Calculating nominal GDP

I would like to understand one detail of how nominal GDP of a country in a given year is calculated. It seems to me that the GDP in each country is first calculated in the national currency and then is converted to dollars. Is this correct?

If yes, how can one make this conversion if a currency can weaken of strengthen itself significantly with respect to dollar within a year?

Here is the list of such GDP's per country per year in US dollars: http://en.wikipedia.org/wiki/List_of_countries_by_GDP_(nominal)

For example take Russia, GDP in 2013 was 2,096,774 million dollars and in 2014 it is 2,057,000 million dollars.

Now, we know that the rouble weakened significantly in 2014, (almost twice). So it looks like Russian GDP in dollars should have dropped significantly as well. But it had not done so. Why?

There is no need to complicate things here - just try to find annual exchange rate used by IMF. Worldbank's data are here. LCU stands for Local Currency Units
You may see a drop of Rub/$ rate from 31.84 to 38.38. (Drop of Ruble, of course.) National statistics provides input figures for international accounts. According to Rosstat, Russian nominal GDP grew in the Y2014: • Y2013 - 66,190 bil. Rub, • Y2014 - 71,406 bil. Rub. Using NGDP dollar evaluation from your post, annual exchange rate for Y2013 is 31.5 Rub/$; in Y2014 rate is 34.7 Rub/$, which makes some sense. But indeed Rub/$ rate was somewhat closer to 38.5 (see here). So, IMF figure for the Y2014 from your wiki reference seems to be more correct: bil. \$1,857,461 for Russian NGDP. (do not forget about Crimea correction to appreciate the difference).
So, in fact Russian nominal GDP did drop significantly in current dollar terms in the Y2014.
Imho, PPP is not applicable for your question as Russian GDP by PPP is of different magnitude.

Institutions may differ in their exact approach, but GDP statistics probably use a moving average to calculate currency exchange rates to avoid rapid fluctuation. The rubel exchange rate only started falling at the very end of 2014, see this graph. Hence its effect was limited on the 2014 GDP. (There is still a 10% drop, which is huge!) The full effects of the current exchange rate will show in 2015 GDP data.

• Thank you for the answer and for the link to Yahoo finance. I wonder still if a precise recipe for this calculation is written somewhere. Maybe something like "average annual exchange rate" is used...? – aglearner May 27 '15 at 7:58
• I am not sure how they do it, but you might find your answer here: imf.org/external/pubs/ft/weo/faq.htm#q1g – Giskard May 27 '15 at 11:09

The answer is mostly tradable goods priced in dollars and purchasing power parity (PPP). PPP:

If goods were perfectly tradable across borders, with no trade barriers or transactions costs, then there would be no reason for prices to differ. This gives rise to the idea of purchasing power parity, a theory of exchange-rate adjustment based on the law of one price.

If the same good sells for one hundred dollars in the United States and one hundred euros in Europe, then according to the law of one price the exchange rate between dollars and euros ought to be one. The theory of purchasing power parity is that this relationship holds for an overall market basket of goods and services.

Empirical tests tend to show only a weak tendency for exchange rates to move in the direction of purchasing power parity. This means that cross-border trade is not nearly friction free. The failure of purchasing power parity to hold, except perhaps in the long run, indicates that transportation costs, language-translation costs, and other factors limit the integration of global markets.

Because PPP doesn't seem to hold always and everywhere, people who wish to compare nominal or real GDP across countries in a way that avoids the nuisance variation of transitory fluctuations in exchange rates report PPP NGDP and PPP GDP rather than RGDP and NGDP at exchange rates. This amounts to pricing tradable good at their prices in current exchange rates but correcting for that fact that non-tradable goods may be too cheap or expensive when converted at market exchange rates. The purchasing power parity Wikipedia page has some nice example of how this works with the price of an iPad and a Big Mac.

In your case of Russia, because the exports of Russia are mostly oil priced in dollars, even as the exchange rate declines the value of their oil is worth more rubles, and so that should not decline much (except through the falling price of oil which probably is major reason why the ruble declined in the first place). The non-tradable sector similarly wouldn't be fully affected by the depreciation, although in this case through a different mechanism, the PPP correction. ...