How Cobra Spreads ETC
When calculating forecast costs, Cobra spreads the ETC for each work package or control account between the estimated start and finish dates (or between the status date and the estimated finish date for in-progress work).
In addition, Cobra lets you explicitly update the spread of ETC in a manner similar to the spreading of the budget in control account planning.
To be accurate, forecast calculations must take into account that schedule or rate variances (either positive or negative) may result in the use of rates different from the ones used to budget the work. Thus, Estimate to Complete (ETC) calculations must take into account how future expenditures will be spread over time.
Cobra attempts to retain the current spread profile when the new forecasts are spread. The profile used depends on both the existence of a current ETC spread and upon the setting of the project option that allows users to use the current budget spread for forecast spreads. If there is no current ETC spread or if the project option is set to spread according to the budget spread, the profile is generated from the budget spread profile. The included budgets for the forecast currently being spread are totaled for the resource by date, and the period budget values are totaled and normalized into percentages, which make up the curve point values. The profile curve has as many points as there are periods in the budget detail, and Cobra stretches or shrinks the curve as necessary if the dates for the work are adjusted at a later date.
Cobra contains special algorithms that preserve the previous spreading of ETC, if at all possible:
- If only the performance factor has changed, the previous spread is retained and the new ETC is prorated against it.
- If the dates have changed but the number of fiscal periods has not changed, the previous profile is simply shifted to match the new dates and is re-rated.
- If the new dates span a different number of fiscal periods than the budget, Cobra attempts to reproduce the existing curve within the new dates. This may result in a stretching or shrinking of the curve, based on whether the number of fiscal periods increases or decreases.
The spread is done by calculating the percentage of the ETC that will go into each of the periods. This is done by mapping the number of periods to be spread into over the number of periods currently making up the profile, weighted by the profile values that are normalized as cumulative percentages (total to 100). Then the number of the period to spread is divided by the total number of periods to be spread into and multiplied by the number of profile points to get a factor. The factor is made up of an integer value and a modulus. The integer value represents the profile point after which the spread point will fall. The modulus is the percentage of the difference between the new two profile points.
For example, assume the following budgeted and forecast costs before statusing:
Periods | ||||||
---|---|---|---|---|---|---|
1 | 2 | 3 | 4 | 5 | Totals | |
Budget | 100 | 200 | 300 | 200 | 200 | 1000 |
ETC | 100 | 200 | 300 | 200 | 200 | 1000 |
After one period, assume that ETC has been calculated as follows:
ETC = BAC Earned Value = 1000 - 50 = 950
This results in the following cost data:
Periods | ||||||
---|---|---|---|---|---|---|
1 | 2 | 3 | 4 | 5 | Totals | |
Budget | 100 | 200 | 300 | 200 | 200 | 1000 |
Earned Value | 50 |
|
|
|
|
50 |
Actual Costs | 100 |
|
|
|
|
100 |
ETC |
|
211 | 317 | 211 | 211 | 950 |
Now, if the estimated finish date for the work package had been pushed back one period (to take the poor performance into account), the spread would be:
Periods | |||||||
---|---|---|---|---|---|---|---|
1 | 2 | 3 | 4 | 5 | 6 | Totals | |
Budget | 100 | 200 | 300 | 200 | 200 | 1000 | |
ETC | 95 | 190 | 285 | 190 | 190 | 950 |
Note, however, that the profile used depends on how much has been earned. Because the earned value was less than the amount of the first period budget, the entire budget profile was used. If, however, the full amount had been earned, the profile used would have been generated from the last four periods of budget. For example, assume a performance factor of 1 and the following cost data:
Periods | |||||||
---|---|---|---|---|---|---|---|
1 | 2 | 3 | 4 | 5 | 6 | Totals | |
Budget | 100 | 200 | 300 | 200 | 200 |
|
1000 |
Earned Value | 100 |
|
|
|
|
|
100 |
Actual Costs | 120 |
|
|
|
|
|
120 |
First, the spread profile curve is created. It is a four-point curve made up of the values of the last four budget amounts: 200, 300, 200, and 200, which total 900. The profile therefore is:
- Period 2 .222
- Period 3 .333
- Period 4 .222
- Period 5 .222
If expressed as a cumulative curve:
- Period 2 .222
- Period 3 .555
- Period 4 .777
- Period 5 1.0
The ETC profile is then calculated by determining each spread point:
First spread point:
- Factor= 4 * (1/5) =.8
- Integer= 0
- Modulus = .8
- Percent= 0 + (.8 * (.222 - 0)) = .178
Second spread point:
- Factor= 4 * (2/5) = 1.6
- Integer= 1
- Modulus = .6
- Percent = .222 + (.6 * (.555 - .222)) = .422
Third spread point:
- Factor= 4 * (3/5) = 2.4
- Integer = 2
- Modulus= .4
- Percent= .555 + (.4 * (.777 - .555)) = .644
Fourth spread point:
- Factor = 4 * 4/5 = 3.2
- Integer = 3
- Modulus= .2
- Percent = .777 + (.2 * (1 - .777)) = .821
Fifth spread point:
- Factor= 4 * (5/5) = 4
- Integer= 4
- Modulus= 0
- Percent= 1
These spread points are then multiplied by the ETC amount to get cumulative spread values for the ETC, and the interval values are then obtained by subtraction of the previous value:
Periods | |||||||
---|---|---|---|---|---|---|---|
1 | 2 | 3 | 4 | 5 | 6 | Totals | |
Budget | 100 | 200 | 300 | 200 | 200 |
|
1000 |
ETC |
|
160 | 220 | 200 | 160 | 160 | 900 |