TOIL AND TROUBLE, DON'T GET BURNED SHORTING BUBBLES

Bubbles are among the most puzzling and controversial phenomena of financial markets. Although rare, their cumulative impact on both investor returns and the broader economy can be great. One particular question that has motivated research is why shrewd short sellers don't prevent excessive price increases. The "limits to arbitrage" idea argues that correcting inefficient market prices is neither easy, cheap nor riskless. The "rational bubble" literature identifies situations in which being long the bubble is a better trade than being short, even if investors know for certain the bubble will pop. And there is a theory that bubbles only inflate after the shorts have suffered significant losses. We examine the "short subprime" trade from 2005 to 2008 to evaluate these and other explanations. We argue that the short subprime trades had more risk than is commonly appreciated. We discuss how the opaque and illiquid nature of subprime mortgages deterred some investors from purchasing CDS contracts and note that other investors assessed the risk of counterparty failure, government intervention and unknown time horizon to be sufficient enough not to purchase CDS contracts. In addition, we describe how factors such as performance convexity and credit convexity made the subprime short more profitable than most ex-ante calculations suggested. We also outline why the subprime short trade was ineffective at reining in the subprime bubble and how buying subprime after the crisis was an equally, if not more, attractive trade that potentially did more to mitigate the harm of the bubble. Looking back at the last major bubble with a decade of hindsight yields insights that might be helpful to market participants and policy makers thinking about future bubbles.