Has anyone seen any study or discussion of inflation arbitrage? Not in the trivial sense of if you predict inflation borrow and invest, but in the case of viewing a country as multiple separate markets and there being inflation differences between them, which in turn means there is opportunity for arbitrage?
It seems to me like a failure to accept that adjacent markets sharing a currency can have different rates of inflation is a large part of why people are so damned bad at understanding and measuring inflation. The article is a good example, every metric and estimate proposed assumes that there is a single inflation rate for the currency. But if you instead thought of it as one good being trades in multiple different market places, then its obvious that there could be differences in price between these markets which traders could exploit. Critically, unless they did so and doing so was a near perfect market, there would effectively be multiple different prices for the commodity and the textbook use of inflation would be such a shitty model as to be near useless. If you instead asked, how many ingots of currencium would I need to buy a bag of other goods, it would be obvious that this would also require a statement of in market A. I doubt the reason this arbitrage opportunity is entirely missed, but it could be that its unusually hard to exploit. However, I suspect it is partly because even among financially literate people, a currency has one rate of inflation is a common idea.
That said, the single market model where currency has a common price provides shitty predictions. For instance, the strongest counter argument against the apparently obvious statement. SNP500 has not increased more in value in 2020 than 2019, its mostly just inflation, is: No metric of inflation say it has been anywhere near 40% in 2020.
However, if we view this as two different markets, A(capital), B(consumer) where the inflation is different for each. Then any metric which is designed to predict inflation assuming its the same in both would by necessity underestimate one and overestimate the other.
The counterargument would be that if this was the case an efficient market would eliminate the arbitrate opportunity. But thats barely true in the most ideal cases, and its easy enough to come up with such arbitrate opportunities.
For instance, we know that historically, whenever there is inflation, stocks respond quickly, but salaries generally lag behind. This is damned near proof of the multi market model on its own, but a model with more parameters always fits the data better. A sufficient, but not necessary proof would be the existence of insurance contracts for and against inflation in another market priced in the same currency. In short, is there a reason that salary futures aren't a thing?