This paper develops a model for understanding end-User order flow in the FX market. The model addresses several puzzling findings. First, the estimated price-impact of flow from different end-user segments is, dollar-for-dollar, quite different. Second, order flow from segments traditionally thought to be liquidity-motivated actually has power to forecast exchange rates. Third, about one-third of order flow's power to forecast exchange rates 1 month ahead comes from flow's ability to forecast future flow, whereas the remaining two-thirds applies to price components unrelated to Future flow. We show that all of these features arise naturally from end-User heterogeneity, in a setting where order flow provides timely information to market-makers about the state of the macro-economy. Copyright (c) 2006 John Wiley & Sons, Ltd.

This dissertation examines the influence of attitudes, beliefs, and preferences of energy industry practitioners on investment decision-making with regard to fuel choice for new electric power plants. The conclusions are based on in-depth interviews and an extensive online survey I conducted of 600-800 energy professionals in the U.S. power sector.

Chapter 1 analyzes the impact of policy uncertainty on investment decision-making in renewable energy, using the federal production tax credit (PTC) and wind energy investment as an example. It is generally understood that the pattern of repeated expiration and short-term renewal of the PTC causes a boom-bust cycle in wind power plant investment in the U.S. This on-off pattern is detrimental to the wind industry, since ramp-up and ramp-down costs are high, and players are deterred from making long-term investments.

The widely held belief that the severe downturn in investment during "off" years implies that wind power is unviable without the PTC turns out to be unsubstantiated: this chapter demonstrates that it is not the absence of the PTC that causes the investment downturn during "off" years, but rather the uncertainty over its return. Specifically, it is the dynamic of power purchase agreement negotiations in the face of PTC renewal uncertainty that drives investment volatility. This suggests that reducing regulatory uncertainty is a crucial component of effective renewable energy policy. The PTC as currently structured is not the only means, existing or potential, for encouraging wind power investment. Using data from my survey, various alternative policy incentives are considered and compared in terms of their perceived reliability for supporting long-term investment.

Chapter 2 introduces the concept of expected payment of carbon as a factor in investment decision-making. The notion of carbon risk (the financial risk associated with CO2 emissions under potential climate change policy) is usually incorporated into investment decision-making by including a cost of carbon in the budget analysis. Most existing literature uses the expected price of carbon as a proxy for this cost, where expected price is a weighted average of various scenarios, often comparing policy proposals and representing either the price of traded permits or level of carbon tax, depending on the type of policy. The literature focuses on the minimum price of carbon required to influence power plant investment decisions.

In contrast, this chapter introduces expected payment as a more accurate measure of carbon cost as it is perceived by industry practitioners. The expected payment of carbon is the expected price of carbon times the probability that this cost would actually be faced in the case of a particular investment. This concept helps explain both the 2005-2006 surge of activity in coal-fired power plant development and the subsequent decline in that interest.

The energy industry has been slow to move away from fossil fuels and towards renewable resources. In chapter 3 I find evidence for a cognitive bias that plays a role in this momentum. Energy executives' expectations of future energy prices are strongly correlated with their own preferences, which I document for the case of natural gas prices. This is an example of wishful expectations, a form of overconfidence in which people are excessively optimistic over uncontrollable future outcomes. This implies energy executives with strong exposure to fossil fuels are excessively optimistic on future prices and so continue to invest despite the presence of superior alternatives.

Macroeconomic models of nominal exchange rates perform poorly. In sample, R 2 statistics as high as 10 percent are rare. Out of sample, these models are typically out-forecast by a naïve random walk. This paper presents a model of a new kind. Instead of relying exclusively on macroeconomic determinants, the model includes a determinant from the field of microstructure-order flow. Order flow is the proximate determinant of price in all microstructure models. This is a radically different approach to exchange rate determination. It is also strikingly successful in accounting for realized rates. Our model of daily exchange-rate changes produces R 2 statistics above 50 percent. Out of sample, our model produces significantly better short-horizon forecasts than a random walk. For the DM/$ spot market as a whole, we find that $1 billion of net dollar purchases increases the DM price of a dollar by about 1 pfennig.

We examine the trading strategies of mutual funds in emerging markets. We develop a method for disentangling the behavior of fund managers from that of underlying investors. For both managers and investors, we strongly reject the null hypothesis of no momentum trading: mutual funds systematically sell losers and buy winners. Selling current losers and buying current winners is stronger during crises, and equally strong for managers and investors. Selling past losers and buying past winners is stronger for managers. Managers and investors also practice contagion trading-they sell (buy) assets from one country when asset prices fall (rise) in another. (C) 2003 Elsevier B.V. All rights reserved.

Dealers need to search for quotes in many of the world's largest markets (such as spot foreign exchange, US government bonds, and the London Stock Exchange). This search affects trading cost. We estimate the share of total trading cost attributable to search. Our experiments show that the share is large -- roughly one-third of the effective spread. Past work on estimating spread components typically omits the search component. Our estimates suggest this omission is important.