Thursday, February 05, 2015

Exchange Rate Forecasts: Could Your Guess Be As Good As Nihon Soken’s?

Having too much time on my hands this morning, I had the choice of ripping into yet another Noah Smith Japan op-ed (not that I currently have a strong objection to the core piece of conventional wisdom itself that he festoons, mind you) or satisfying my curiosity as too just how good or not experts were at forecasting exchange rates. I decided to go with the latter, since I thought that it would be more useful (to you, not me; I don’t play the markets). For that, I turned to the Japan Research Institute, Limited, arguably the most reputable economic think tank in Japan, and its exchange rate forecasts during this decade (the regular ten-year one that began in 2011, not the 12-year one that began in 2010), which it posts on its website here. Specifically, I looked at its January yen/dollar and yen/euro forecasts for fourth quarters and compared them with the actual outcomes. It was not pretty.


2011 (10-12)
2012 (Oct-Dec)
2013 (Oct-Dec)
2014 (Oct-Dec)
2015 (Oct-Dec)

2011.01
2012.01
2012.01
2013.01
2013.01
2014.01
2014.01
2015.01
2015.01
2016.01
Y/$
range
 80-92
75.35-79.54
77-82
 77.8-86.8
82-89
   96.6-105.4
100-105
105.19-121.85
116-127
?
Y/$
average
86
77.35
80
81.2
85
100.4
105
114.46
123
?
Y/€
range
102-112
99.41-111.6
 92-104
99.8-114.7
100-112
131.1-145.7
136-148
134.14-149.80
138-150
?
Y/€
average
107
104.2
98
105.3
106
136.7
142
142.9
144
?

In the above chart, I’ve marked out in red the figures in the years in which there was no overlap between the forecast and the outcome ranges for the exchange rates. Four out of eight turned out to be complete duds. The yen/euro forecast for 2012 4Q didn’t turn out that well either, which I’ve marked out in yellow. Had you taken the fourth quarter of the previous year, take a five yen band, and expanded it a few yen both ways depending on if and how uncertain you were, and you wouldn’t have done much worse if at all than the JRI experts.

Incidentally, when I entered the 123 yen average forecast for 2015 Q4 in the Excel sheet that I used, it came out in red and refused to change to black, so I had to devise a workaround. Was my Office software trying to tell me something?

2 comments:

Anonymous said...

It's amazing how lacking erudite foreign analysts/observers of Japan are that Noah Smith can even position himself as some kind of authority on the country and its business and macro situation
Some of his claims on shareholder rights are plain wrong... not just idiotic but based on factually incorrect premises

Jun Okumura said...

Anonymous:

I keep running into perfectly capable economists who do reputable work on Japan without professionally functional Japanese language skills. Indeed there is much useful data readily available that can and do benefit from dissection by experts free of the conventional wisdom and skin in the game—say, appointments to government advisory councils, administrative agencies, and, more recently, corporate boards, government and corporate grants, etc. But I have also seen these people make fools of themselves as they venture beyond their realm of competence. And that is only to be expected if your prime sources of information source are limited to English-language publications and English-speaking locals.

I plead guilty to the sin myself. I just got called out when I made a crack about Australian politics. But at least I had enough command of the English language to have the wits to qualify my comment with “unlikely.”