↓
“There exist around fifty major commodity exchanges that trade in more than ninety commodities. Trading on exchanges is however concentrated. In 2009, the top five exchanges accounted for 86% of all contracts traded globally (TheCityUK 2011). Soft commodities are traded around the world and dominate exchange trading in Asia and Latin America. Metals are predominantly traded in London, New York, Chicago and Shanghai while energy related contracts are predominantly traded in New York, London, Tokyo and the Middle East (TheCityUK 2011). In terms of future contracts traded in 2009, China and the UK accounted for three out of the top ten exchanges while the United States accounted for two and Japan and India for one. China and India have gained in importance in recent years with their emergence as significant commodity consumers and producers. London, New York and Chicago remain however the main centers of commodity future trading.”

↓
“Indeed, focusing on the interface between arbitrageurs and noise traders, De Long et al. (1990) analyse the process by which excess volatility is generated by noise traders. They suggest that the unpredictability of noise traders’ beliefs and expectations, which can be erroneous in the light of fundamentals, could create a ‘noise trader risk’ – a risk in the asset prices, which deters rational arbitrageurs from aggressively betting against them. Hence, ‘arbitrage does not eliminate the effects of noise because noise itself creates risk’ (De Long et al., 1990: 705), since arbitrageurs are likely to be risk-averse, acting with a short time-horizon. As a result, ‘prices can diverge significantly from fundamental values even in the absence of fundamental risk’ (De Long et al., 1990: 705). Moreover, bearing a disproportionate amount of risk enables noise traders to earn a higher return than rational investors who engage in arbitrage against noise. Clearly, their model challenges the standard proposition made by Friedman (1953) that irrational noise traders are always counteracted by rational arbitrageurs who could drive asset prices close to fundamental values.”

↓
“Karl Pearson, the first Professor of Statistics in Britain, wrote two quite different histories of correlation and regression analysis separated by some twenty-five years. The first in 1895 attributed correlation to the French mathematician Auguste Bravais, a mid-19th century error theorist, interested in assessing the accuracy of astronomical measurements. For any given point in space, a separate set of measurements could be made for both the x and y coordinates. The overall pattern of error is then given by multiplying together the two independent laws of error associated with each coordinate. Bravais, however, was interested in the more difficult case where the same set of measurements were used to calculate both x and y together. To do that required him to calculate a joint law of error, the equation of which is remarkably similar to Galton’s later equation for correlation.
[…] By 1920, however, Pearson…had changed his mind, arguing that it was Galton who was the true originator of correlation because the problem to which Bravais applied his work was completely different from the one Galton was trying to solve. Bravais devised his equation…in order to get rid of the amount of statistical variation of error around the true values of the variables, whereas for Galton it was precisely the statistical variation—the error—that needed to be kept. Once explained, the variation could be made the source of intellectual progress: specifically for Galton’s purposes [as a eugenicist,] geniuses could be bred.
[…] Galton is interested primarily in the deviations around the mean, and not the mean itself. As Galton himself says, “The primary objects of the Gaussian Law of Error were exactly opposed, in one sense, to those which I applied them. They were to get rid of, or to provide a just allowance for errors. But those errors or deviations were the very things I wanted to preserve and know about.” (1908, p. 305, note 2).”

↓
“Cartwright (1999, ch. 7) argues that probabilities are not there for the taking, but are characteristics of quite particular set-ups (e.g., of roulette tables or particular configurations of unstable atoms). Only in such designed set-ups do objects display well behaved probabilities. The role of economic theory (and of quantum-mechanical theory) is to provide the conditions that articulate such a well-defined set-up: a nomological (or law-generating) machine.”

↓
“In its attempts to attain its many objectives, economic theory was helped by greater abstraction – preference theory supplies an example again. Significant research efforts were expended on solutions of the integrability problem. That problem can be bypassed altogether, and greater simplicity can be achieved by moving from the commodity space to the more abstract space of the pairs of its points. In this space, whose dimension is twice the number of commodities, the pairs of commodity points indifferent to each other are now assumed to form a smooth (hyper)surface. As another instance of the generality permitted by abstraction, consider the notion of a commodity, which can be treated as a primitive concept, with an unspecified interpretation, in an axiomatic economic theory. A newly discovered interpretation can then increase considerably the range of applicability of the theory without requiring any change in its structure. Thus, by making the transfer of a good or service between two agents contingent on the state of the world that will obtain, Arrow (1953) made possible the immediate extension of the economic theory of certainty to an economic theory of uncertainty by a simple reinterpretation of the concept of a commodity. The theory of financial markets has been influenced by that view of uncertainty, and their practice has not been unaffected. Finally, take the problem of existence of a general equilibrium, once considered to be one of the most abstract questions of economic theory. The solutions that were proposed in the early 1950’s paved the way for the algorithms for the computation of equilibria of Herbert E. Scarf (1973) and for several of the developments of applied general equilibrium analysis (Scarf and John B. Shoven, 1984).”