Chapter 37: Derivatives Mishaps and What We Can Learn from Them

Why Study Mishaps?

"Those who cannot remember the past are condemned to repeat it." — George Santayana

Studying derivatives failures reveals patterns of risk management breakdown that transcend specific instruments or market conditions. The same mistakes recur because human and organizational behaviors repeat.

Lessons for All Users of Derivatives

1. Define Risk Limits Clearly

Many losses occurred because traders exceeded limits without detection. Firms must:

• Set unambiguous position limits
• Monitor limits in real-time (not just at day-end)
• Escalate and investigate limit breaches immediately

Case: Barings Bank (1995) — Nick Leeson was both the head trader and the back-office supervisor in Singapore. With no separation of duties, he hid £827 million in losses from Nikkei 225 futures trading in an error account. The bank collapsed when the losses were discovered.

2. Take Risk Limits Seriously

Do not assume a trader who exceeds limits can "trade out" of the position. Limit violations are a red flag, not a negotiation starting point.

Case: Allied Irish Banks (2002) — John Rusnak accumulated $691 million in FX options losses over five years by fabricating trades and exceeding limits. Management ignored early warning signs.

3. Don't Assume You Can Outguess the Market

Trading strategies that assume you know better than the market (directional bets) are dangerous. Hedging strategies are different from speculative strategies — confusing the two leads to trouble.

Case: Metallgesellschaft (1993) — The company used a stack-and-roll hedging strategy with short-dated oil futures to hedge long-term delivery contracts. When oil prices fell, massive margin calls ($1.3 billion) created a liquidity crisis. The strategy was arguably correct in the long run but the company couldn't survive the short-term cash flow strain.

4. Beware When Everybody Is Following the Same Strategy

Crowded trades create liquidity risk — when everyone tries to exit simultaneously, prices gap. The 1987 crash, the 1998 LTCM crisis, and the 2008 financial crisis all featured crowded trades unwinding.

Case: Long-Term Capital Management (1998) — LTCM's convergence trades (e.g., long off-the-run Treasuries, short on-the-run) were based on solid principles. But when Russia defaulted in August 1998, a flight to quality caused spreads to widen rather than converge. LTCM was leveraged 25:1 and lost $4.6 billion in months. The Fed orchestrated a bailout to prevent systemic risk.

5. Financial Markets Are More Volatile Than "Normal" Distributions Suggest

Models that assume normal distributions underestimate tail risk. The 1987 crash was a 20+ sigma event under the normal distribution — effectively impossible. Markets experience fat tails, and risk models must account for this.

6. Don't Use "Black Box" Models Without Understanding Them

Traders and managers should understand the assumptions, limitations, and failure modes of the models they use. The Gaussian copula model for CDOs was widely used with an assumed correlation of 0.15-0.30. When defaults clustered during the 2007-2008 crisis, realized correlations were far higher, causing massive losses.

7. Liquidity Risk Can Be Fatal

Models typically assume continuous, liquid markets. In a crisis, markets freeze:

• Bid-ask spreads widen dramatically
• Some securities become untradeable
• Margin requirements increase
• Funding sources dry up

The interaction of funding liquidity and market liquidity creates a "death spiral."

Lessons for Financial Institutions

Segregate Responsibilities

• Front office (trading), middle office (risk management), and back office (settlement) must be independent
• No trader should control or influence the P&L verification process
• Risk management must have authority to stop trading activity

Monitor Traders' Behavior

• Unusually high profits with low reported risk are suspicious (Leeson, Rusnak, Kerviel all showed this pattern)
• Traders who never take vacation may be hiding something
• Lifestyle changes inconsistent with known compensation warrant investigation

Understand the Products

• Senior management at many firms that suffered derivatives losses didn't understand the instruments
• The complexity of modern derivatives demands specialized expertise in risk management and audit functions

Case: Société Générale (2008) — Jérôme Kerviel, a junior trader, took unauthorized directional positions worth €50 billion (more than the bank's market cap). He used his knowledge of back-office systems from his previous role to circumvent controls. Loss: €4.9 billion.

Case: UBS (2011) — Kweku Adoboli lost $2.3 billion on unauthorized ETF trades, hiding the risk by booking fake hedging trades. The bank had failed to upgrade its systems despite warnings.

Diversification Benefits Can Disappear in a Crisis

Diversification that works in normal markets often fails during crises. In the 2008 crisis:

• All risky assets fell simultaneously
• Correlations spiked toward 1
• Only the safest assets (Treasuries, gold) provided refuge

Lessons for Nonfinancial Corporations

Make Sure You Fully Understand the Trades

Companies should not enter derivatives trades they don't understand. If you can't explain the trade simply, don't do it. The board of directors should approve all material derivatives transactions.

Case: Procter & Gamble (1994) — P&G lost $157 million on a complex leveraged swap with Bankers Trust. The swap's payoff formula involved a complicated function of interest rates. P&G argued they were misled about the risks; Bankers Trust was found to have used misleading sales practices.

Case: Orange County (1994) — The county treasurer invested in structured notes and inverse floaters that performed well as long as rates stayed low. When the Fed raised rates in 1994, the portfolio lost $1.6 billion. Orange County declared bankruptcy — the largest municipal bankruptcy in US history at the time.

Don't Make Treasury a Profit Center

Corporate treasury departments exist to manage risk, not to generate profits. When treasuries are evaluated on their trading profits, they take risks that can destroy the company.

Don't Assume Financial Engineering Can Convert Losses to Profits

Some companies have been sold complex derivatives under the premise that they can reduce borrowing costs with no additional risk. In reality, these structures typically embed a hidden bet on rates, currencies, or other variables. The apparent cost reduction is compensation for risk-bearing.

Summary of Major Losses

The common threads: inadequate oversight, concentrated risk, misunderstanding of instruments, overreliance on models, and the toxic combination of leverage and illiquidity.