Comment 82: Recession Watch 2011 (2) (August 2, 2011)

The official arbiter of U. S. economic cycles is The National Bureau of Economic Research. By majority vote, a panel of seven prominent economists in the bureau determines the timing of peaks and troughs of U. S. economy. The period from a peak to the following trough is naturally defined as the period of “recession". On the other hand, many media use a folklore like definition of recessions; two consecutive negative growth quarters of Real GDP is defined as the start of a recession and the recession ends when Real GDP returns to positive growth. However, the folklore type recessions deviate substantially from the official declaration of recessions from The National Bureau of Economic Research. For example, the bureau declared that the recession after the burst of the infamous dot-com bubble had started at March, 2001 and ended at November, 2001, an eight month long recession, whereas according to the folklore type definition the painful economic downturn after the burst of the dot-com bubble is not even a recession since there was no two consecutive negative growth quarters when Real GDP is concerned. By studying the reality of the past economic ups and downs, it is clear that the determinations of The National Bureau of Economic Research are much more appropriate than the folklore type simple definition that has the tendency to declare a recession too late and the exit from a recession too early.

The determination of recessions by The National Bureau of Economic Research, though is appropriate, has a drawback. There is no fixed criteria to decide when is the peak and when is the trough of the economic cycle. It is all up to the personal judgment of the economists on the panel of The National Bureau of Economic Research. The lack of an objective gauge to determine the onset and the end of recessions is inconvenient for the detailed analysis of recessions. We have searched for such an objective criteria for the start and the end of recessions such that the outcomes are much more closely aligned with the declarations of The National Bureau of Economic Research than the folklore type definition discussed above. The process of the search are discussed in Comment 64, article 12 and Comment 71. The criteria is summarized in the following three steps:

Step 1: For each quarterly Real GDP data, compute the year-to-year percentage change by comparing with Real GDP of the same quarter one year before. Let us call this “year-to-year % change of Real GDP". The calculated “year-to-year % change" is rounded up to the first digit after the decimal point. If this “year-to-year % change" drops to or below 1.0% after a sustained period of healthy growth, then the economy is already in a recession.

Step 2: Suppose the “year-to-year % change" of Real GDP satisfies the criteria of Step 1 at Quarter X, then move back two quarters from Quarter X, and define the new moved-back quarter as the starting quarter of the recession.

Step 3: After the economy falls into a recession, suppose at Quarter Y the “year-to-year % change" of Real GDP moves above 1.0%, then move back Quarter Y by two quarters and define the moved-back quarter as the one that the recession has ended.

Let us use the most recent severe recession as an illustration to see how the above mentioned criteria applies at the same time to explain why the process of “moving back by two quarters" in Step2 and 3 is necessary. After four years of boom and bubble, “year-to-year % change" of real GDP dropped to 1.0% at the second quarter of 2008 so that Step 1 of the criteria was satisfied. By moving two quarters back from the second quarter of 2008, that is, the fourth quarter of 2007 is defined as the starting quarter of the recession. Why move back by two quarters? “Year-to-year % change" of real GDP calculated at the second quarter of 2008 is comparing Real GDP of the second quarter of 2008 with Real GDP of the second quarter of 2007. The result is representing the average condition of the whole year from the second quarter of 2007 to the second quarter of 2008. This average condition can be approximated as the economic condition of the mid point of this one year period, that is, the fourth quarter of 2007 that is two quarters back from the second quarter of 2008. After entered the recession, “year-to-year % change" moved above 1.0% line at the first quarter of 2010. Moving back two quarters according to Step 3, the third quarter of 2009 is defined as the quarter that the recession had ended, a six quarter long recession.

Let us compare the starting and the ending date of the most recent recession with the determination of The National Bureau of Economic Research. The most recent recession was started at December, 2007 and ended at June, 2009, an 18 month long recession, according to the bureau. The starting point and the duration of the recession thus matches with the three step criteria outlined here. How does the recession look like under the folklore type definition that requires two consecutive negative growth quarters of Real GDP to trigger? “Quarter-to-quarter % change" of Real GDP as used in the folklore type definition plunged to -3.7% at the third quarter of 2008, and continued its negative growth rate of -9.2% at the fourth quarter of 2008 so that a recession was triggered according to the folklore type definition. Let us assume that the recession had started at the third quarter of 2008 instead of the fourth quarter of 2008 to make its duration one quarter longer. The recession had ended at the third quarter of 2009 as the growth rate returned to +0.4%, thus making the recession a four quarter long one. We can see that the starting date of the recession was delayed by more than six months according to the folklore type definition to make the duration of the recession two quarters shorter compared to the determination of The National Bureau of Economic Research and our definition. By studying the recessions from 1947 when the quarterly data of Real GDP have become available, folklore type definition has the tendency to make the durations of recessions substantially shorter than both the determination of The National Bureau of Economic Research and our definition. We consider the folklore type definition of recessions out of touch with the real economy and is a kind of fools' self- comfort by making recessions artificially short.

Once we have an objective but reliable way of defining recessions based on the indicator of “year-to-year % change of Real GDP" as discussed above, forewarning signs of upcoming recessions are searched in Comment 71. It is found that 1.8% line of the indicator serves such a purpose. After a prolonged period of healthy growth with the indicator above 1.8%, if the indicator drops below 1.8%, in the most of cases since 1948 the indicator immediately plunges to or below 1.0%, indicating that a recession is already underway. Even in the cases that the indicator bounces back above 1.8% level after dropping below 1.8%, a rather rare occurrence, the indicator drops into the recession level within six quarters. In the latter case the drop below 1.8% for the indicator serves as a nice forewarning signal of the upcoming recession. The most recent recession belongs to the latter category. The indicator dropped below 1.8% line in the first half of 2007, but subsequently it bounced back above 1.8% only to see it dropped to 1.0% at the second quarter of 2008, indicating the entry to the recession at the fourth quarter of 2007.

After the lengthy review of the indicator of “year-to-year % change of Real GDP", we are now ready to discuss the main topic of this article, why a recession watch needs to be issued here. In Figure 1 posted at the right, the behavior of the indicator from 1984 to the second quarter of 2011 is shown. Black dots are the cases that the indicator is at or above 1.8% and the red dots are the cases when the indicator is below 1.8%. The green line indicates this 1.8% line and the purple line is 1.0% recession border line. The attention of readers is directed to the point labeled with a red letter “A" that is the “year-to-year % change of Real GDP" of the second quarter of 2011, based on the annually revised data published on July 29, 2011 by The Bureau of Economic Analysis. The value of the indicator is 1.6%, below the critical 1.8% line. Does this indicate that a recession is already at hand or the recession is near? We will look into the behavior of Real GDP more closely below.

In Figure 2 graph at the right, quarterly values of Real GDP since 2002 are plotted as black dots, in the conventional logarithmic scale. We should note that the indicator of year-to-year % change of Real GDP at the second quarter of 2011 represents the economic condition of the fourth quarter of 2010, following the rule of “two quarters back" as discussed before. Looking at Real GDP of the fourth quarter of 2010 in Figure 2, the black curve is indeed rounding at that time, indicating the slow down of the growth of Real GDP. The registered value of the indicator, 1.6%, at the second quarter of 2011 is implying this slowed growth of Real GDP. Since the growth rate of Real GDP has slowed further in the first and the second quarters of 2011 as indicated by the black curve in Figure 2, the indicator of year-to-year % change of Real GDP should stay below 1.6%. With this understanding we can play some numbers game next to see what kinds of growth rate of Real GDP we should expect in near future.

Assuming that the value of Real GDP at the second quarter of 2011 will not be revised substantially in coming months, what percentage growth of Real GDP should be for the indicator to stay at 1.8% at the third quarter of 2011? The answer is annualized growth rate of +3.2%. If the indicator stays at 1.8% at the fourth quarter of 2011, the required average growth rate of Real GDP in the third and the fourth quarter of 2011 should be +2.8%. On the other hand we can play the similar game by asking what kind of growth rates should be if the indicator drops to 1.0%, indicating the onset of a recession. If the indicator should drop to 1.0% at the third quarter of 2011, the annualized growth rate of Real GDP should be 0.0% in the third quarter, whereas if the indicator should drop to 1.0% at the fourth quarter of 2011, the average growth rate of the third and the fourth quarters should be +1.2% for Real GDP. With this kinds of view insight, we believe that it is prudent to issue a stronger recession watch signal than that of Comment 81 here.