Abstract
World tuna prices exhibit substantial fluctuations over time. We studied monthly tuna and preceding crude oil prices from 1986 to 2018, using linear regression models with autoregressive and moving average (ARMA) errors. Results indicated that a model including an increasing linear trend, the oil price 21 months earlier, and a simple ARMA(1,1) error process could predict the monthly tuna price reasonably well for recent years, but not prior to 1999. This suggests that oil prices began to affect tuna process only after the global financial crisis, but it takes nearly two years before a change in the oil price affects the tuna price.
| Original language | English |
|---|---|
| Pages (from-to) | 1221-1226 |
| Number of pages | 6 |
| Journal | Songklanakarin Journal of Science and Technology |
| Volume | 42 |
| Issue number | 6 |
| DOIs | |
| Publication status | Published - 2020 |
Keywords
- Tuna prices
- Crude oil prices
- ARMA process
- Regression with autocorrelated errors
- Time series forecasting