Oops

How reliable are data from the National Bank of Kazakhstan?

In some ways, Kazakhstan is lucky to have the central bank that it does.  The National Bank (NBK) is not deliriously reckless, as is the Central Bank of Iran.  Neither is it paranoid about rising prices – unlike the Reserve Bank of New Zealand, where a 1989 law bound the governor to hit the rate of inflation negotiated with the finance minister, or lose his job. 

The NBK handles policy with care; if only it would do the same for statistics.  Exhibit A of its neglect is its estimate of the exchange rate.

First, some background.  Most people are familiar with the nominal exchange rate, since this is reported by the daily press.  For example, yesterday one could sell 153.6 tenge in exchange for a United States dollar.  But this exchange rate is not truly important.  Most of us don’t care about the number of tenge or dollars that we hold; instead, we care about the goods and services that we can buy with them.  The purchasing power of the tenge is expressed by the “real exchange rate”, which adjusts the nominal rate for prices in Kazakhstan and abroad. 

As it is usually defined, a real exchange rate of 2 (say) implies that a foreign bundle of goods costs twice as much as a similar bundle in Kazakhstan.  A rise in the rate to 3 would suggest that the foreign bundle now costs three times as much as ours; that is, the tenge is losing its purchasing power over foreign goods.  That’s depreciation.  We now must sell three domestic bundles in order to buy a foreign bundle; before the depreciation, we had to sell only two domestic bundles.   

We could express the real exchange rate with respect to any other country.  But since we trade with many countries, it makes sense to take into account all of these exchange rates.  We can do this by calculating the weighted sum of all the bilateral rates, where each weight is that country’s share of our total volume of trade.  For example, suppose that we have two trading partners: Country A, which accounts for 60% of our total trade; and Country B, which accounts for 40%.  Suppose that our real exchange rate is 2 with respect to A and 3 with respect to B.  Then our weighted exchange rate is .6*2 + .4*3, or 2.4.

Almost half of our trade is with Russia, China and Italy, in that order.  But we also trade with 16 or 17 other countries that each claim more than 1% of our current account.  In recent years, Uzbekistan has generated 1.6% of our trade, usually exporting fruit, vegetables and textiles.  In terms of its trading weight, Uzbekistan is between the United Kingdom and Poland.  From 2003 through 2012, it was our 14^th largest trading partner.  Its exports to us increased last year by more than $1 million, reported the weekly Kazakhstani newspaper Kapital.

Yet another “coding error”?

Inexplicably, the National Bank excludes Uzbekistan from its estimates of the weighted real exchange rate for the entire 10-year period.  That was discovered by a KIMEP graduate student in economics, Kairat Beisenov.  To double-check on the Bank’s work, he used data from the official source – the Customs Control Committee of the Ministry of Finance.  Uzbekistan doesn’t show up anywhere in the Bank estimates for the top 24 trading partners, although it is in the control committee’s dataset.

One may think this a small error since Uzbekistan accounts for less than 2% of our trade.  The omission matters for three reasons.  Currency traders, as well as import and export dealers, find that their profits are sensitive to small changes in the exchange rate.  Also, we need Uzbekistani data to answer such questions as the impact of Kazakhstan’s customs union (formed with Russia and Belarus in 2010) on trade in Central Asia.  Finally, and most important, the Bank’s failure to discuss anywhere its reasons for excluding the Uzbekistani data for 10 years, raises questions about its diligence in verifying its own estimates.

This is not the first time that the Bank’s calculations have been called into question.  As Beisenov notes, Tengrinews in 2012 noted a $3.6 billion difference between the trade estimates of Kazakhstan and China in 2011.  That was 3.1% of Kazakhstan’s entire trade that year.  Somebody needs a new abacus.  --Leon Taylor, tayloralmaty@gmail.com

 

 

References

Beisenov, Kairat.  Estimation and impact of the real effective exchange rate on the goods market in Kazakhstan.  KIMEP University manuscript.  August 2013.

Bernanke, Ben S., and Frederic S. Mishkin.  Inflation targeting: A new framework for monetary policy?  National Bureau of Economic Research Working Paper #5893.  1997.  Online.  Briefly discusses the Reserve Bank of New Zealand.

Gayfutdynova, Venera.  Uzbekistan uvelychyt eksport v RK bolee chem na $1 million.  Kapital.  July 25, 2013.   

 

Tengrinews.  Raznitsa tamojennoy statistike kazahstana-kitaya sostavila 3.6 milliarda dollarov.   http://tengrinews.kz/kazakhstan_news/  June 30, 2012.