Saturday, June 28, 2014

Caveat Emptor

Regular readers of this blog and of my previous ramblings know that I am a skeptic when it comes to the reliability of government statistics.  Such figures are important to investors trying to understand the true state of the economy in which they are investing and, unfortunately, today are so manipulated (even by the government's own admission) as to be increasingly useless.

In the September 2013 post to this blog, "Lies, Damn Lies and Statistics", I took a look at the United States' monthly jobs report and noted how seasonal adjustments applied to the figures have gotten bigger over time and how the adjustment is almost always larger than the underlying number itself.  When the seasonal adjustment is stripped away and the underlying figures are smoothed, the health of the jobs economy can be quite different from what is reported as truth each month.  Similarly, while penning the quarterly CIO Letter at Artio Global Investors/Julius Baer, the figure that bothered me the most was the CPI, the Consumer Price Index, which I argued was no longer a price index with any relevance to the pocketbook of consumers.  Changes in the way the index has been constructed over time (including the 40% weight given to housing, of which 3/4 is based on a survey question asking homeowners what they think they could rent their house for) leaves the current release unrecognizable and not comparable to historic readings.

Today, though, I'd like to talk about GDP, Gross Domestic Product, or the value of goods and services produced within the borders of a given country.  And for this, we will look overseas.  GDP per capita is often used as a measure of a society's standard of living and changes in GDP are looked at to measure the health of an economy.  Therefore, you might agree, it would be a useful number for investors to know.

As reported by the Wall Street Journal on June 9th, Italy, the UK and Ireland will be changing the way they calculate GDP.  Going forward, they will included the "value" of the illegal economy -- i.e. drugs, prostitution, etc..  As a result, GDP will be larger, not because the economy has improved, but rather because the illicit portions of it will now be counted.  Debt to GDP figures will depict greater economic "health" as the denominator in the equation is increased.  In other words, links to the past will change.

Granted, one can argue whether legal or illegal, such activities are truly a part of the economy and their inclusion is overdue.  Fair enough.  However, when measuring illegal activity, assumptions have to be made.  Lots of assumptions - many more than in the reported economy.  And to the extent the underlying activities are large relative to the size of the economy, the assumptions used can distort the true state of being.  So let's look at how these numbers will be computed and adjusted.

Drugs

In the UK, the trade in six substances will be added to GDP - heroin, powder cocaine, crack, ecstasy, amphetamines and marijuana.  And here's where the math gets interesting.  According to a UK crime survey, the number of heroin users climbed to 38,000 in 2009 from 36,000 in 2003 (assumption #1).  The purity of the heroin has climbed to 43% from 33%, yet its price has fallen by 31% (assumptions #2 and #3).  So to measure the value of heroin in GDP, you take the base year (2003) sales of GBP 1.349 million and multiply by the increase in users (38,000 divided by 36,000) and then by the increase in price to get a base-level value of product sold.  Then you further multiply by the 2003 purity level and divide by the 2009 purity level to get the total value of heroin consumed (the purity adjustment is the hedonic, or quality adjustment, so prevalent in many economic statistics compiled today).

GBP 1.349 * (2009 heroin users/2003 heroin users) * price * (2003 purity/2009 purity)

You are not done yet.  Now you have to account for imported heroin and subtract that out of the total number as GDP is only value produced within a country's borders.  So based on another set of calculations using UK Border agency seizures and a UN estimate of the imported price level, this figure is calculated.  The total value of heroin to the UK economy is thus "exactly" GBP 536 million.

The value of the largest drug, crack, weighs in at GBP 2.9 billion - or as the Wall Street Journal points out, more than the GBP 2.5 billion in profits from Barclays, the nations largest bank and employer of approximately 60,000 Britons.

Prostitution

To measure prostitution, the measurements get a little more complex.  A 2004 survey estimated there were 58,000 prostitutes at work in the UK.  Since then, the population of men in the country older than 16 years has increased by 5%, so it is assumed the number of prostitutes has kept pace.  Using data from the Netherlands, it is further assumed each prostitute sees 25 clients each week and that in 2004 the price was GBP 55 per visit.  Lap dance prices in the country have climbed by 22% since then (and we know this how?), so it is assumed so has the price received by prostitutes.  These figures must then be adjusted by the purchases prostitutes make in the course of doing business as GDP only measures final sales, not intermediate sales (for example, if the Brown Forman company sells a bottle of Jack Daniels to my local liquor store and the liquor store then sells that bottle to me, only my purchase is included in calculating the value of GDP).  So, when a prostitute buys a condom it is not included in the value of GDP.  However, when a private person buys a condom, it is.  So total condom sales must be adjusted downward based on an assumption between professional and personal useage.  The same holds true for hotel rooms and any other consumables used in the provision of the prostitute's service.

Is there a point to all this?

The point I am making is that economic statistics are important and their estimation is hard enough.  When multiple assumptions are then mixed in, and the size of the underlying segment changed is large relative to the whole, the value of the resulting number has to be questioned - especially when making comparisons with previous measures of the same.  Unfortunately, many economic series including jobs, inflation and GDP are now all suspect.

Happy July 4th