Fischer Black

Fischer Black was an American economist, born in Washington DC in 1938. His training was not in economics: he took a first degree in physics at Harvard and a doctorate there in applied mathematics in 1964. He moved into finance through consulting work at Arthur D. Little, where Jack Treynor introduced him to the subject.

He taught at the University of Chicago from 1971 and at the Massachusetts Institute of Technology from 1975. With Myron Scholes he wrote The Pricing of Options and Corporate Liabilities, published in the Journal of Political Economy in 1973 after being rejected elsewhere. The paper derives what an option should cost from the observation that a position in the option can be offset continuously by a position in the underlying share, so that the combination carries no risk and must therefore earn the risk-free rate. Robert Merton published a related paper the same year giving an alternative derivation and generalising the result, and it was Merton who attached the name Black-Scholes to it. The Chicago Board Options Exchange opened for trading in the same year.

In 1984 he did something unusual for a leading academic and joined Goldman Sachs, where he stayed for the rest of his career. With Robert Litterman he developed the Black-Litterman model, an approach to portfolio allocation that starts from the returns implied by market weights and then adjusts them for an investor's specific views, which addressed the tendency of earlier optimisers to produce extreme allocations from small changes in inputs.

His 1986 presidential address to the American Finance Association, published as Noise, is the paper investors should read. In it he argues that trading on information that is not really information is what supplies liquidity and simultaneously prevents prices from being precise. He offered a definition of an efficient market that is startling coming from a founder of modern finance: one in which price is within a factor of two of value, at least most of the time. He passed away in 1995, aged 57.

The option pricing argument rests on a single idea that is easy to state and was not obvious before. If you hold an option and simultaneously hold an offsetting position in the underlying share, adjusted continuously as the price moves, the combination has no risk. Anything with no risk must earn the risk-free rate, or there is an arbitrage. That constraint alone is enough to determine what the option must cost, and notably it does not require knowing whether the share will rise or fall. The expected return of the underlying drops out of the formula entirely, which is the part that surprised people.

What makes Black interesting rather than merely important is that he spent the second half of his career explaining the limits of that kind of precision. Noise argues that markets need participants who trade on things that are not really information, because without them there would be nobody to take the other side and no liquidity. The same noise that makes trading possible is what stops prices from being accurate, so liquidity and precision are in tension rather than being two names for a healthy market.

His definition of efficiency follows from that and is unusually humble. Rather than treating price as equal to value, he suggested that a market is efficient if price sits within a factor of two of value most of the time, and observed that by this standard essentially all markets are efficient essentially always. The point is not cynicism; it is that the standard being defended in the literature was far stronger than the evidence supported.

The Black-Litterman work applies the same scepticism to portfolio construction. Earlier optimisers took expected returns as inputs and produced wildly concentrated allocations, because small errors in those inputs dominated the result. His approach starts instead from the returns implied by what the market already holds and adjusts for specific views, which keeps the output stable when the inputs are uncertain, as they always are.

Option pricing became infrastructure rather than a theory. Derivatives markets, employee share options, convertible securities and the practice of quoting implied volatility all depend on the framework this paper established.

Noise had a different kind of influence. It gave later behavioural and market-microstructure researchers a respectable footing for arguing that prices are only approximately right, made by someone whose credentials in the efficient-market tradition were beyond question.

The Black-Litterman model is standard in institutional allocation, largely because it produces portfolios that remain sensible when the return estimates feeding it are uncertain.