Kieron Bradley, one of my colleagues at Sun, during a piece of client consulting recently had reason to use financial option theory & language to justify why CPU’s in Sun’s large systems are more expensive to buy than those in the smaller ones. He and the customer had examined all the TCOO factors they thought were relevant and the fact remained that if one wanted to take a utility view of CPU supply, it was cheaper to buy and run smaller systems rather than larger ones. (This particular analysis did not perform a variable utilisation analysis. It was assumed, (or defined as policy) that all CPUs would run at a given % utilisation. Contradicting this assumption, it is a fact that large (and flexibly partitioned) systems are easier to keep busy.)
It should be clear that a large, flexibly partitioned systems cost more to manufacture and incur higher R&D costs by the vendors, and when customers buy these systems, they get this additional technology and functionality. Most importantly, they get the system bus. (You can’t buy these from Dell or Tiny). The SunFire™ system bus cost Sun a lot to invent, prototype and test and delivers to its customers reliability (by offering fault isolation and superior recovery), size (SunFire domains can scale to very high numbers of CPU, and offer very large vertically scaled UNIX execution environments) and flexibility.
Kieron’s challenge was to bring home to some very hard-nosed and practical banking IT people that flexibility could be expressed in monetary terms i.e. it was more than just nice. This is where the use of options language & theory came in. Buying larger systems than required today is also buying a system with an option to expand. This planning option allows the growth within a operating system instance, so permitting a change in scale, leveraging the size benefit as well as the flexibility benefit of these large systems. The planning option also allows the creation of additional domains or partitions. The details of when and how the resource arrives and is deployed, together with what and when the price is agreed are matters of technical & commercial detail.
Amran & Kulatilaka in their book “Real Options”1, argue that options & flexibility create value. These project planning options need to be evaluated and factored into investment plans. The Real Options guru’s develop a financial market proxy for the project planning option and then evaluate the investment in the light of this analysis. The key is that flexibility creates value and that static appraisal techniques such as NPV% will underestimate the value of the flexibility, opportunities for profit are missed and investment funds underused. They also argue that a projects option/value curve needs to be built up out of decision modules of which there are only four, based on whether the enterprise wishes to buy or sell the underlying asset, or buy or make the contract.
One example of a real option is to buy a Sun StarFire F25K, which is half full, and eighteen months later fill the remainder of the card cage & system bus. By buying the Starfire, they have (probably) paid more than buying this capability using smaller systems, but they have the opportunity to scale the systems to meet market growth. The difference in cost between the two investment plans is the Real Option price. Once a purchaser factors in the revenue opportunities of scalability, then the sums can be fairly simple. However Amran & Kulatilaka argue that by using a financial markets proxy and appropriate financial markets valuation tools the revenue opportunity and thus the underlying asset value can be established. This involves both care in framing the question and more complex mathematics. The answer is often different using real options analysis rather than NPV%
1. “Real Options, Managing Strategic Investment in an Uncertain World”, Amran & Kulatika, Harvard Business School Press ISBN 0-87584-845-1
First posted on my sun/oracle blog, and republished here in Feb 2016.