Evolution of French Wheat Prices, a Breakdown by Protein Level

In this instance, the curve and points in Figure 1 showcase a very different situation to Germany:

• France 11% protein Wheat product is unambiguously driving the market trends and prices
• The fit between the curve and the data points is least strong in early 2018, during the bear market (Dec 2018 – Sep 2019), and from May 2020 onwards
• The seasonal rally in Sep 2018 and Sep 2019 for their respective campaigns highlight the impact of the droughts on Wheat prices during the 2018-2019 campaign

The observations for French products are straightforward. Wheat prices for the 11% protein level is the top French product driving the Wheat prices. However, for certain periods – particularly in a bear market – the fit is not as strong, implying other Wheat prices for other protein products.

Moreover, the seasonal market rally behind 11% protein Wheat for both the 2017-2018 and 2019-2020 campaigns highlight the impact droughts had in the Summer of 2019. That year the rally stopped around July, with prices dropping to their lowest point in this time frame at the end of the campaign. This drop in France’s Wheat prices is due to the hard drought of summer 2019, just the same as for Germany’s Wheat prices.

Remarkably, the start of the bear market for these products is also the inflexion point for 11.5% protein Wheat prices, with CNF going bull. Similarly, the emergence of 13.5% Wheat prices and growth synchronized with the 11% Wheat price curve when the market bulls also indicated this trend.

Figure 2 displays the Wheat price curves by protein levels for the French dataset. Although there aren’t many data points for several products, it is still possible to highlight the following: 

• 13.5% Wheat and 11.5% Wheat price first appear during strategic periods
• Wheat price curves for  9% and 10.5% protein products show links with 11% Wheat price curve in other specific periods too

These two observations provide great insight into the French market. Indeed, for the first point, the fact that these two products appeared at crucial inflection points supports the idea that these products’ purpose was to propose alternatives to the top product and stabilize or boost the market at the end of the 2018-2019 campaign.
Moreover, the lack of data for  13.5%, 11.5%, and 12.5% Wheat prices stresses these products’ strategic purpose even further.

From May 2020 onwards, the introduction of 9%  and 10.5% Wheat price quotes appeared during the rally behind the 11% Wheat product until the end of September for the 2019-2020 campaign. This event is not new, as it corresponded to the similar introduction of 12% Wheat price in May 2018 for the 2017-2018 campaign.

Therefore, France’s 11% Wheat is predominantly driving the French market, since at least 2018, which already differs from the German Wheat market.

However, where Germany had numerous products with various protein levels, the norm for the French market is to view other protein products’ quotations  – namely 9%,  10.5%, and  12% Wheat – during the Spring rallies, probably due to seasonality and better climate conditions, as well as boosting the market rally before the end of the campaign. That said, 13.5% and 11.5% Wheat prices appeared with the necessity to both boost and stabilize the 11% Wheat prices during the 2018-2019 campaign, in bear and bull markets alike. Nonetheless, the limited amount of data for these Wheat prices only shows that it was not a sought-after alternative.

Ultimately, the French Wheat market showed more flexibility with cash Wheat prices, as various – situational – protein Wheat products have quotations at strategic times.

A similar table to the Germany Wheat Price one also provides excellent insight into the French Wheat Price relationships. Following the trends found previously, the table can highlight the trend around 11% Wheat prices.

The combinations computed in the table for Figure 3 exclude obvious perfect correlations (comparison with the same parity), leaving only combinations between the top products for each origin. Moreover, combinations that had correlations below 0.725 were excluded. 

The changes for collinearity statistics between similar combinations are due to the variable dependence in multicollinearity. Perfect correlations between some combinations don’t show relevant numbers for the collinearity test, as (almost) perfect combinations’ p-values are just rounded to 0 since they are so small, and t-values tend to minus infinity.  On the other hand, while some price curves are significantly correlated, they may not be collinear.

Remarkably, there were some combinations between German Wheat prices showing strong correlations but weak collinearity relationships. For France Wheat prices, some combinations show the same pattern; namely, combinations between France 11 % Wheat, France  12.5% Wheat, and France 13.5% Wheat (i.e., fr_11 | fr_125, fr_11 | fr_135, and fr_135 | fr_11). This opposition is relatively normal as the data for these Wheat prices followed the same direction as the 11% Wheat prices – hence the strong correlation. Nonetheless, the lack of data explains why there are extremely poor collinearity relationships between these combinations.

Ultimately, there are mostly strong collinearity relationships for these combinations, weighted by the t-static variations. The p-values also consolidate the t-statistic, with most combinations falling between the 1% and 10%  critical values, with most around the 5% critical value (or above 95% confidence level), showing high predictability overall between them.

This data then shows that France Wheat prices align, correlate, and are collinear with the  11% Wheat price curves. Nevertheless, the long term relationships between Wheat prices are not particularly relevant, as the Wheat prices for products other than 11% Wheat are situational.