Meyrick, I wrote about a breakdown of relationships between commodity currencies and commodity prices over at Money Inside and Out. There may be a read through to your observation on Cu and 5/5. One thing that springs to mind for the overall time series (not just the recent curiosity) is that the copper price broke out of its historical nominal range due to the cost shock of the 2021-23 period, and that the new SRMC level creates a floor that prevents it from adjusting back to pre-pandemic nominal levels ($8000/t is the new $4000/t in LME terms). In other words, there is a structural break.
Hi Huw, thanks for your comments. You make some interesting and useful observations. I’m sure I can learn from your approach and would gladly compare notes, especially as our methodologies may differ significantly. I look at the top 4 principal components across a broad global portfolio of equities, currencies, bonds and commodities on a rolling 12m window. There are obvious criticisms of such an approach but it does allow structural changes to be taken into account quickly. What stood out for me was singular dislocation of copper from the rest of the portfolio in a highly unusual fashion. Normally there is some kind of read-through to the behaviour of parts of the global portfolio. I understand that a good part of the copper price decline may have been triggered by reassessment of inventory requirements but that is clearly a macro-economic consideration too. Sometimes things don’t fit easily, which is maybe when they contain the most useful information.
Meyrick, there could be a view that a tariff disturbance of considerable scale would create just this sort of quirk. Tariffs hurt the economies of export-oriented manufacturers - they consume the copper. Add that to a double whammy on the inflation side - immediate lift in the price level from tariffs, medium term cost premium from onshoring - and you have your 5y/5y move. Going back a little further, if you decomposed moves in 5y/5y into oil plus growth expectations, I suspect there wouldn't be much left to explain. Now that the inflation and growth view can be plausibly separated due to supply side issues, a pro-growth asset like copper can plausibly de-couple from an inflation signal which may have actually been a growth signal in the past. Happy to chat offline at a time of your convenience.
Having been involved in the markets for more than 40 years, I sense that the current situation of correlation breakdowns is unprecedented. while it was not unusual in the past for one or two long-term correlations to break, it is almost as though every relationship has busted. What exactly does risk-on or risk-off mean anymore? Ultimately, I cannot look at the copper market, and take my understanding on what is required to dig more of it out of the ground (which while not exhaustive is reasonably robust as I continue to try to learn more about the processes) and see that over time, the red metal must rally. but these days, timing is incredibly difficult. it could take 6 months of further decline before something catalyses a rebound.
Meyrick, I wrote about a breakdown of relationships between commodity currencies and commodity prices over at Money Inside and Out. There may be a read through to your observation on Cu and 5/5. One thing that springs to mind for the overall time series (not just the recent curiosity) is that the copper price broke out of its historical nominal range due to the cost shock of the 2021-23 period, and that the new SRMC level creates a floor that prevents it from adjusting back to pre-pandemic nominal levels ($8000/t is the new $4000/t in LME terms). In other words, there is a structural break.
Hi Huw, thanks for your comments. You make some interesting and useful observations. I’m sure I can learn from your approach and would gladly compare notes, especially as our methodologies may differ significantly. I look at the top 4 principal components across a broad global portfolio of equities, currencies, bonds and commodities on a rolling 12m window. There are obvious criticisms of such an approach but it does allow structural changes to be taken into account quickly. What stood out for me was singular dislocation of copper from the rest of the portfolio in a highly unusual fashion. Normally there is some kind of read-through to the behaviour of parts of the global portfolio. I understand that a good part of the copper price decline may have been triggered by reassessment of inventory requirements but that is clearly a macro-economic consideration too. Sometimes things don’t fit easily, which is maybe when they contain the most useful information.
Meyrick, there could be a view that a tariff disturbance of considerable scale would create just this sort of quirk. Tariffs hurt the economies of export-oriented manufacturers - they consume the copper. Add that to a double whammy on the inflation side - immediate lift in the price level from tariffs, medium term cost premium from onshoring - and you have your 5y/5y move. Going back a little further, if you decomposed moves in 5y/5y into oil plus growth expectations, I suspect there wouldn't be much left to explain. Now that the inflation and growth view can be plausibly separated due to supply side issues, a pro-growth asset like copper can plausibly de-couple from an inflation signal which may have actually been a growth signal in the past. Happy to chat offline at a time of your convenience.
Having been involved in the markets for more than 40 years, I sense that the current situation of correlation breakdowns is unprecedented. while it was not unusual in the past for one or two long-term correlations to break, it is almost as though every relationship has busted. What exactly does risk-on or risk-off mean anymore? Ultimately, I cannot look at the copper market, and take my understanding on what is required to dig more of it out of the ground (which while not exhaustive is reasonably robust as I continue to try to learn more about the processes) and see that over time, the red metal must rally. but these days, timing is incredibly difficult. it could take 6 months of further decline before something catalyses a rebound.
Maybe but we're not dealing in conventional correlations.
I wonder if there are any more conventional correlations
I have always preferred unconventional correlations - at least that's what my wife says.
😂