The volatile movements in commodity prices during the last years have raised new questions to be discussed in both policy and academic spheres. Up to the financial slump of the second quarter of 2008 commodity prices grew fast for several consecutive years in a highly volatile context. The international financial crisis suddenly broke this trend. However, after the abrupt correction during the most traumatic months of the crisis, commodity prices started a recovery, although the rate of growth was slower than in the previous cycle. Thus, this paper enters into this re-energized debate by developing a theoretical and empirical model that tries to help in understanding both those factors that affect long-run movements in prices and those that influence their short-run dynamics. In particular, the so-called financialization of commodities phenomenon is analyzed claiming that the impact of this process might be reflected in short-run movements in prices, but not in long-run or equilibrium values. In technical terms specifically, the main hypothesis is that the process of financialization of commodities could generate a non-linear adjustment pattern of prices towards their long-run equilibrium values. In order to theoretically support this hypothesis, a model of heterogeneous agents that interact in commodity markets is developed. The empirical evaluation is done with a smooth transition vector autoregressive model. Concerning long-run fundamentals of prices, this study focuses on the macroeconomic determinants of them. Among fundamentals, there is a clear role for world income (proxy variable for global demand). According to the view of several analysts this demand, would be growing at a pace that is not consistent with the expansion of international supply. The value of the US Dollar also plays an important part in the story. The influence of this variable has been stressed in other historical cycles of commodity prices. Additionally, easy monetary conditions in developed countries and the context of excessive liquidity have been pointed out as important elements by several authors. They would tend to add inflationary pressures that are more rapidly transmitted into goods with flexible prices as it is the case of commodities. In relation to the empirical front, the estimated model highlights the importance of both the macroeconomic fundamentals included in the long-run relationship and the significance of the hypothesis of non-linear price adjustment to the equilibrium value. The results suggest that large misalignments between spot and fundamental prices tend to be corrected relatively fast, while small gaps persist in time since in this case there are not endogenous correcting forces. Finally, this document covers two processes that were originated thirty years ago, but have intensified during the last decade. These processes cannot be omitted when the objective is to use the empirical results of the paper to project future trends in commodity markets. On the one hand, the issue of biofuels as a substitute of more traditional sources of energy is analyzed. On the other hand, the changing pattern of world demand due to structural transformations that are taking place in increasingly important emerging economies such as China and India is considered. Regarding policy implications for commodity-dependent developing countries, it must be stressed that commodity price misalignments should be carefully monitored: price reversions tend to be abrupt when the gap between actual and fundamental price is higher than 20-25%. It is also important to note that factors affecting commodity prices (like real international interest rates and the US real exchange rate) are similar to those that influence capital flows. This explains why it is hard for developing countries to cushion terms of trade shocks with external finance. The same fundamentals that worsen terms of trade affect negatively the access to international credit market.