Introduction

My first job as a trainee research analyst started at the end of 1985. Since then, many things in economies and society have changed beyond recognition. The world has become more interconnected; the Cold War ended and the Soviet Empire unravelled, heralding an era of 'globalisation'. When I began my career, the UK had only recently, in 1979, removed restrictions on foreign exchange controls (for the first time in 90 years), while France and Italy still had them in place, only abolishing them in 1990.1 Economic conditions have also transformed, and several key fundamental macro drivers have shifted dramatically: over the past three decades, inflation has fallen persistently and interest rates have collapsed; 10-year government bond yields in the United States have come down from over 11% to 2%, Fed funds rates have fallen from over 8% to 1.5% and currently one-quarter of all government bonds globally have a negative yield. Inflation expectations have become well anchored and economic volatility has declined.

Meanwhile, technological innovations have also altered how we work and communicate, and computing power has revolutionised the ability to process and analyse data. The most powerful supercomputer (the Cray-2) in 1985 had a similar processing ability to an iPhone 4.2 The scale of the digital revolution and the quantity of available data since then would have been unimaginable at the time, and this seems to be accelerating. Microsoft's president Brad Smith recently signalled that 'this decade will end with almost 25 times as much digital data as when it began'.3

Over the same period, there have been three major recessions (in most economies) and several financial crises, including the US Savings & Loan crisis of 1986, the Black Monday stock market crash of 1987, the Japanese asset bubble and collapse between 1986 and 1992, the Mexican crisis of 1984, the Emerging Market crises of the 1990s (Asia in 1997, Russia in 1998 and Argentina in 1998-2002), the ERM currency crisis of 1992, the technology collapse in 2000 and, of course, the most recent global financial crisis, starting with the subprime mortgage and US housing declines of 2007, and the European sovereign debt crisis of 2010/2011.

Despite the huge changes in economic conditions and technology over the past three decades, and occasional financial and economic crises, there has been a tendency for similar patterns to repeat themselves in financial markets, and for cycles to emerge, albeit in slightly different forms. In a 2019 paper, authors Filardo, Lombardi and Raczo noted that, over the past 120 years, the US has gone through the Gold Standard period, when inflation was low, and the 1970s, when inflation was high and volatile, and that over this long historical period the price stability credentials of central banks has shifted and fiscal and regulatory policies have varied considerably, but that 'through all of this, the financial cycle dynamics have remained a constant feature of the economy'.4

It is these cycles, and the factors that drive them, that this book explores. Its purpose is to show that, despite significant changes in circumstances and environments, there still appear to be repeated patterns of performance and behaviour in economies and financial markets over time.

But, although acknowledging the changes, and trying to assess how much of the change we observe is cyclical and how much is structural, the main body of this book aims to examine what there is about financial markets that is predictable, or at least probable.

Interest in economic cycles, and their impact on financial markets and prices, has a long history and there are many theories on how they function. The Kitchin cycle, after Joseph Kitchin (1861-1932), is based on a 40-month duration, driven by commodities and inventories. The Juglar cycle is used to predict capital investment (Clement Juglar, 1819-1905) and has a duration of 7-11 years, whereas the Kuznets cycle for predicting incomes (Simon Kuznets, 1901-1985) has a duration of 15-25 years and the Kondratiev cycle (Nikolai Kondratiev, 1892-1938) has a duration of 50-60 years, driven by major technological innovations. There are, clearly, problems with all of them and the fact that there are so many different descriptions of cycles points to the fact that there are many different drivers. Several of them, such as the very long Kondratiev cycle, are difficult to test statistically given the existence of so few observations.

Although the traditional focus on cycles has related mainly to the economy, the focus in this book is on financial cycles, their drivers and different phases - a topic discussed in detail in chapter 3. The idea that there are cycles in financial markets in general, and in equity markets in particular, has been with us for a very long time. Fisher (1933) and Keynes (1936) both examined the interaction between the real economy and the financial sector in the Great Depression. Burns and Mitchell found evidence of the business cycle in 1946 and later academics argued that the financial cycle was a part of the business cycle, and that financial conditions and private sector balance sheet health are both important triggers of the cycle and factors that can amplify cycles (Eckstein and Sinai 1986). Other research has demonstrated that waves of global liquidity can interact with domestic financial cycles, thereby creating excessive financial conditions in some cases (Bruno and Shin 2015).5

More recent studies suggest that measures of slack in the economy (or output gaps - the growth rate versus potential output) can be explained partly by financial factors (Boria, Piti and Juselius 2013) that play a large part in explaining fluctuations in economic output and potential growth, as well as 'determining which output trajectories are sustainable and which are not',6 thereby implying a close link and feedback loop between financial and economic cycles.

That said, although interest in economic and financial cycles has a long history, views on whether they can be predicted are widely contested. One set of ideas about the inability to anticipate future price movements in markets stems from the efficient market hypothesis (Fama 1970), which argues that the price of a stock, or the value of a market, reflects all of the information available about that stock or market at any given time; the market is efficient in pricing and so is always correctly priced unless or until something changes. Following on from this idea is the argument that an investor cannot really predict the market, or how a company will perform. This is because no individual will have more information than is already reflected in the market at any time, because the market is always efficient and prices change in fundamental factors (such as economic events) immediately.

But theory is one thing and practice is another. Nobel Laureate Robert Shiller, for example, showed that while stock prices are extremely volatile over the short term, their valuation, or price/earnings ratio, provides information which makes them somewhat predictable over long periods (Shiller 1980), suggesting that valuation at least provides something of a guide to future returns. Others have argued that the returns one can expect from financial assets are linked to economic conditions and therefore the probability of certain outcomes can be assessed even if accurate forecasts are not particularly reliable.

Although there are relationships between financial cycles and economic cycles, mainly because bonds are affected by inflation expectations and equities by growth, there are some patterns of human behaviour that reflect and sometimes amplify expected economic conditions. It is the way in which economic and corporate fundamentals (expected growth, profit, inflation and interest rates, for example) are perceived by investors that is the crucial mix. Academic work has increasingly shown that risk-taking appetite has been a key channel through which supportive policy (for example, low interest rates) can affect cycles (Borio 2013).7 Willingness to take risk and periods of excessive caution (often after a period of weak returns) are factors that tend to amplify the impact of economic fundamentals on financial markets and contribute to cycles and repeated patterns.

The emotions of fear and greed, of optimism and despair, and the power of crowd behaviour and consensus can transcend specific periods of time or events, supporting the tendency for patterns to be repeated in financial markets even under very different circumstances and conditions. There is also a tendency for errors to be repeated when investors fail to heed some of the important warning signals of overheating and excess that can develop when conditions are supportive and there is a powerful narrative. I discuss this topic in chapter 8, which looks at the role of sentiment in developing speculative excesses and financial bubbles.

Of course, although there are repeated patterns in markets over time, there are also events and economic conditions that are unique to each cycle or circumstance. In reality, no two periods are ever precisely the same; even faced with fairly similar conditions, the precise permutations of factors are unlikely to be repeated in the same way. Structural changes in industries and in economic factors, such as inflation and the cost of capital, can shift relationships between variables over time. For example, the behaviour and performance of a stock market cycle in an era of high inflation and interest rates...