I’ve recently come back from a week on Ibiza—the smaller and cooler of the main Spanish Mediterranean isles—enjoying what has to be one of the most fantastic climates on earth. I come back to the realisation that I could have been more spendthrift in the pool bar despite its grotesquely overpriced drinks and snacks. Stocks are flying, credit spreads are narrow and volatility has plunged to a new low for the year. My relatively defensive portfolio is currently tracking a punchy 3.8% monthly gain for May, just shy of the 4.4% rise in the S&P 500. Long may it continue. I will have more to say about this in due course, but in the first instance, my recent work suggests that this rally has one strong tailwind on its side; the cyclical picture in the global economy has improved. My measures of global cyclical activity hit a new high at the end of Q1, and into Q2, from a trough last year, and cyclical equity returns are now re-accelerating, after softening a touch at the start of the year.
Read MoreThe BIS has a Bulletin out on the usefulness of AI and large language models. They’re not terribly impressed.
Read MoreWhen posed with a logical puzzle that demands reasoning about the knowledge of others and about counterfactuals, large language models (LLMs) display a distinctive and revealing pattern of failure.
The LLM performs flawlessly when presented with the original wording of the puzzle available on the internet but performs poorly when incidental details are changed, suggestive of a lack of true understanding of the underlying logic.
Our findings do not detract from the considerable progress in central bank applications of machine learning to data management, macro analysis and regulation/supervision. They do, however, suggest that caution should be exercised in deploying LLMs in contexts that demand rigorous reasoning in economic analysis.
I have just returned from two weeks of holiday, and I have a lot on my mind. First things first, on the war between Israel and Hamas; emotions are running high and as a result, the quality of initial opinion and analysis is clouded and governed by hard-held priors. That always make for a treacherous information environment, especially in a situation as complex as is the conflict between Israel and its border states, not to mention the political situation in the Middle East as a whole. I am making the following initial assumptions. As long as it is Israel and Hamas pounding each other to a pulp—with devastating consequences mainly in Gaza as the IDF brings the heat—markets will eventually stop caring. The obvious risk is that a bloody conflict between Hamas and Israel spills over into wider military conflict in the Middle East. It is grim irony that Iran recently warned how a heavy-handed response by Israel “could spiral out of control and ricochet into far-reaching consequences”. Teheran is right, and I suspect that it is exactly what it, Hamas, other key actors in the Middle East, not to mention Russia, want.
Read MoreThe prevailing mood in global macro discussions seems to be as follows; inflation is past its peak, but it is set to remain a lot higher for a lot longer than initially anticipated, forcing central banks to continue hiking, keep rates higher for longer, or a combination of the two. The interest rate shock in the UK, as markets have adjusted their expectations for the BOE bank rate higher, and hawkish comments from the ECB are the two most obvious cases in point in developed markets. But a surprise hike by the Bank of Canada, and a larger-than-expected hike in Norway have added to the sentiment. We only really need the Fed to be forced into a hawkish turn to complete the narrative. This shift is important for investors. We are not just trying to calibrate when central banks will pause their hiking cycles—probably soon—but we’re also increasingly discussing, and pricing, how long rates will stay elevated, and whether central banks will have to resume hiking before they cut. Higher-for-longer, or #H4L, is already a trending hashtag on FinTwitter.
Read More