Constructing a Market — MacKenzie & Millo on Performativity

Constructing a Market — MacKenzie and Millo on Performativity and Legitimacy in Economics

Derivatives, marijuana and thalidomide

“Financial weapons of mass destruction”. That was the affectionate term that Warren Buffett gave to the market for derivatives. In his 2002 letter to Berkshire Hathaway shareholders, he added that “the derivatives genie is now well out of the bottle, and these instruments will almost certainly multiply in variety and number until some event makes their toxicity clear.”

He was right. Between 2005 and 2008, the total value of OTC derivatives more than doubled, reaching a peak “just shy of $35 trillion”—just in time for the collapse of Lehman Brothers and the eventual financial implosion. A decade on, options and derivatives, in all shapes and sizes, continue to reach far and wide in finance.

But it wasn’t always like this, as Donald MacKenzie and Yuval Millo explain in Constructing a Market. True, options and futures have long been a feature of financial markets. But the 1929 stock market crash and the Great Depression had made people wary of anything that looked remotely speculative.

“Even as late as the 1960s, market regulators such as the Securities and Exchange Commission remained deeply suspicious of derivatives”, write MacKenzie and Millo. Then SEC Chair Manuel Cohen had likened options to “marijuana and thalidomide”. (Thalidomide was a popular drug to treat nausea in pregnant women during the 1950s. However, “it became apparent in the 1960s that thalidomide treatment resulted in severe birth defects in thousands of children.”)

Cohen was not exaggerating. Back then, options trading relied on qualitative instincts and arbitrary heuristics. The Babylonian kings who shook arrows to invoke divination would not find themselves out of place in the early days of options trading.

The Black-Merton-Scholes model

But headway was made in 1973, when Fischer Black, Robert Merton, and Myron Scholes developed their theory on options pricing. Their reasoning was relatively straightforward once the building blocks came together.

As MacKenzie and Millo explain, if we assume amongst other things that stock prices follow “the standard model of a lognormal random walk in continuous time”, then it is “possible to construct a continuously adjusted portfolio of the underlying stock and government bonds or cash that would replicate the option.” It follows that “the price of the option must equal the cost of the replicating portfolio.” If this was not the case, then “arbitrageurs would buy the cheaper and short sell the dearer. Their activities in turn would push asset prices efficiently to their correct levels.”

Here, “what was being bought and sold in an options market was reconceptualized”, Mackenzie and Miller add. “It was the Black-Scholes-Merton model’s key parameter, volatility. If stock volatility increased, options became more valuable; if it decreased, they became cheaper.”

Unveiling the black box of finance

And just like that, the Black-Merton-Scholes (BSM) model had lifted the veil on the black box that was options trading. As Burton Rissman observes in his interview with MacKenzie and Millo, BSM “gave a lot of legitimacy to the whole notions of hedging and efficient pricing.” It was no longer “speculation or gambling, it was efficient pricing.”

But the formula was not an immediate empirical success. Early testing of the option pricing formula in 1972 “found only approximate agreement.” At the time, the price of call options according to the BSM model “were maybe 30-40 percent overvalued.” Either the model was wrong, or a lot of money stood to be made.

But “by 1986, typical deviations [between observation and theory] had fallen to less than 1%” for index options that traded on the CBOE. The theory’s fit had improved. And many hailed it as a triumph for economics and finance. Merton and Scholes would go on to win the 1997 Nobel Prize for their contributions.

Computation and culture

It’s worth noting, however, that the empirical success of option pricing theory is nuanced. For starters, the assumptions were initially “wildly unrealistic”. Portfolio management at the time was neither as cost-effective nor instantaneous as the theory demanded. But as computing power, data feeds and trading costs improved, “the model’s assumptions… became more accurate.”

Behaviourally, there was cultural resistance and machismo to overcome as well. Traders who brought theoretical price sheets to the trading floor were often mocked and scorned. Barbs like “you’re not a man if you’re using those theoretical value sheets” were commonplace.

Of course, “seat-of-the-pants trading” could not compete with more sophisticated approaches over the long run. After all, who are people more likely to trust? The trader with a clairvoyant stomach, or the quants and their Nobel-winning formulas? As finance and risk management grew manifold, “expensive errors made pricing models seem indispensable.”

Performativity of economics

Some have taken the success of option pricing theory as evidence in favour of market efficiency. MacKenzie and Millo suggest instead, however, that the rise in legitimacy and popularity of pricing models gave rise to the very prices predicted by BSM.

Think about it. If more and more traders begin to use the same theories, assumptions, and models, then the market as a whole will think about valuations and arbitrage in the same way. Prices were converging on the models that traders were using.

The result hints at a more general phenomena of “performativity” in economics. That is to say that economics is more than a descriptive theory. It creates and shapes the systems of which it is apart. Its parallel in philosophy is the performative utterance—a statement that changes the very reality of the things it refers to.

Indeed, financial markets persist in part because investors and lenders share similar beliefs about the future. Such expectations in turn produce the financing and real business activity that is necessary to realize such beliefs. So it should not surprise us to hear about theories that posits a value or future that it helps in turn to create.

That’s not to say, however, that economic theories are always self-fulfilling. While the empirical fit of the Black-Scholes-Merton model was strong for some time, much of it changed after Black Monday in 1987. Following the one-day 20 percent plunge across major indices, the market abandoned the log-normal assumption and “has come to expect crashes.” A “volatility skew” emerged in option prices as investors updated their models accordingly—no longer conforming to the predictions of the original BSM.

Performativity, and reflexivity

So we can categorize the evolution of option pricing, MacKenzie and Millo note, into “three distinct phases”. In the beginning, “Black, Scholes, and Merton did not describe an already existing world.” But over time, “financial markets changed in a way that fitted the model.” This seemed true at least until the “collective trauma” of Black Monday in 1987, which led investors to update their models.

Financial history reveals to us the complex relationship between beliefs, expectations, and reality. Options pricing in particular is a microcosm of the performativity that pervades economics. It seems to me then that Merton and Scholes had won their Nobel prize in part for developing a theory that contributed to the very results they were predicting. Beliefs shape actions, while actions shape beliefs. They can take a life of their own. Indeed, as George Soros noted long ago, it would be unwise to ignore the reflexivity and fallibility that is inherent to economic theory and financial markets. To that, MacKenzie and Millo would add performativity.

Sources and further reading

MacKenzie, Donald., & Millo, Yuval. (2003). Constructing a Market, Performing Theory: The Historical Sociology of a Financial Derivatives Exchange.
Buffett, Warren. (2002). Annual Letter to Berkshire Hathaway Shareholders.
Soros, George. (2009). General Theory of Reflexivity. The Soros Lectures at the Central European University. Open Society Foundations.