How to return to the equilibrium set by the adjusted Big Mac Index?

External question — Economics Stack Exchange Author: heinzlee Original post: https://economics.stackexchange.com/questions/60967 License: CC BY-SA 4.0 — https://creativecommons.org/licenses/by-sa/4.0/ Adaptation: HTML converted to plain text; contact email addresses removed. I read the economist article on the adjusted Big Mac index. Correct me if I am wrong, the idea is to compare the ratio of price divided by a linear function of GDP per capita fitted on price-GDP per capita pairs. But what exactly does this index estimate? The long term equilibrium exchange rate? What could people do to profitably force the exchange rate to return to this equilibrium? In the case of USD-CNY exchange rate, the adjusted and raw index gives 'fair' estimates two times apart. Which one should I trust? I believe a fairer estimate might be to to subtract the price of a Big Mac by $\beta\cdot \text{GDP per capita}$ where $\beta$ is the fitted constant in $Price=\alpha + \beta\cdot\text{GDP per capita} + \epsilon$ . This way the statistic measures the price people pay for the same tradable goods that go into a Big Mac in different countries. If MacDonald's overpay in a country, traders in that country buy more of that tradable good and vice versa. Change my mind.

External answer — Economics Stack Exchange Author: 1muflon1 Original post: https://economics.stackexchange.com/a/60968 License: CC BY-SA 4.0 — https://creativecommons.org/licenses/by-sa/4.0/ Adaptation: HTML converted to plain text; contact email addresses removed. Is the adjusted Big Mac Index considered serious economics? Yes, although it started as a playful idea of the Economist (UK based non-academic newspaper focusing on economic/financial news), there is an actual economic logic behind it and there are even academic research papers published that feature this index (see references below). More specifically the original Big Mac index is based on economic concepts of law of one price and purchasing power parity (which imply that exchange rates, conditional on things like transportation costs, tend to equalize prices) which are serious economic concepts and the GDP adjusted Big Mac index is based on the concept of Penn effect (the positive cross-sectional empirical association between economies’ composite price level and real per capita GDP, expressed in a given reference currency <- this empirical observation is also explained by the Balassa-Samuelson theorem). Correct me if I am wrong, the idea is to compare the ratio of price divided by a linear function of GDP per capita fitted on price-GDP per capital duals. It does sound mostly correct. GDP adjusted Big Mac Index is simply the ratio of price of Big Mac in one currency, divided by estimated price of Big Mac price in other country, i.e. for an index between US and EU it would be $aBMI =\frac{P_{\\\$}}{{E[P_E]}}$ . The expected prices can be calculated using linear regression $P_E= \alpha +\beta GDP +\epsilon$ ( Brien & Vargas 2015 ). But what exactly does this index estimate? The long term equilibrium exchange rate? The index is used to see if countries' currencies are under or overvalued. A misevaluation of currency can be simply calculated as $m=aBMI -1$ since aBMI of 1 indicates that there is no misvaluation. Do economics academics treat this index seriously? Yes, as already mentioned above, although the idea is playful, it is a serious idea that is based on real research of top international trade economists. This doesn't mean its necessarily the only way how to estimate whether currency is over/under-valuated. There are way more fancier statistical models that can be more accurate. The utility of Big Mac index is that it can be computed from very little publicly available data, whereas other fancier methods are far more data intensive and elaborate. A lot of economic data is published with a severe lag, even for example true GDP data area published with about 5 year lag (the headline numbers are forecasts and estimates that are subject to multiple revisions over the years). And GDP is one of the statistics that is published most promptly compared to others. As a result the utility of raw or adjusted Big Mac Index is that you can use it in real time (especially the raw one) to have a crude measure of over/under valuation. In the case of USD-CNY exchange rate, the adjusted and raw index give 'fair' estimates two times apart. Which one should I trust? You seem to misunderstood things. It is not like one of these numbers is correct and another is wrong. The USD-CNY exchange rate is the exchange rate you can use now to buy Yens. The Big Mac Index, is a measure of whether this price is currently over or under valuated compared to predicted equilibrium price. However, deviation from the long term equilibrium price doesn't mean the exchange price is unfair (whatever that is even supposed to mean in this context). Due to temporary shocks economic equilibria are constantly being perturbed. An long-term equilibrium price is simply the price that price will tend to converge over time, and it would also be equal to that price if there never would be any shocks or changes to the economic system, but real life economy is constantly perturbed by various exogenous shocks, which would lead to price deviations, even if these shocks are not structural and hence the long term equilibrium price is still the same. This doesn't make the short-term price incorrect. Personally, I believe a fairer estimate might be to to subtract the price of a Big Mac by β⋅GDP per capita where β is the fitted constant in Price=α+β⋅GDP per capita+ϵ. This way the statistic measures the price people pay for the same tradable goods that go into a Big Mac in different countries. If MacDonald's overpay in a country, traders in that country buy more of that tradable good and vice versa. Change my mind. This doesn't make any sense prima facie. It doesn't measure what you claim it measures. I am not sure where you got the idea that $P_{\\\$}-\beta GDPc$ measures prices of tradable goods. You are just throwing random numbers together. That is like if I would say that distance is measured by subtracting weight from Hight. That is total non sequitur.