How do benchmark prices “influence” the AI-derived pricing? Is there a correlation effect?

The benchmarks provide the reference for daily movements in prices from the model. The concept of correlation is a measure of “linear dependence”, as widely known in statistics.

In VAI’s model, various non-linear and implicit (not closed-form) relationships in the data are captured, including among benchmarks. As such, what the model captures is much more powerful as it implicitly infers “nonlinear dependences” among benchmarks and with prices from the data for various combinations of the attributes that define a carbon project.