Forecasting lessons from an artist

Forecasting lessons from an artist

I logged into my email Friday night after three days with no Internet to write my Monday column on my economic impressions of Cuba. I had found out that my undergraduate thesis adviser, Bob Shiller, had won the Nobel Prize the last time I connected to the net, but the news turned out to be much grimmer this time.

The first email I read solemnly stated that Ali İhsan Gelberi, head of economic research at Garanti Bank, had passed away. Although I never got to meet him face-to-face, we thought similarly on many matters and had pleasant exchanges on Twitter. 

While he was well-known among economists and finance professionals, Gelberi became famous after predicting correctly at the end of December that credit ratings agency Moody’s would upgrade Turkey the same day Prime Minister Recep Tayyip Erdoğan would meet with U.S. President Barrack Obama.

Newspapers labeled Gelberi as a soothsayer ‎after he hit the bull’s eye on May 16, and so he explained in a blog post that his prediction was based solely on analysis. I asked to publish that article on my own blog at Nouriel Roubini’s Economonitor, and he kindly wrote a short piece in English for me. I never got to use that, and so I would like to share parts of it now:

“For an upgrade, it is much more important to convince U.S. officials than rating agency analysts. But to make that demand, Turkey would need to have solved its main disagreements with the U.S. The two main problem areas between the two countries were regarding Turkey’s relations with Israel and the central Iraq government. I correctly forecasted that Turkey would approach the U.S. stance on these matters. 

“During my time at the Turkish Treasury, I learned that leaders do not discuss anything other than major issues during face-to-face meetings. Since Turkey’s rating is too small a matter to be decided in a meeting between Turkish and American leaders, it would have to be sorted out before or after.

“The vital part of my prediction was the timing of the upgrade. Based on the analysis above, I could have concluded that it would come a month or so before or after the Erdoğan-Obama meeting. But I also thought that the Turkish government would specifically ask the U.S. for the upgrade to be announced at the same time as the meeting to convey a strong domestic message: ‘The rating agencies were not handing us our well-deserved upgrade. Therefore, we flexed our muscles and used our bargaining power with the U.S. to grab it.’

“It is important to note that announcing the upgrade exactly during the Erdoğan-Obama meeting is to declare the world that it is political. Moody’s would definitely not have preferred something like that, which would have tarnished its already-weakened credibility. They were evidently forced by the U.S. government.”

Despite my fancy econometrics, I could have never made that prediction. That’s why forecasting is, at its essence, more of an art than science. I will remember Gelberi the artist as much as Gelberi the economist.