The Nature of Markets

Abstract

Time is the ultimate constraint on mortals. Lack of time contributes also to limited information. Shortage of time and information make it harder for agents to achieve Pareto equilibrium matchings of wants with endowments. This had led, as described in Section 1, to the evolution of several social artifacts to relax such constraints, namely the establishment of both organised markets and money.If markets are only held at infrequent discrete occasions, potential buyers and sellers lose the advantage of immediacy, a term coined by Demsetz. On the other hand, the flow of new orders into a continuously functioning market may be small and relatively unbalanced in any short period. One common response to this problem has been for certain market participants to assume the function of market making, acting as principals and quoting prices at which they will buy or sell the good or asset being traded.There are considerable resource costs involved in establishing a market. In practice, many (perhaps most) agents who act as market makers — e.g., retailers — also provide associated functions. Nevertheless, one can abstract from the costs arising from such associated functions and enquire as to the nature of the particular costs involved in market making itself, and how those costs will be reflected in the size of the margin, or spread, between the bid and ask price. This forms the subject of Section 2. There are two such main costs. The first involves resource costs, including the costs of carrying inventory necessary to fulfil the role of market maker. The second follows from the fact that the market maker’s quoted prices make him the potential victim of anyone with ‘inside information’ whether the likely future price is liable to lie outside the quoted price spread.Despite the fact that new order flows themselves carry information about future prices, one can observe that market makers do not always adjust prices instantaneously. There is a whole spectrum of speeds of price adjustment in differing markets, ranging from very fast indeed in standardised asset markets — e.g., the foreign exchange market — to comparatively very slow in labour markets. In Section 3 we ask what determines the speed with which market makers adjust prices. This remains a conceptually difficult subject. We conclude that the main considerations are: 1. The costs/risks involved in absorbing inventory fluctuations in each market. 2. The extent and distribution of information about the asset, good or service being traded in each market. 3. The relative risk aversions, and hence preference for insurance, among traders and market makers. There are economies of scale in the establishment of markets. Moreover, as noted in Section 2, the higher the ratio of regular traders to those with ‘inside information’about future price levels, the lower will be the spread that the market maker can quote; these low spreads induce more trading and therefore support the continued existence of markets. There are thus centripetal tendencies towards the centralisation of trading through successful markets, although there is the possibility, as noted in Section 4, that markets may collapse and disappear through lack of support.It is perhaps unusual to commence a book on money with an explicit account of the market micro-structure being assumed throughout the rest of the book. I argue, in Section 5, not only that the system described here of market makers adjusting their quoted prices is more realistic than the alternative Arrow—Debreu Walrasian auctioneer paradigm, but also that many practical aspects of monetary phenomena can be properly analysed only against such a more realistic background.