Ståle Johansen

𝐇𝐨𝐰 𝐰𝐞 𝐮𝐬𝐞 𝐀𝐈 𝐭𝐨 𝐠𝐞𝐧𝐞𝐫𝐚𝐭𝐞 𝐜𝐨𝐧𝐬𝐢𝐬𝐭𝐞𝐧𝐭, 𝐫𝐞𝐩𝐞𝐚𝐭𝐚𝐛𝐥𝐞 𝐩𝐞𝐫𝐟𝐨𝐫𝐦𝐚𝐧𝐜𝐞 𝐢𝐧 𝐨𝐮𝐫 𝐄𝐦𝐞𝐫𝐠𝐢𝐧𝐠 𝐌𝐚𝐫𝐤𝐞𝐭𝐬 𝐃𝐞𝐛𝐭 𝐬𝐭𝐫𝐚𝐭𝐞𝐠𝐲. We have not been shy about our enthusiasm for the 𝗔𝗜-𝗽𝗼𝘄𝗲𝗿𝗲𝗱 models driving our investment process at Danske Bank’s Emerging Markets Debt team. But what does this look like 𝑖𝑛 𝑝𝑟𝑎𝑐𝑡𝑖𝑐𝑒? Let me give a concrete example. Our proprietary machine-learning country rating model is estimated twice a year. The process is forward-looking; once the model has been trained, we feed it forecasts of macroeconomic data, and the model produces an assessment of where sovereign credit risk is headed over a 12-month and 24-month horizon. One country that stood out during this exercise in the summer of 2023 was Albania. According to our model, credit risk in 𝐀𝐥𝐛𝐚𝐧𝐢𝐚 was roughly one rating notch better than the assessment of the rating agencies. Markets were still pricing Albanian bonds in line with the rating agencies, 𝘀𝗼 𝘄𝗲 𝘀𝗮𝘄 𝗮 𝗺𝗶𝘀𝗽𝗿𝗶𝗰𝗶𝗻𝗴 𝘄𝗲 𝗰𝗼𝘂𝗹𝗱 𝗲𝘅𝗽𝗹𝗼𝗶𝘁 𝗯𝘆 𝗽𝘂𝗿𝗰𝗵𝗮𝘀𝗶𝗻𝗴 𝘀𝗼𝗺𝗲 𝗔𝗹𝗯𝗮𝗻𝗶𝗮𝗻 𝗘𝘂𝗿𝗼𝗯𝗼𝗻𝗱𝘀. That investment has proven to be quite profitable. Between June 2023 and the biannual model update in June 2024, Albanian Eurobonds returned an average of 11.4% (in EUR terms). In contrast, the benchmark EMBI Global Diversified index (EUR hedged) returned 8.8% over this same period. We use the model to guide country allocation decisions across emerging markets; sometimes the model determines that a country is better than its official rating, and sometimes that a country is worse than its official rating. But taken together, these allocation decisions have contributed to an excess return of +447 bps (gross of fees) over the relevant period. Meanwhile, the rating agencies are slowly catching up with our forecasts. In March 2024 S&P upgraded Albania one notch to BB- and in April, Moody’s improved the outlook on its rating for Albania from stable to positive. These improved ratings have been the primary driver of Albania’s outperformance over this period. 𝐖𝐡𝐢𝐥𝐞 𝐰𝐞 𝐡𝐚𝐯𝐞 𝐢𝐥𝐥𝐮𝐬𝐭𝐫𝐚𝐭𝐞𝐝 𝐣𝐮𝐬𝐭 𝐨𝐧𝐞 𝐞𝐱𝐚𝐦𝐩𝐥𝐞 𝐡𝐞𝐫𝐞, 𝐨𝐮𝐫 𝐬𝐮𝐢𝐭𝐞 𝐨𝐟 𝐀𝐈-𝐩𝐨𝐰𝐞𝐫𝐞𝐝 𝐦𝐨𝐝𝐞𝐥𝐬 𝐬𝐞𝐫𝐯𝐞 𝐚𝐬 𝐭𝐡𝐞 𝐛𝐚𝐜𝐤𝐛𝐨𝐧𝐞 𝐨𝐟 𝐨𝐮𝐫 𝐢𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭 𝐩𝐫𝐨𝐜𝐞𝐬𝐬, 𝐚𝐥𝐥𝐨𝐰𝐢𝐧𝐠 𝐮𝐬 𝐭𝐨 𝐠𝐞𝐧𝐞𝐫𝐚𝐭𝐞 𝐜𝐨𝐧𝐬𝐢𝐬𝐭𝐞𝐧𝐭 𝐚𝐧𝐝 𝐫𝐞𝐩𝐞𝐚𝐭𝐚𝐛𝐥𝐞 𝐩𝐞𝐫𝐟𝐨𝐫𝐦𝐚𝐧𝐜𝐞 𝐢𝐧 𝐨𝐮𝐫 𝐄𝐦𝐞𝐫𝐠𝐢𝐧𝐠 𝐌𝐚𝐫𝐤𝐞𝐭𝐬 𝐃𝐞𝐛𝐭 𝐬𝐭𝐫𝐚𝐭𝐞𝐠𝐲.

177

13 kommentarer

Very engaging meeting this morning in Stockholm 🇸🇪with Portfolio Manager Joakim Ahlberg . Joakim holds a Master in Finance from Stockholm School of Economics, 17 years in equity management and 8 years in the buy-side. He joined Nordea Asset Management as head of research in the industrial and mining sectors in March 2014. In 2016 he became co-PM of the #NordeaGlobalStars strategy and in 2018 Lead PM of the #NordeaNorthAmericanStars strategy. AI and Big Tech are the catalysts of the record-setting rally across U.S. equities. However, the US economy is slowing the pace of growth and if there is pressure on employment, then the consumer is going to have some issues. As long as consumers have jobs, the markets are generally pretty well supported. #GlobalStars 🌎 #NorthAmericanStars 🇺🇸

95

Fed insurance policy? Over the past few weeks, bond yields have turned sharply lower in the US, adding fresh support for the stock market. But will the Fed confirm hopes of a less hawkish stance at next week’s FOMC meeting? I think it would make sense to keep the option of faster cuts open in order to take out some insurance against slowing growth. Consumer confidence has been falling amid stagnation in both household income and spending and the evidence of a cooling labour market is also intact after May’s jobs report. Employment is still rising, but the trend is slowing and it is not fast enough to prevent a gradual increase in unemployment. There is a delay from the first cut to the macro effect materializing, so the Fed has to be prepared to move before serious recession triggers materialize. For now, it still looks like a soft landing and in thst scenario, lower rates open for a summer rally in equities.

35