1 PRICE STABILIZATION USING BUFFER STOCKS ∗ Iasson Karafyllis, George Athanasiou 1 and Stelios Kotsios Dept. of Economics, University of Athens, 8 Pesmazoglou Str., 10559, Athens, Greece Abstract The price stabilization problem is stated and solved for a mildly nonlinear cobweb model for a single commodity with stocks. It is shown that if the storage capacity for the particular commodity is sufficiently large then there exists a simple stabilization policy, such that the equilibrium price is a global attractor for the corresponding closed-loop system. JEL classification: C60; C61; C62; D40; L52 Keywords: Price Stabilization; Buffer Stocks; Cobweb Model 1. Introduction It is a well-known fact of economic reality that commodity prices are extremely volatile (Deaton and Laroque, 1992; Larson, Varangis and Yabuki, 1998; Osbourne, 2004). One of the instruments for price stabilization, which is found frequently in the economic literature, is the so-called buffer stock scheme. This instrument was implemented not only in developing countries but in the United States and Europe as well (He and Westerhoff, 2005, p. 2). For example, since the Second World War, the United States implemented buffer stock programs for various strategic materials and especially for oil (Nickols and Zeckhauser, 1977, p.p. 66-67). Jha and Srinivasan (2001) underline the importance of the operation of buffer stock programs for India. A retrospect in economic reality shows that the introduction of such programs, both in national and international level, faced difficulties with the funds that supported them (Larson, Varangis and Yabuki, 1998). The mechanism of a buffer stock program is to store a certain amount of the commodity in boom periods, when the price is low, and release a certain amount of the stored commodity in bust periods, when the price is high. Early theoretical work on the function of such programs include Keynes (1938), Kaldor (1939), Working (1949), Porter (1950) and Brennan (1958). We can discern many lines of research within the literature according to: i) who is conducting the stockpiling, ii) the policy rule, iii) the hypothesis about expectations and iv) the incidence of a speculative attack. The majority of the research is focused on the effect of government stockpiling on the stockpiling decision of risk-averse rational private agents with no market power (Wright and Williams, 1982b; Miranda and Helmberger, 1988; Wright and Williams, 1988; Newbery, 1989; Miranda and Glauber, 1993; Jha and Srinivasan, 1999; Jha and Srinivasan, 2001; Brennan, 2003). Under this framework, the stockpiling decision of the policy maker (the government or some international organization) can be conducted by following two rules: a “price band” (or “bandwidth”) rule or a “price peg” rule. There is also the possibility of implementing a mixture of these policy rules (Van Gronendaal and Vingerhoets, 1995, p. 260). When the policy maker follows the “price band” rule, the price is constrained in a predefined range (Miranda and Helmberger, 1988; Jha and Srinivasan, 1999; Jha and Srinivasan, 2001; Srinivasan and Jha, 2001; He and Westeroff, 2005). Whenever price hits the boundaries, the manager intervenes by increasing or decreasing the level of the stored commodity. In other words, the buffer stock agency provides a floor or incentive price to producers and a ceiling price to consumers. When the agency follows the “price peg” rule, it intervenes by augmenting or disposing stocks in order to keep the price as close as possible to a predefined target price path (Van Groenendaal and Vingerhoets, 1995). In other words, the manager adjusts his stocks so that the market price converges to the target price. The main methodology in order to derive the optimal private and government storage rules is the application of stochastic optimal control in a linear or non-linear setting. (Arzac, 1979; Teisberg, 1981; Wright and Williams, 1982a; Wright and Williams, 1982b; Newbery and Stiglitz, 1982; Wright and Williams, 1988; Jha and Srinivasan, 1999). Van Gronendaal and Vingerhoets (1995, p. 259) argue that we should search for buffer stock rules that are robust in a realistic setting. In order to support their argument they realized a comparison between the price behavior under an optimal buffer stock policy and the closed-loop price behavior under a buffer stock ∗ Research sponsored by the University of Athens contract 70/4/5026 1 Corresponding author. Granted by the State Scholarship Foundation. E-mail: gathanas@econ.uoa.gr