The following chart is one of hundreds that I use in my day-job as Chief Eurozone Economist at Pantheon Macroeconomics. It plots a normalised Z-score index of surveyed new manufacturing orders in Germany alongside year-over-year growth in factory orders, ex-major orders. It’s worthwhile spelling out the meaning of this chart in the world of economic research and forecasting. The factory orders numbers are so-called hard data, which in this case means that they’re official numbers of real activity reported by the statistical office. The PM new orders index, by contrast, is my home-cooked index of so-called soft data. Specifically, these are survey data, compiled by the likes of the EU Commission, IFO, S&P, and national statistical offices. We’re only interested in these numbers to the extent that they tell us something about the official/hard new orders data, which in turn could help us pin down trends in industrial production, exports, GDP growth, employment and so on. From simply eye-balling the chart, the two series look coincident, but note that the surveys are released ahead of the official data, so that we always have survey numbers that are one-to-two months ahead of the official data. In other words, when it comes time to forecast new orders for the month of December, we will already have survey data for that month. This should, in theory, help us to better forecast the official real new orders data.
Read MoreI want to thank everyone for bearing with me in my AI advent calendar project this year. I know it probably isn’t to everyone’s taste. But it was a fun project, and it helped me get better acquainted with AI, which will becoming an increasingly inevitable force to be reckoned with in 2024, for all of us , for better or worse. I am going to spend a few days over Christmas in Denmark with my wife, who is from Canada, and my extended family. I asked ChatGPT 4 to come up with a picture capturing that mood.
Have a break Christmas break and a Happy New Year!
Read MoreThis year’s advent calendar project was primarily done on ChatGPT 3.5 and DALL E2. I only recently acquired access to GPT 4.0, which is infinitely more powerful, especially in its ability to respond to real-time events—it is trained on current data—to create more detailed and realistic pictures and, as I will show here, to analyse data. So, I thought that I would pivot on this entry and show an example of some empirical and statistical economic analysis I’ve done with GPT 4.0.
The first step was to create my own GPT optimised for time series analysis done on data uploaded in excel or CSV format. This didn’t take long, though it is not clear to me that my little GPT is any better than the generic "Data analysis” GPT, but over time, it should become better at the task that I am interested in, in theory.
Read MoreI am coming into 2024 in a decent position. My MinVar equity portfolio, designed to extract the best from both worlds in the perennial battle between growth and value, has done largely what it is supposed to do. It has offered positive, but below-beta, returns with below-beta volatility, the latter which means that your humble blogging investment analyst has been able to sleep calmly at night. In bonds, I moved my exposure onto the front early in 2023 in line with the yield curve inversion. At this point I see no reason to change that strategy. Why buy negative carry in duration when you don’t have to? There will be a time to take a strong bet on duration, but I can’t really see that point until either the front-end has collapsed under the weight of global central bank easing, or unless the curve rinses everyone by bear-steepening sufficiently to restore a positive roll and carry in the long bond. In other words, I don’t see any reason to buy duration as long as the curve is still deeply inverted.
Read More