Blending Optimization is widely used by refineries and traders for scheduling and product blending decision making. In a financial context, one apparent benefit is enhanced profit. But its impact on risk and hedging is relatively not well known (at least to me!), so I decided to do some investigation by myself.
Consider a business scenario. (1) on day one, the operator runs optimizer, and get best product mix, based on the price forecast. Let’s call this one as “planned” mix. (2) at the same time, he calculates risk (say 95% one-sided VaR) using the planned mix, and sells financial swap to hedge his scheduled production (3) at day 20, the operator runs optimizer again, using realized prices and get an optimized solution (“final” solution). Then he executes blending and deliver the product.
What is unique in blending is, his final mix may differ from the planned one. Consequently, (1) his estimate of the risk (VaR) is incorrect; in fact, it’s over-stated because he can always choose better solution (if exists) than his planned one through optimization. (2) P&L is not hedged, as the hedging transaction is based on the planned mix, not the final mix which is not known at the planning stage. Continue Reading in LinkedIn