Price Forecast for El Dorado Wine Grapes for 2017 Harvest
Recently, I've had the chance to focus a bit on the El Dorado AVA. I did work for a client there and also put together a presentation for the El Dorado Wine Growers Association. Click here to get the report I put together for the growers. In the process, I noticed that the reported El Dorado grape prices look pretty easy to forecast. In fact, the report and presentation included a regression that could be used to forecast prices for 2016 (the El Dorado Wine Grape Survey had not quite been released yet.) The forecast ended up being off by only about 1%.
So, I decided to give it a try. My forecast for the average price per ton paid for all grapes from El Dorado County is $1,601.02 for the 2017 harvest. A few caveats:
1) This prediction is for the price reported in the El Dorado County Wine Grape Survey. El Dorado County and the El Dorado AVA are not one and the same. The average price in El Dorado AVA may be a bit higher than for the county as a whole. Nor is the reported price necessarily the true, average price, due to the limitations of gathering data.
2) I did not put much effort into this forecast, so take it with a grain of salt. In fact, this forecast is based purely off a simple linear regression of the price's time series.
3) This forecast is based off of only 9 data points, as the report only goes back through 2008. For such a small sample size, however, the numbers look good. The correlation is .9221; the p-value of both intercept and the trend variable are roughly .0004. The standard error is 49.16. The empirical results for jack-knifed testing can be found below:
OK, so there's a lot in caveat number 3. Basically, though, it looks to me like this method is likely to be very accurate going forward. In concrete terms, I would estimate that, for next year, there is a 50% chance that the price will be accurate within 1% and a 70% chance that it will be accurate within 3%. There is no more than a 1 out of 4 chance that it will be off by more than 5% and no more than a 1 out of 8 chance it will be off by more than 10%.