2010-09-22

The USDX and US Real GDP: Has the US been in recession since December 2002?

I'm a big fan of tying down absolute values. For me, it is no reason to cheer GDP growth in the US if the US Dollar has dropped by enough to wipe out relative gain internationally. This is because I see currency value as an international way of valuing. Let me give you an example.

Let's pretend that Country A has a GDP of 100 while country B has a GDP of 100, and that both countries have parity in currency value. If the currency remains locked together, then any increase in GDP in either country can be compared to one another. So country A, for example, may increase their GDP to 102, while country B increases to 101. Thus country A has the better GDP.

But what happens if you take currency changes into account? Let's say the while Country A's GDP has increased as per above, what would happen if their currency had dropped during that same period? In order to make an accurate comparison between country A & B, this change of currency must be taken into account. So the equation would look like this:

(Change in Real GDP) x (Change in Currency Value) = (Actual GDP Change)

And here's a table that explains how the difference might work in practice:



So while Middle Earth's GDP has increased faster than Narnia, the devaluation of Middle Earth's currency ensures that, in comparison with Narnia, it has actually experienced economic decline while Narnia has grown.

This sort of activity is naturally the case when GDP is compared internationally. Of course this has been augmented by the use of "Purchasing Power Parity" to ensure that differences in GDP comparison are more realistic.

Anyway, a while back I began to add the uncertainty of currency value into the analysis of US GDP. The USDX is the index best used for comparing the US Dollar's value compared to other currencies. So into the spreadsheet I went, multiplying real GDP by the 12 month USDX average from 1981 Q4 until now. This is the graph that results:



Historically you can see the effects of the Plaza Accord in the mid 1980s, followed by the mid 90s expansion and the tail-off since the early 2000s. According to this graph, US GDP peaked in 1986 Q1 (170.89), bottomed out in 1991 Q1 (134.67), peaked again in 2003 Q3 (239.54), and bottomed out again in 2009 Q3 (189.34). Since then it has risen only slightly (currently 191.91).

Of course, population issues need to be added into the equation, which is where GDP per capita comes into play. Here is Real GDP per capita adjusted by the USDX in the same way:



This obviously follows the same trend as the graph before, but you can see that the effects of a higher population have led to even less of a growth in GDP per capita from 1981 until today. According to these figures, Real GDP per capita, adjusted by the USDX, is currently 143.03, which implies only a 43.03% gain since 1981. In fact the current situation is lower even than the mid 1980s peak that was brought down by the Plaza Accord. This graph shows that a peak was reached in 1986 Q1 (164.52), which bottomed out in 1993 Q1 (121.81), peaked again in 2002 Q4 (190.29), and bottomed out again in 2009 Q3 (142.02).

What is notable from these last two graphs is that the last peak goes back eight years. In other words, the recent downturn, extreme in nature, has been cushioned by a rise in the dollar. Conversely, the "mild" recession of the early 2000s has been made worse by a long term decline in the value of the US dollar. These two assertions fit into the USDX history.

It would be fair to say that if Real GDP and Real GDP per capita (both adjusted by the USDX for international - absolute - comparison) did advance and decline in the years shown on that graph, then surely that would've been felt at a national level. Without really giving any links, consider this:
  • US GDP per capita adjusted by the USDX grew in the early 80s. This implies that between 1981 and 1986, there was a period which could be described as "a good economy". That certainly fits.
  • US GDP per capita adjusted by the USDX declined between 1986 and 1993. This implies that the period that could be described as "a bad economy". That fits too.
  • US GDP per capita adjusted by the USDX then grew between 1993 and 2002. This implies that the period could be described as "a good economy". Despite a few issues (notably the 2001 recession), that fits as well.
  • US GDP per capita adjusted by the USDX has then declined between 2002 and 2009. This implies "a bad economy" which fits too, though the property bubble helped smooth out the 2005-2007 period. Certainly this period had below-average GDP growth.
Though this issue has been on my mind for some months now, I am not fully convinced that my logic or argument is correct, which is why I am writing a disclaimer here. Part of my thinking is that a 1% growth in the value of a currency is the same as a 1% growth in the value of GDP, which may be in error. Certainly there is clear evidence that a GDP decline results in higher unemployment while a drop in currency value may not result in such an event. It is also affected by my opinion that common currencies like the Euro should be a logical economic goal, or at least have fixed currency rates governed by a currency board and operating by international standards.

Anyway, tell me where I'm wrong on this... if I am wrong.

2 comments:

Jacob said...

you are right, but also wrong.
you are right in the way that compared to other economies/foreigners, americans lost purchasing power. if you look at it this way, then GDP is a zero-sum game amongst all people, which it isn't.
say, the american income goes from twice the average, to only 50% above average. you'd say that they lost 33% PP. but what if world-output doubled at the same time? then his PP actually went up 50%!
of course, this is only my humble opinion.

Neil Cameron (One Salient Oversight) said...

I'm not comparing US GDP to international GDP, but adjusting it according to currency rates, which are not GDP dependent (at least directly)

I'm sure that a counter graph can be made to show the growth of the currencies that make up the USDX fared in relation to the US Dollar.