Abstract
The price of farmed Atlantic salmon from Norway has increased in recent years. This new regime follows several years of consistently falling prices. At the same time price volatility has increased substantially. This article models the volatility of salmon prices and establishes empirically that volatility is on an increasing trend. Further empirical analysis suggests that the volatility trend is largely accounted for by the common trend in other food prices relevant to salmon, including meats, cereals, oils and fish meal observed in recent years. Other potentially contributing factors to volatility are also discussed. This includes the role of the 2005 maximum total allowable biomass restriction, the 2006 introduction of the Fish Pool ASA futures market for salmon, the Chilean Salmon crisis and the increasing use of bilateral contracts.
Keywords:
Notes
Note: The Q-statistic is the Box-Pierce test for the null of no-remaining residual autocorrelation. Lags are given in parenthesis. The Q-statistic is χ2(nlags) distributed under the null, where. The ARCH statistic is Engle's LM ARCH test for presence of residual ARCH effects. The test regresses squared residuals on own lags and tests for significant coefficients using the F(nlags,nobs-nlags) distribution.
Market shares for 2009 and 2010 were 60% and 65%, respectively; these numbers are naturally inflated by the Chilean disease issues.
Logarithmic return of price at time t is defined as ln(p t ) − ln(p t − 1).
For an example of a firm specific harvest model for salmon, see Guttormsen (2008).
Please note that due to the seasonal variation in the biological growth pattern (Asche & Bjørndal, Citation2011), full utilization of the MTB is impossible.
It is also of interest to note that the introduction of the futures market is so close in time to the introduction of the MTBs, that it is difficult to separate the potential effects of these two measures on price volatility.
Numbers are from the Fish Pool ASA annual rapport. (http://fishpool.eu/uploads/%C3%85rsregnskap_2010_Fish_Pool_ASA.pdf).
For details on the FAO fish price index, see Tveterås et al. (2012).
The specific seasonal representation for the weekly data is: where αsin, j and αcos, j are coefficients to be estimated for each seasonal cycle (k 1 = 52(annual), k 2 = 26(semiannual), k 3 = 13(quarterly)).