I. Optimum trade in public goods, 63. — II. The all-or-none offer, 70.— III. The core of Pareto-optimal outcomes, 73. — IV. Summary, 78.

We document spillover effects of the 2000–2002 Treasury Buyback program on Treasury returns and the composition of the Federal Reserve’s System Open Market Account (SOMA) portfolio. The reduction in bond supply due to the buybacks contributed an average of 95 basis points to the yields of bonds bought back and bonds of similar maturity over the course of the program. Each $10 billion of purchases corresponded with an average yield increase of 7.8 basis points. At a higher frequency, prices of purchased and near substitute bonds increased on settlement dates. Changes to the SOMA portfolio were smaller for securities exposed to the buybacks and tended to occur outside of auction weeks, consistent with the Federal Reserve attempting to avoid exacerbating Treasury supply shortages. We relate our findings to the theoretical literature on asset supply in preferred habitats models of the term structure. Our results suggest that the proposed reintroduction of the Treasury buyback program will have limited effects due to its size and proposed composition.

This study applies Lance Girton and Don Roper's (hereafter G-R) monetary model of market pressure to the postwar Brazilian monetary experience. The model was designed specifically for the Canadian managed float during the period 1952-62. The object of their model is to explain what they term exchange market pressure; that is, the pressure on foreign reserves and the rate when there exists an excess of domestic money supply over money demand in a managed floating rate regime. The basic theoretical proposition is that any such excess supply of money can be relieved by an depreciation, a loss in foreign reserves, or, in the context of a managed float, by some combination of the two. In this sense, the G-R managed float model used here is firmly rooted in the modern monetary approach to rates and the balance of payments.' Brazil provides a particularly good example for testing this approach, not only because it is in many senses a unique example of a postwar managed float system, but also because it can be treated as a small, open economy in the sense that world prices and monetary conditions faced by Brazil are taken as given. This particularly suits the purpose of most modern monetary models which make this assumption and obviates the problems of monetary dependence and neutralization dealt with in the pioneering G-R paper. Specifically, the small-country assumption permits us to devise a simple one-country equation of managed floating which depends upon four essential ingredients: 1) money demand, 2) money supply, 3) purchasing power parity, and 4) monetary equilibrium.2 Furthermore, in Brazil a much greater proportion of market pressure was absorbed by rate depreciation than in the Canadian case where changes in reserves were large relative to rate movements. In short, postwar Brazil provides a singularly good opportunity to test the monetary model of market pressure. Section I briefly states the essential elements of the monetary model, and derives the equation to be tested for the Brazilian experience from 1955 to 1975. Section II reports empirical results for the market pressure model, and Section III examines the applicability of the relative version of purchasing power parity for the time period considered. Section IV summarizes the results and discusses the merits of the monetary approach in light of the Brazilian experience.

The purpose of this study is to introduce the element of time into the analysis of optinial taxation of international capital movements. As with Irving Fisher, in his Theory of Interest, the . . supply and demand we have to deal with are . .. the supply and demand of future income, and we shall interpret a rate of interest as . . that sort of price which links one point of time with another point of time in the markets of the world (pp. 32-33). The analysis stresses the formal identity between capital theory, involving exchange over time, and trade theory, involving exchange at the same point of time and argues that, as a consequence, the theory of the optimum tax on capital movements may be regarded as a branch of optimal tariff theory. The correspondence between trade theory and capital theory has appeared in many guises throughout the past century, and was recognized at least as early as 1907 by Fisher. His classic graphical apparatus anticipates much of the later opportunity cost approach to international trade, as well as the more recent work in general equilibrium encompassing both theories.1 In this latter work, time is broken up into a (finite) number of successive time periods, and the same physical item available at different time periods and locations is treated as a different commodity. Future production and consumption possibilities are assumed to be perfectly known today and economic agents acting as price takers engage in markets in which prices for delivery in future periods are quoted, and, under certainty, will correctly reflect future scarcity values. As an immediate consequence of this approach, optimum tariff theory is no less applicable when trade over time is included with trade at a given point in time. Prices quoted in the present reflect marginal rates of substitution not only between present commodities, but also between future commodities, and between present and future commodities. A country then stands to gain by taxing exchange over time in the same manner as it may in taxing exchange at the same point in time. Indeed, it has the same purpose-to influence the terms of trade. That the terms of trade over time are called interest rates, or that a lender is treated as an exporter and a borrower as an importer of present income makes little difference. The resulting optimum taxes on trade are simply those taxes which equate the marginal rates of transformation through trade at the same point in time, and over time, with the marginal rates of transformation in domestic production and consumption. This is a somewhat different approach to the problem of optimal taxation of capital movements than that recently pursued by Murray Kemp and Ronald Jones. The major difference is the explicit role played in our analysis by the element of time-of waiting -which allows for pure borrowing or lending, in addition to the financing of imported capital goods by current exports. In a later section we will relate the two approaches.