Miller focuses on several well-known derivatives disasters—Procter & Gamble, Metallgesellschaft AG (MG), Orange County, and Barings—and evaluates the related social costs. First, he notes the importance of distinguishing between the private loss of a counterparty and the aggregate social loss. Derivatives, by their very design, have two sides that always exactly offset each other. If one party wins, then, by design, the other party loses. Hence, wealth transfer takes place between counterparties, but no aggregate loss of social wealth is associated with the transaction.

Similarly, a bond market crash also involves only wealth transfers between creditors and issuers. For example, the losses incurred by Orange County, California, resulted in gains to the issuers, such as the Home Loan Bank Board, the Federal National Mortgage Administration, and the U.S. Treasury. Effectively, the taxpayers of Orange County transferred wealth to U.S. taxpayers. The aggregate social cost of such a transfer was negligible.

MG's losses are more complicated. MG was hedging its long-term fixed-price forward commitments. Unfortunately, futures contracts are marked to market daily but forward commitments are not. Thus, theoretically, MG's hedging program should have offset its risks. When the margin calls reached $1.3 billion, MG liquidated its position. Even in this case, the losses incurred by MG resulted in gains for those futures traders who were on the opposite side of the transaction. Thus, in aggregate, very little of the world's wealth was lost from these two dramatic derivatives debacles.

Miller turns next to the question of whether one should be concerned with large wealth transfers. An uncompensated wealth transfer (e.g., robbery) is cause for concern because someone has been unjustly harmed. Derivatives, however, are not robbery but, rather, a voluntary transaction that must be mutually beneficial. In the end, derivatives transactions are zero-sum transactions, in terms of dollars trading hands, or are slightly negative, once trading costs are considered. Before the transaction, the derivatives deal must have been a positive-sum transaction for both parties; otherwise the parties would not have agreed to the deal.

Miller considers three types of social costs in order to assess the indirect social costs related to derivatives disasters: litigation costs, bankruptcy costs, and systemic risk. Although MG is involved in at least six major lawsuits and numerous other suits related to derivatives disasters are on the books, historically, the amount of litigation related to derivatives activities is low. This finding is surprising given that every derivatives trade, by definition, involves a loser and hence a potential plaintiff. Ultimately, the cost of litigation will be borne by the customers in this market. Miller argues that private litigation seems to be the most efficient form of redress for claims of fraud when compared with the alternatives of regulatory intervention.

Actual bankruptcy costs are an indirect social cost related to derivatives disasters. The loss of real social capital in bankruptcy is low, apart from the legal and negotiation costs. When a business is liquidated, the physical assets do not simply disappear: They are often used in other productive capacities.

The social costs related to systemic risk—the risk that failure of one major bank may bring down the world's financial systems—are real but extremely remote; banks are heavily capitalized and highly diversified, and they are constantly monitoring their aggregate risk exposures.

Miller concludes that the recent increase in the volume of derivatives traded suggests that market participants view this market as having integrity and that further regulation is unnecessary.