Cloudonomics: The Business Value of Cloud Computing, + Website

Relationships between Demands

We can categorize four interesting relationships between two or more demand functions: (1) independent and uncorrelated, (2) negatively correlated, complementary, and/or countercyclical, (3) positively correlated, (4) generally uncorrelated but with simultaneous peaks. These relationships impact service provider Cloudonomics.

Independent

When demands are independent, aggregation creates a smoothing effect, in the same way that a diversified portfolio has a lower volatility than the most volatile stock in the portfolio. Aggregating the demand from increasing numbers of customers results in increased smoothing—sanding the jagged edges of volatility, as it were—that can be easily characterized, as we shall see.

Negatively Correlated

When demands are negatively correlated—either complementary or countercyclical—it only takes two or three demands to potentially create a relatively smooth aggregate demand. Winter ski resorts that turn into golf and tennis resorts in the summer maintain relatively high occupancy compared to those resorts that are just ski resorts. Such complementary demands may be achieved through luck or planning. Samuel Insull, for example, built and ran one of the first electric utilities, Chicago Edison, in the early 1900s. He strategically balanced the demands of the “traction” companies—streetcars and elevated railways—whose peak demand was during morning and evening rush hours, with offices and factories—whose peak demand was during ...