investment stories reveal their missing assumptions.
A final hypothetical review concerns the exit assumption in a valuation. An investor estimates a holding at £500,000 using the last small trade, but the only firm bid for the full position is £450,000. Neither number should be silently substituted for the other. The first may serve as a reference mark; the second answers a particular liquidation question at a particular time. If the investor owes £470,000 tomorrow, the distinction is immediately material. If there is no forced sale, the lower bid still provides information about current liquidity without proving the asset’s ultimate value. State the purpose of the number and the time over which it must be realised. A decision becomes more reviewable when each amount is attached to an executable action or an explicitly identified estimate.
Decision exercise (hypothetical)
A small underwriting fee is attached to a large binding purchase commitment. What belongs in the risk review?
- Only the expected fee
The fee does not measure the principal at risk.
- Only last year’s successful deals
Past completion does not remove the present obligation.
- The unsold position, funding needs and contractual exit terms
Yes. Trace the obligation through a failed distribution scenario.
Recall and self-check
Does a planned quick resale remove inventory risk?
No. The buyer or market may disappear before resale.
Does a later price rise prove an earlier exit was wrong?
No. Evaluate the information, mandate and financing constraints at the time.
Reflection
What would you require in a one-page summary of all outstanding commitments?
Include amounts, deadlines, counterparties, conditions, cash demands and named decision authority.
Open interactive chapter & notes