James Effinger of TPI's Financial Analysis Services Group will be "blogging about the bottom line" this week.
Deadbands are generally defined as a small range around a contracted baseline volume against which no price adjustments (ARC/RRCs) are made. For example, a given volume is 1,000 and the deadband is given as 5 percent. Thus, for any volumes between 950 and 1,050, there are no price adjustments.
But chargeback can become problematic due to fluctuating unit prices around the upper and lower limits of the deadband. Using the example above, assume the base charges for the contract volume of 1,000 is $1,000,000 (or $1,000 per unit).
- The effective unit rate at the upper limit (1,050 units) is $952.
- The effective unit rate at the lower limit (950 units) is $1,053.
This creates a difference in the effective unit rate between the upper and lower limit of approximately 10 percent. So when does it make sense to implement deadbands?
Typically, when you have a very high volume on a resource baseline, such as mainframe print pages, which could be in the millions each month. A deadband could reduce the hassle of making very small dollar adjustments every month around a very large baseline volume.
But there can be hiccups during implementation. Here's why:
- You still have to count the items every month, quarter, etc.
- You still have to compare against the deadband and determine if an adjustment is needed.
- Once the deadband is exceeded, do all of the units above the baseline receive the ARC/RRC or just the units outside the deadband? You must provide clarity in the contract language on how the ARC/RRCs work in conjunction with the deadband.
- You have to perform nearly all functions except applying the ARC/RRC rate if inside the deadband. Then in the following months you have to track which of the prior months received the ARC/RRCs adjustments and which months did not because they were inside the deadband. This historical tracking of when ARC/RRCs were applied and when they were not can be problematic for the governance team especially as turnover occurs.
The bottom-line: Deadbands serve a practical purpose but shouldn't be used without an understanding of the rationale for use, potential problems, and long-term impact on governance.
What are your experiences with using deadbands?