B.), a Columbia University Undergraduate WEP grant (S.K.C.), a 5U01 MH078844-05 grant (Z.J.H.), and the Howard Hughes Medical Institute (S.A.S.). “
“In February 1637 in Amsterdam, the cost of a single exotic tulip bulb reached a price equal to ten times what a skilled craftsman earned in a year. The price of the same bulb collapsed a few days later. The dramatic rise and fall of tulip bulb prices is a famous historical example of a financial bubble (Kindleberger and Aliber, 2005). A bubble is conventionally defined by active trading of an asset at prices that are considerably higher than its intrinsic fundamental value. Examples of modern bubbles

include Japanese stocks in the 1990s, the US high-tech sector in the late 1990s, and housing prices, which rose and crashed in many countries from 2000–2008. All of these bubbles Akt cancer (especially the housing crash) caused long-lasting macroeconomic disruptions (Shiller, 2005). Modern bubble episodes have also led to a substantial shift in thinking about the capacity of prices to act as sober information aggregation mechanisms that guide efficient allocation of capital. Policy makers, academics, and market participants alike are now more familiar with, and groping to understand, the ways that prices can reflect pathological valuation and are actively debating Ku 0059436 whether policy

interventions can help (Akerlof and Shiller, 2009). Despite these dramatic historical and modern examples, there is no well-accepted theory of how bubbles start and end. One common definition of bubbles is rapid price appreciation followed by a crash (Brunnermeier, 2008). However,

this definition has no predictive power for identifying an ongoing bubble, since it does not identify a bubble before it crashes. Furthermore, fundamental asset values are rarely known with precision, so it is difficult to identify a bubble if bubbles are defined as prices above an elusive fundamental value. One way to learn about bubbles is to observe trading in an experimental market for artificial assets that have a known fundamental value. In these markets, price variation cannot be explained by changes TCL in fundamentals. In fact, several carefully controlled economics experiments have shown that certain classes of asset markets do generate price bubbles quite regularly, even when intrinsic values are easy to compute and are known to traders (Smith et al., 1988, Camerer and Weigelt, 1993, Porter and Smith, 2003 and Lei et al., 2004). The nature of bubbles has also been intensely investigated in theory (Abreu and Brunnermeier, 2003 and Yu and Xiong, 2011), but empirical reasons why bubbles arise and then crash are still not well understood in economics (Xiong, 2013). Recent work in neuroeconomics has shown how financial decision theory can be informed by neuroscientific data (Bossaerts, 2009). In particular, studies have started to dissect the neural mechanisms by which risk processing (Preuschoff et al.