No recession---impossible to coordinate the next one and responses solve
Cristino et al 20 (Cláudio Tadeu Cristino, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing – review & editing, Piotr Żebrowski, Conceptualization, Methodology, Supervision, Writing – review & editing, Matthias Wildemeersch, Conceptualization, Formal analysis, Methodology, Supervision, Writing – original draft, Writing – review & editing, published 5/7/20, accessed 7/23/22, “Assessing the time intervals between economic recessions”, https://doi.org/10.1371/journal.pone.0232615) KCabrera
Recessions are economic contractions following a period of economic expansion. They are part of the business cycle, and do not feature a regular periodicity. Market adjustments and policy measures are capable to revitalize the economy and turn around the trend of economic decline. In the modern economic history of the U.S., policy makers have tried time and time again to address unemployment and declining incomes with recovery measures in the New Deal, the employment act, new institutional frameworks, the Bretton Woods agreement, etc. [1, 2]. In addition to reforms and new regulations, different instruments of fiscal control can be activated, such as adjustments in taxation, public expenditure and debt management to foster economic growth [3]. The efficiency of all these policies can be measured with so-called fiscal multipliers, which behave differently during expansions and recessions [4, 5]. However, these multipliers do not allow us to estimate the time and probability of the next recession. The lifetime of a system is defined as the time that a system can carry out its function before going into a failure state [6, 7]. In this work, the failure state corresponds with a recession event, and the age of the system is the time that the system has already worked properly, i.e., the time that the economy is in expansion. There are multiple approaches to predict business cycle turning points, such as econometric techniques [8], the spread between short-term and long-term interest rates captured in the yield function [9], or stochastic simulation [10, 11]. Typically, the probability of a recession is determined based on the several indicator variables, such as stock prices, credit market activity, or employment and interest rate [11]. These techniques are valuable, but they do not capture explicitly the effect of system age, system aging, or the effects of policy and market interventions on the state of the economy. There is in fact evidence that the termination probability of recessions is dependent on its duration [12, 13]. This dependence can be captured by different types of survival models, such as the proportional hazard model [14] and the accelerated failure time model [15], which include the effect of covariates in the hazard function. Aforementioned models are however not able to include the effect of interventions following an event on system age. Business cycle durations have typically been modeled using Markov switching processes [16, 17]. Here, we use an alternative approach based on Generalized Renewal Processes (GRP) to model the inter-arrival times of recessions. This method allows us to integrate explicitly the effect of policy interventions and to determine if aging is a real process happening in economic systems. GRPs integrate simultaneously two sources of stochastic variability in system performance: deterioration through aging and restoration by means of policy interventions. System deterioration can occur through different types of aging processes. No aging indicates that the age of the system has no effect on the residual lifetime. The no-aging property implies a system with no memory, a constant failure rate, and lifetimes distributed according to the exponential distribution. In reliability analysis, positive aging is much more common and means that the expected residual lifetime decreases with age. System restoration deals with repair actions that improve the system after failure. The literature often distinguishes between perfect repair and minimal repair [18]. Perfect repair is applicable, for instance, to engineering systems that consist of a single component that can be replaced, resulting in a system as good as new. Minimal repair requires the replacement of only these components that failed; minimal repair is a reasonable assumption for systems consisting of multiple components all of which have their own failure properties, leaving the system in a state as bad as before the failure. Minimal repair and perfect repair are however specific cases and in realistic settings the degree of restoration can be better described by general repair and the corresponding stochastic process, the GRP. The effectiveness of interventions in general repair can take arbitrary values, and general renewal theory extends the classical renewal theory by including the notion of rejuvenation, indicating the performance of the system after interventions. This idea was firstly presented in [19] and requires to introduce the concept of virtual age, a function that reflects how interventions affect the real age of the system. In the literature, a Weibull-GRP (WGRP) model is usually fitted to observed data in order to infer lifetime probabilities [20–23]. The restoration process is commonly captured by a rejuvenation parameter [22, 24, 25], whereas the deterioration of the system is reflected by the shape and scale parameters of the Weibull distribution regardless of the restoration process [20–22]. The GRP can also be used in an optimization context, where policies are searched for that minimize intervention costs while maintaining certain levels of system performance [24, 26–28]. In reliability analysis and engineering, the Weibull distribution has typically been used to describe the functioning of a serial system that operates well as long as all components operate correctly [29]. In this work, we propose to model the arrival time of recessions by means of a GRP based on the Gumbel distribution. The Gumbel distribution is able to describe a parallel system that stops operating once all components fail, and we use the abstraction of a parallel system as a stylized representation of the economy. The duration of an expansion period is affected by market adjustments taking place during the preceding recession as well as by the policies and reforms introduced to mitigate the causes of this recession. We therefore argue that the periods between consecutive recessions do not constitute a stationary process. We deduce the distribution of the GRP based on the Gumbel distribution, and show how the parameters of the statistical model can be estimated. Importantly, we propose a novel goodness of fit test to demonstrate that the observed recessions of the American and European markets can be represented by the Gumbel-based GRP. We show that the GRP outperforms several stationary distributions, and in addition to the recession incidence rate over time, we also estimate the expected time of the next recession.