In this article I elaborate on an earlier work to suggest a new set of rules of thumb for determining the occurrence, magnitude and intensity of a global crisis. This approach also draws on previous attempts to define a recession using quarterly data.
As is well known, in general, a recession is defined as two consecutive quarters of negative real economic growth, following the original idea put forward by Julius Shiskin in the New York Times in 1974. However, this definition is not accepted by the National Bureau of Economic Research (NBER), in charge of the Business Cycle Dating Committee (BCDC), which expresses a recession in terms of
"a significant decline in activity spread across the economy, lasting more than a few months, normally visible in real GDP, real income, employment, industrial production, and wholesale-retail sales. A recession begins just after the economy reaches a peak activity and ends as the economy reaches its trough. Between trough and peak, the economy is in an expansion. Expansion is the normal state of the economy; most recessions are brief and they have been rare in recent decades."
The NBER gives three reasons to explain why the BCDC doesn't agree to take the Shiskin procedure: "First, we consider the depth as well as the duration of the decline in economic activity. Recall that our definition includes the phrase, 'a significant decline in economic activity.' Second, we use a broader array of indicators than just real GDP. One reason for this is that the GDP data are subject to considerable revision. Third, we use monthly indicators to arrive at a monthly chronology."
In what follows I will try to demonstrate that the set of rules of thumb to determine a global crisis developed in this piece generates somewhat similar results to the definition of recession by the BCDC. This is particularly apparent with the two most important crises identified in my previous contributions, namely: The Gulf War and the Great Recession. But it's not necessarily the case with the other three recessions detected since 1986: The Stock Crash, the Early 2000 Recessions, and the current crisis. The article closes pinpointing some reasons as to why the new approach might be useful in predicting both relatively weak and new crises not ascertained or anticipated by the BCDC method.
Let's start with the new rules of thumb. They can be visualized in Figure 1, where the five global crises previously identified are depicted. To begin with, what these crises share in common is that they all reflect a certain number of consecutive WTI oil price coefficients of variation above the average for the whole period of analysis. Secondly, these deviations could be added up to reveal the relative importance of each crisis. And thirdly, in only two global crises the Shiskin criterion of recession is also met.
Next in Table 1 these regularities are converted into rules of thumb to determine the occurrence, magnitude and intensity of global crises. First, the occurrence of a global crisis is expressed in terms of at least three consecutive quarters of the largest positive differences between WTI Oil Price Coefficients of Variation and their average for the whole period of analysis (1986 Q1 - 2016 Q2). Second, the magnitude of a global crisis is denoted by the sum of at least three consecutive quarters of those positive differences. Third, the intensity of a global crisis is detected whenever it also shows at least two consecutive quarters of real negative growth, a recession, as defined by Shiskin. Lastly, a ranking of global crises is then established on basis of both the magnitude and intensity of a global crisis, where 1 represents the worst crisis; 2 a less severe one, and so on and so forth.
Rules of Thumb to Determine the Occurrence, Magnitude and Intensity of a Global Crisis
1986 Q1 - 2016 Q2
In Table 2 we can find a summary of results of the methodology. Note that the occurrence of the five global crises previously identified is confirmed, the Early-2000s Recession and the Great Recession being the longest crises followed by the Current Crisis, as well as the Stock Market Crash and the Gulf War. But the longest crises aren't necessarily the largest ones. In effect, as our rule of thumb for the magnitude of global crises indicates, the largest crisis was by far the Great Recession followed by the Gulf War, the Stock Market Crash, the Current Crisis and the Early-2000s Recession. In addition, as expected, the intensity measure applies to the two harshest global crises only, namely the Gulf War and the Great Recession, where the latter appears to be more intense than the former. Finally, the ranking of crises gives the first place to the Great Recession, the second to the Gulf War, the third to the Stock Market Crash, the fourth to the Current Crisis and the fifth to the Early-2000s Recession.
Summary of Results
Source: Table 1.
We are now in a position to compare the occurrence results with NBER's Business Cycle Reference Dates. This is performed in Table 3, where it's clear that there's a perfect match between the two methodologies for one crisis (i.e. Gulf War); an almost complete (86%) correspondence for another (i.e. Great Recession); a total miss with a third one (i.e. Early 2000s Recession); an unnoticed crisis (i.e. Stock Market Crash) by the NBER procedure which makes the comparison not possible; and a not-yet-determined one (i.e. Current Crisis) by the NBER approach.
Comparison between NBER's Procedure and this Approach
Sources: nber and Table 1.
It's indeed remarkable that the two most important global crises in the last three decades were picked up quite well by the approach suggested in this contribution. Nevertheless, a few words on the other three crises are in order.
First, regarding the Stock Market Crash, in another article I have already suggested some reasons as to why the high WTI oil price coefficient of variation in 1986 wasn't reflected in any important fall in the U.S. GDP growth rate thereafter. Here we are left with the impression that relatively weak global crises such as the Stock Market Crash may not be captured by the NBER Business Cycle Reference Dating procedure because of their relatively mild impact on GDP growth. But this raises some doubts as to the effectiveness of the NBER approach in not relying on that very variable to determine the timing of a recession. Another concomitant point here appears to be that only the most severe crises would imply negative economic growth rates at the same time, which is consistent with the findings of the present investigation.
Second, in the Early 2000s Recession the six consecutive positive differences for the period 1999 3Q - 2000 4Q aren't concurrent at all with a negative real growth of the U.S. economy which only shows up in 2001. But this doesn't invalidate my rules of thumb for determining a crisis. It only demonstrates that not all crises are created equal. As argued in a previous contribution, here the causal relationship from WTI oil price coefficients of variation to real economic growth doesn't seem to be particularly clear-cut. In fact, in this case two oil price volatility peaks (one in 1999 and another in 2000) would have produced a recession in the U.S. that started with a deceleration in 2000 followed by a fall of GDP growth in 2001 and a final recovery only two years later. This boils down to a rather extensive global crisis and a rather retarded effect of oil price fluctuations on GDP growth. Again, because the NBER system relies, despite arguing otherwise, heavily on negative growth rates to determine whether we are in a recession, that methodology failed to pick up that slow impact. Lastly, it would be too presumptuous on my part to think that oil price volatility is the only factor causing a global crisis, though I have shown in my analysis sufficient elements to argue that every time we have oil price volatility (either because of an oil price hike or an oil price fall), we should have reasons to worry about the near future of the economy, which make the rules of thumb put forward here particularly useful for policy makers.
Third, last year I had already found an annual coefficient of variation for 2014 above the total average for the period 1986-2014 and growth rates (for a number of years, including 2014) consistently below the total average for the whole lapse under scrutiny which - I thought - might be an indication of a global economic/financial crisis. I then wondered "whether sound economic policy will be capable of avoiding this time that negative outcome, just as it apparently did in 1987 during the Stock Market Crash." I am now more inclined to believe that the Current Crisis appears more akin to the Early 2000s Recession than to the Stock Market Crash. For one thing, contrary in part to my intuition in another piece published a few months back, it does seem like it will be more extensive than originally thought. For another, it also begins to reflect a somewhat retarded effect which may eventually translate into negative growth rates only by next year. The question remains though as to whether policy-makers will grasp these insights and act in a way so as to lessen or offset its already foreseen impact.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.