This is a rare work of serious statistics made accessible to general readers who want to know the interrelationship of risk (quantifiable chance or outcomes), and uncertainty (what is unknown and unquantifiable). Authors Dembo, a serious academic who writes the rocket science programs used by large banks and brokerages to structure and trade risk-weighted financial instruments, and Freeman, a writer for The Economist, have produced a work with a huge potential audience.

Seeing Tomorrow brings regret to risk assessment. If a decision turns out to be less than optimal for a given time horizon, then one regrets it. Selling a stock one day, only to see it rise the next, produces regret. If the seller captures the best price for a time interval, there is no regret.

The regret reference point is a triviality until it is quantified in a trading model. The authors provide measures for equations that compare the regret-adjusted trading positions for two parties doing an exchange. The equations adjust for the phenomenon that each side may substitute intuition or some acquired subjective knowledge, such as “X is a smart guy whose trades pay off handsomely,” for known baselines, such as trades that come out losers 95% of the time. The result is a uniform risk/uncertainty model that allows people to trade across different asset categories, evaluating in one equation life insurance, stocks, bonds, and cash.

Dembo and Freeman present models with some important properties. Their equations can be manipulated with a little algebra. They are what mathematicians say is continuously differentiable, rather than on/off processes of game theory, a branch of math that can produce similar answers through different but non-continuous models. Readers who remember a bit of algebra should have no problem getting through Seeing Tomorrow; those who grasp its main concepts will have an understanding of what could turn out to be a cult of cognoscenti.