As financial crises became more frequent during recent decades, so also have books about them, as well as about crises that occurred centuries earlier. These studies tend to come in three styles: (1) narratives of one or more crises, usually by historians and journalists; (2) financial-economic-heuristic models of the typical patterns or essential stages of crises, as illustrated by historical examples, often adduced by financial economists with historical interests; and (3) large-scale data assemblies and econometric analyses covering decades or centuries of crises by economists.1Boom and Bust, written by two financial economists who observed firsthand from their posts in Belfast the devastating effects of the global financial crisis from 2007 to 2009 on Ireland, combines styles 1 and 2. The authors call their simple model “the bubble triangle.” Its three sides (with analogies to a fire) are marketability (oxygen), money and credit (fuel), and speculation (heat). Financial bubbles form when it is easy to buy and sell assets, when ample money and credit is available to finance speculative purchases, and when people are in a mood to speculate. These, however, are only necessary conditions for a bubble. They become sufficiently hot to inflate a bubble only with the right spark, which the authors contend can come from either of two sources, a captivating technological breakthrough (for example, the internet leading to the “dot com” bubble) or governmental policies that cause a rise in asset prices (for example, home-ownership policies leading to the sub-prime bubble).The authors develop narratives of several bubbles and then analyze each of them using their bubble-triangle framework. Some of the cases—the Mississippi and South Sea bubbles of 1719/20, the British railway mania of the 1840s, the Wall Street stock bubble of the late 1920s, Japan’s stock and real-estate bubble of the 1980s, and the recent “dot com” and sub-prime bubbles—are familiar to financial historians. Others such as the Australian real-estate bubble of the 1880s and the British bicycle mania of the 1890s are more obscure.The book’s discussion and analysis of two stock-market bubbles in China, in 2007 and 2015, is particularly useful because these bubbles are recent and not well known. In each instance, the Chinese authorities engineered a bubble to advance a policy goal. In 2007, the authorities wanted to privatize government-owned untraded shares in state-owned enterprises into privately owned tradable shares, and a bubble eased achievement of the goal. In 2015, after encouraging massive private-sector borrowing to alleviate the impact of the global crisis from 2007 to 2009, China’s authorities hoped that a new stock bubble would allow companies to raise equity capital on easy terms and use it to reduce their bloated debts. China’s 2015 bubble promoted debt for equity swaps to ease corporate debt burdens in the same way that John Law’s Mississippi bubble and the South Sea Company’s bubble did in 1719/20 to ease the debt burdens of the French and British governments. Financial history is cyclical, forever blowing bubbles.Quinn and Turner tout their model as predictive. A test of its predictive power seems underway (as of mid-2021), when the S&P 500 stock index hit an all-time high. All the authors’ necessary conditions for a bubble scenario appear to be in place. Competition of financial firms for assets and trades has made the marketability of stocks and other assets easier and cheaper than ever before. The Federal Reserve roughly doubled its balance sheet over the past year, vastly expanding the supply of money and credit. To fight the economic slowdown caused by the Covid-19 pandemic, Congress enacted trillions of dollars of fiscal stimulus, with more on the way. Signs of speculative excesses abound, with names such as Archegos, Bitcoin, Credit Suisse, GameStop, Greensill, non-fungible tokens (nfts), Nomura, Reddit, Robinhood, and special-purpose acquisition companies (spacs) attached to them. So far, none of these potential sparks has started the deflation of a bubble, if indeed there is one in mid-2021. Conventional wisdom, challenged by Quinn and Turner, holds that we can know a bubble only after it bursts. By the time this review is published, we may have some test results.
Read full abstract