*Editor's note: This article was originally published on November 29, 2016 by Menzie Chinn here.*

Among the many promises made by President-elect Trump, one was to declare China a currency manipulator on his first day in office. Besides the logistical difficulties of doing so without a Treasury Secretary in place, there are the minor difficulties of what the data indicate (I know, I know, facts seem of little important these days, but what the heck). In addition to the legally defined concerns Brad Setser has raised, I think it is useful to assess China's currency using a commonly used measure of currency misalignment.

Recall, in recent work with Yin-Wong Cheung and Xin Nong, we found that that as of 2011, the yuan was not misaligned using the Penn effect allowing for non-linearity. Now 2011 is a while ago, but we were constrained by the availability of internationally comparable price level data in the Penn World Tables (PWT).

Our findings using World Bank World Development Indicators (WDI) data up to 2014 indicated substantial *overvaluation* (in both 2011 and 2014); since the 2011 estimates of misalignment differ so much between the PWT and WDI datasets, that suggests an alternative check to determine misalignment. I use the Big Mac index as of July 2016, and WDI PPP estimates of per capita income for 2016.

Using a sample of 55 countries (Data [XLSX]), I estimate the following regression:

*r* = **0.630***y* + **0.158***y*2

Adj-R2 = 0.12, SER = 0.273. **bold face** denotes significance at 5% msl, using Huber-White standard errors.

The use of the quadratic specification is justified in Cheung, Chinn and Nong (2016), following Hassan (2016, *JIE*).

In Figure 1, I show the scatter plot for July 2016, and the curve fitted through the data, forcing the intercept to go through the origin, thereby forcing the US dollar to have no misalignment against itself.

Figure 1: Log relative dollar price of Big Mac against dollar price of US Big Mac (July 2016) versus log relative per capita income in PPP terms (2016 estimates); regression fit from quadratic specification (red dots), and 90% prediction interval (gray dots).* Source: **Economist**, World Bank World Development Indicators, and author's calculations.*

The implied undervaluation of the Chinese yuan is 3.7% (log terms) as of July 2016. Of course, the estimate is surrounded by high degrees of uncertainty (as evidenced by the prediction interval) (see Cheung et al. (2007)).

Since July 2016, the yuan has continued to drop against the US dollar, in nominal bilateral terms (see Figure 2):

Figure 2: Log USD/CNY exchange rate (blue), and log trade weighted real value of CNY against broad basket of currencies, both normalized to 2015M07 = 0 (dashed vertical line). Solid vertical line at July 2016 (Economist Big Mac date.) Up denotes increase in value of CNY for both series. *Source: Federal Reserve Board, **BIS**, and author's calculations.*

However, the drop is small - 2.3% through November, and has actually appreciated in real trade weighted terms by 0.6% through October. (Figure 2 also highlights the fact that the Chinese yuan actually appreciated along with the dollar before the recent depreciation - so much so that the real trade weighted yuan remains above mid-2014 values.)

Hence, using a price metric (admittedly subject to large estimation error), it hardly seems easy to argue the Chinese currency is currently significantly undervalued.

For more on the various frameworks to assess currency misalignment, see this post. Brad Setser discusses "currency manipulation" in the context of American legislation regarding the determination of offending parties.