(different person) Just eyeballing the graph, If we can attribute the entire spike to a "pulling forward" effect, then I'm guessing it might end late 2026 or so

(different person) Just eyeballing the graph, If we can attribute the entire spike to a "pulling forward" effect, then I'm guessing it might end late 2026 or so (as a gradual slope over longer period, not a hard noticeable line). Then give it a few months for stabilization / to ignore random normal fluctuations that just happen to be at that time. So without more information, early 2027 we can call it a long-term trend for certain? Extending it back to 1995 or so like the first comment suggested would give us a few more upsets in the market to see more easily if this effect actually exists in this data. That said, their source is https://fred.stlouisfed.org/series/IHLIDXUSTPSOFTDEVE https://fred.stlouisfed.org/series/IHLIDXUSTPSOFTDEVE so that data probably doesn't exist.