A straight line going nowhere fast, despite tweets and likes

Ken B., another Australian reader, wasn't too proud of this effort, apparently excerpted from an HSBC report by the Sydney Morning Herald (link):

Ken: If you plot ranking by ranking it magically turns into a straight line.

***

There are a few other annoyances. Gridlines, data labels, double-edged arrow, bars all based on the same data, which can easily be conveyed with a ranked table. In fact, just turn the chart 90 degrees clockwise, get rid of everything else except the names of countries, and you have a much more readable figure.

The completely unnecessary legend is an Excel special. If only one data series is plotted, it should be automatic to suppress the legend.

The three-letter acronyms for different currencies is a futile educational lesson kind of like plotting geographical data on maps (in many cases). For most readers, the message of the chart does not require knowing the names of the currencies, nor their acronyms. For those who care about acronyms, say currency traders, they most likely already know those letters.

***

Just like I don't understand how we can define "over-rated" or "under-rated" restaurants (see this post and this), I also don't understand how we can define "over-valued" or "under-valued" currencies given the impossiblity of knowing the "true value" of any currency. 

***

I just had to point your attention to the fact that 123 people tweeted this article, and 221 liked this item on Facebook. And these actions form part of the so-called Big Data revolution.

Comments

[derek]

Check out this this food-wine-pairing chart. Isn't it a-maze-ing?

I'd do it as a crosstab instead: food along one side, wine down the other, and spots at the intersection. Do a bit of Bertin-style reordering before you send it out, to get the pattern of spots as close to a diagonal band as you can.

[conchis]

I also don't understand how we can define "over-valued" or "under-valued" currencies

This is actually a pretty standard economic concept, which you can read all about if you google 'purchasing power parity'. The basic idea is that if exchange rates were 'properly' valued then the cost of a given bundle of goods in, say, Australia, converted to US dollars, would be the same as the cost of the same bundle of goods in the USA. If the cost of the bundle is higher in Australia, then the Australian dollar is overvalued, and vice versa. (The Economist's famous 'Big Mac Index' conducts this exercise with Big Mac prices around the world.)

None of which excuses the graph, particularly when you could use the actual numbers behind the rankings (e.g. ratios vs. 'neutral').

[Kaiser]

conchis: But what are all the assumptions behind the "purchasing power parity" concept? And where is the proof that PPP is a valid theory? I'm not talking about theoretical proofs; I'm talking about data. I'm raising the question of whether "fair value" is an intrinsic value or just an opinion.

[conchis]

Kaiser: There's probably a bit too much complexity behind PPP for me to address this adequately in a blog comment, but a few points seem worth making:

(1) There are lots of measurement issues here, but most of them are well known issues that apply equally to the construction of any price index such a the CPI. (How to define the bundle of goods, adjust for quality, etc. etc.) They're not trivial, but they're also not so severe as to make the concept meaningless, or the exercise pointless.

(2) The 'fair value' concept is somewhat arbitrary. Really all you're doing here is picking a base currency, and saying other currencies are over or undervalued relative to that currency. The ranking of most overvalued to undervalued, and the relative ratios would be preserved for any choice of base, but the scale is arbitrary.

(3) The degree to which PPP is a valid theory depends on what you're trying to use it for, but let's assume for that we're trying to use it to predict future exchange rate movements. The idea is that large relative over- or under-valuations present arbitrage opportunities because you can buy stuff cheap in one country, and sell it for more overseas (and the process of doing this will gradually cause the relative exchange rates to equalise). Of course, there are lots of reasons why this might not happen in practice (transactions costs, trade-barriers, and governments that fix their exchange rates being some). Nonetheless (and while this is not my area of expertise) I understand that the general view of the evidence is that while PPP is not very useful at predicting short-term exchange rate movements, it is useful at predicting longer term exchange rate movements (potentially over a 1-3 year horizon), once you take account of some of the additional factors above.

[Kaiser]

conchis: thanks for the detailed explanations. I was thinking about point (2) when I compared this exercise to identifying "under-rated" and "over-rated" restaurants. Some kind of base has to be defined, which makes the exercise completely subjective (with scaffolding to provide a facade of rigor). Of course, many similar things (quality of teaching, quality of wine, etc.) are also not measurable in a similar way and we still want to come up with subjective measures. What annoys me is when the people who use concepts like "over-valued" currencies (i.e. CNBC talking heads) speak as if they are the holder of the "truth".

[conchis]

Kaiser: Fair enough, although I wouldn't necessarily class this as 'completely subjective'. It's just recognising that in a value-neutral sense, saying 'the yuan is overvalued vs the dollar', is the same thing as saying the 'the dollar is undervalued vs. the yuan'. From the perspective of predicting future exchange rate movements, it doesn't matter which way you think about this, and the anticipated correction is also the same whether you decide to call it an appreciation of the yuan against the dollar or a depreciation of the dollar vs the yuan.

That said, the context in which the CNN talking heads use the word 'undervalued' in relation to the yuan is a bit different, and it's actually not entirely illegitimate. Where governments are actively managing the value of their currencies rather than letting them be determined by the market, then it's possible to define a more objective concept over/undervaluation as the difference between actual and (hypothetical) market-determined exchange rates.

The difficulty here obviously lies in figuring out what the market exchange rate should be, absent 'manipulation'. PPP type measures can legitimately be used for this sort of thing, but you do need to adjust them for factors other than manipulation that might legitimately drive price differentials between different countries. For an example of a very simple adjustment along these lines (which argues that the yuan wasn't undervalued in mid 2011), see e.g. this Economist blog post.

[conchis]

Whoops, the first para would make slightly more sense if it read: "saying 'the yuan is undervalued vs the dollar', is the same thing as saying the 'the dollar is overvalued vs. the yuan'".

[Kaiser]

conchis: your example is a good one that illustrates my deep discomfort with the concept. When CNN uses the phrase "Yuan is over-valued against the dollar" as opposed to "the dollar is under-valued against the yuan", there is an unspoken value judgment, in each case the "base" currency is interpreted as "fairly valued". One can nitpick to say the statement makes no such claims but there is no doubt that most readers interpret it as such.
Besides, if we strictly interpret the comparison as relative, then when someone says "the yuan is over-valued against the dollar", is the course of action supposed to be the yuan should depreciate or that the dollar should appreciate? Moving from an insight to an action requires making an unwarranted assumption that one of these two values are "correct".

[conchis]

Hmm... my first attempt to post this comment appeared to have been eaten by the system. Take 2.

I agree that the CNN talk is value-laden, but I don't it's arbitrary in the way you're claiming. Whether or not CNN decides to say 'the yuan is undervalued vs dollar' or 'the dollar is overvalued vs. the yuan', the implication is the same: assuming it is true, this provides evidence that the Chinese Central Bank has managed move the actual exchange rate between the two currencies away from what adjusted PPP would predict. The only value judgment involved is that this deviation from PPP is a bad thing.

Of course, you can contest this value judgement, but you can't contest it by arguing that CNN's chosen base is arbitrary. The only relevant fact is whether there's a differential; it doesn't matter what base we use to denominate that differential, the differential will still be there.

if we strictly interpret the comparison as relative, then when someone says "the yuan is over-valued against the dollar", is the course of action supposed to be the yuan should depreciate or that the dollar should appreciate?

Clarity fail on my part apparently. My whole point was that depreciation of the yuan vs the dollar and appreciation of the dollar vs. the yuan are, by definition, exactly the same thing, so no unwarranted assumptions need enter into it.

To take yet another example of how the choice of base doesn't have any practical implications, assume we're just interested in identifying arbitrage opportunities. Again, for these purposes, all we care about is identifying price differentials between countries. If stuff is cheaper in country A than country B then (subject to transaction costs) we can make a profit by buying stuff in A, and selling it for more in B. It makes no difference whether we say that A's currency is undervalued vs. B, or B's is overvalued vs. A.