So I've done a ton of internet research and lost. I am trying to forecast dayrates for offshore drillships. and I have two exogenous variables that have close correlation with my historical dayrate data. (around 0.75 correlation to the dayrate data for both the demand and supply indices). lots of data feed into these. The thing is, this data is very boom-bust. My dataset is from 2010-2024 for the dayrates. I have forecast data out to 2029 because a lot of these value are somewhat "known" (new build ship que, unproven / unsanctioned E&P activity, etc). What is the absolute best model for my problem? I believe my demand supply indices are very strong predictors of these rates.
I've tried SARIMAX and later Markov Switching ARIMAX but all of them seem to not account for the stages of my bust. Ultimately, I do believe my supply and demand indices are strong predictors historically--how should I approach this?
I really just am asking for a suggestion of what models might fit well to this data. The supply and demand indices are from 1-150 around (min-max) for demand in billions USD of total deepwater investment, and 0-1 for supply tightness that comes from several other datasets.
Here is a small sample of my data:
Date Average Lead Dayrate SupplyTightnessIndex DemandIndex
6/1/2010 500869.5381 0.578840012 94.82266832
