Business cycle

Business cycles are intervals of general expansion followed by recession in economic performance. The changes in economic activity that characterize business cycles have important implications for the welfare of the general population, government institutions, and private sector firms. There are numerous specific definitions of what constitutes a business cycle. The simplest and most naïve characterization comes from regarding recessions as 2 consecutive quarters of negative GDP growth. More satisfactory classifications are provided by, first including more economic indicators and second by looking for more informative data patterns than the ad hoc 2 quarter definition.

Definitions of business cycle fluctuations depend heavily on the specific set of macroeconomic variables examined and on the particulars of the methodology. In the United States, the National Bureau of Economic Research oversees a Business Cycle Dating Committee that defines a recession as "a significant decline in economic activity spread across the market, lasting more than a few months, normally visible in real GDP, real income, employment, industrial production, and wholesale-retail sales.". This has the advantage of incorporating multiple indicators and different assessments made by a group of experts. A few drawbacks are that recessions are commonly announced with a long time lag, that the specific judgment of committee members may have ad hoc elements or biases, and that the decisions can be hard to reproduce into a general rule. Nevertheless, the NBER recession dates are in widespread use as key timing indicators for historical business cycles.

Business cycles are usually thought of as medium term evolution. They are less related to long-term trends, coming from slowly-changing factors like technological advances. Further, a one period change, that is unusual over the course of one or two years, is often relegated to “noise”; an example is a worker strike or an isolated period of severe weather. This suggests that we remove these two components from the data in estimating the cycle movements. It would be difficult to determine the particular effects of long-term or noisy components by looking at complicated details for each case. However, a statistical approach can provide valuable insight.

Band-pass filters have been developed for economic data to extract mid-frequency fluctuations. Such filters also have the attraction that they offer more information about the state of the business cycle; the statement about the path of cyclical GDP as it comes out of recession adds interesting facts beyond just the labelling of when the switch from recession to expansion occurs. An example of a band-pass filter attempting to isolate business cycles is the Christiano-Fitzgerald filter However, such a fixed filter runs a substantial risk of spurious output, which renders any subsequent business cycle study misleading. The approach is also limited to a single indicator.

Adaptive band-pass filters have been used to extract business cycles coherent with the dynamic properties of the indicators. The filters introduced by Harvey-Trimbur have been applied in numerous studies examining diverse national economies. Unlike a fixed band pass filter that can only be applied to a single indicator, this more flexible approach can use multiple variables as inputs. Further, forecasts can be computed (on a timely basis). Lastly, uncertainty in business cycles can be gauged, making them useful for assessing macroeconomic risk.

The individual episodes of expansion/recession occur with changing duration and intensity over time. Typically their periodicity has a wide range from around 2 to 10 years. The technical term "stochastic cycle" is often used in statistics to describe this kind of process. Such flexible knowledge about the frequency of business cycles can actually be included in their mathematical study, using a Bayesian statistical paradigm.

There are numerous sources of business cycle movements such as rapid and significant changes in the price of oil or variation in consumer sentiment that affects overall spending in the macroeconomy and thus investment and firms' profits. Usually such sources are unpredictable in advance and can be viewed as random "shocks" to the cyclical pattern, as happened during the 2007–2008 financial crises or the COVID-19 pandemic. In past decades economists and statisticians have learned a great deal about business cycle fluctuations by researching the topic from various perspectives. Examples of methods that learn about business cycles from data include the Christiano–Fitzgerald, Hodrick–Prescott, singular spectrum, and Harvey-Trimbur filters.