Assured value analysis
earned value extended
The objective of this paper is to introduce and describe Assured Value Analysis, a proposed new extension to earned value management (EVM) theory and methodology that utilizes two new measures to expand and improve on existing EVM concepts. This paper demonstrates how Assured Value Analysis (AVA) can provide more representative performance measurement, and improved estimates of the total project cost, than is possible with EVM alone – particularly for projects that include a significant portion of procurement.
This investigation begins with a review of relevant literature on EVM and issues related to its implementation. It also considers methods for cost control on projects that contain a large proportion of procurement. That is followed by a review of the key elements in EVM: the measures, the formulae, and their application.
The new Assured Value measures and formulae are presented in detail, then applied to two hypothetical project situations – one very simple, the other more detailed. These demonstrate the ease of application of the AVA extensions, and provide results that can be compared with those available from EVM techniques.
Earned Value Management and Procurement
Defining Earned Value Management
The Project Management Institute (PMI®) defines earned value management (EVM) as “a method for integrating scope, schedule, and resources, and for measuring project performance. It compares the amount of work that was planned with what was actually earned with what was actually spent to determine if cost and schedule performance are as planned.” (PMI, 2000, P201) This definition is actually quite limited, for EVM comprises not only those three comparisons, but also the use of those three measures to create a new forecast the total cost of the project, termed estimate at completion.
This paper will not dwell on the history or complexities of earned value management; there are several sources that can provide an excellent background. The text by Fleming and Hoppelman (1996) provides a useful summary of the main elements of EVM, together with many references to its history and application. A recent paper in the Project Management Journal by Anbari (2003) not only summarizes the standard measures and formulae, but also proposes a number of interesting extensions related to time. The terms used in this paper are compatible with those contained in Anbari's paper.
The Project Management Institute has issued an Exposure Draft of its Practice Standard for Earned Value Management (PMI, 2004). Even though that is not yet a finished document, it was consulted during the preparation of this paper, with the objective of using compatible terminology and concepts.
For three decades in the United States, EVM was referred to as Cost/Schedule Control Systems Criteria (C/SCSC) when applied to US Government projects in the procurement of major systems from the defense and aerospace industries. Due to the complexity of those evolving systems, the risk of exceeding the cost budget was accepted by the US government, and therefore C/SCSC was adopted as a means of monitoring that risk exposure (Fleming & Koppelman, 1996).
Adoption of EVM
Earned value adoption has been demonstrated in the defense and aerospace industries, particularly in the United States. The Department of Defense in the USA has developed guidelines for the adoption of EVM, and has been active in both promoting and requiring the use of EVM on its projects (Fleming & Koppelman, 1996). However, despite its demonstrated benefits, EVM has not been comprehensively adopted in all industries where projects are managed.
For example, it has not been widely applied on software development projects (Christian & Ferns, 1995) possibly due to difficulties in documenting progress, or due to the extent of “nebulous tasks” (Prentice, 2003).
There are numerous factors that could have slowed the adoption of EVM, and those have been identified in other papers (Kim et al., 2003). EVM requires, of course, a degree of maturity in project management methodology that (from the author's personal experience) is not consistently available in many organizations. For example, many organizations do not include the cost of internal staff time when estimating project costs. The use of timesheets to track actual staff time devoted to specific projects (much less specific activities) is rare outside of consulting firms and other groups that charge clients for their time. Project schedules may be established, but project managers are under no pressure to set a baseline, without which one cannot perform an earned value analysis. And finally, poor scope definition and change control practices can make identification of work packages and related costs very difficult.
In the construction industry, earned value concepts could be readily applied. Almost all construction work is routinely divided into identifiable packages, most deliverables are tangible, and typically costs are carefully tracked (True, 2003). But it is quite evident that earned value has not been accepted as the preferred means of monitoring, controlling, and reporting cost status on construction projects. The use of subcontracting for the vast majority of the work in most construction and engineering projects could facilitate the application of EVM, but also makes it less necessary, as will be described further in this paper.
One might suggest that EVM has been adopted in the US defense/aerospace industries for the simple reason that the US government and its agencies, such as NASA, have required its use. However, that simply begs the question: Why is EVM considered so valuable in that context that its use has been mandated?
The full answer to that question is beyond the scope of this paper; however, it is valid to recognize that defense/aerospace projects are highly complex, very lengthy, and subject to constant change in requirements during their life cycle. EVM provides a useful methodology for maintaining measures of the planned work, the progress achieved, and the use of resources. Due to the highly dynamic environment, obtaining those measures would be difficult without the inherent logic of EVM.
That said, the success of EVM in the defense/aerospace industries does not negate the possibility that improvements could be made to EVM that not only make it more valuable there, but also in other industries that have shown less enthusiasm to date.
EVM might be seen as primarily a technique to measure performance in relation to plan, then forecast project cost results based on those measurements. However (as noted above) it is also a risk management technique. By determining variances and trends through EVM, the project manager is also establishing the probability that the project will exceed its cost budget or time schedule, and degree to which it could do so.
Procurement also serves as a risk management technique. Through contracts with reputable vendors, the project manager transfers risk related to the delivery of specific work packages. This paper therefore addresses potential links between EVM and procurement, given the similarity of content and aims.
Procurement and Cost Control
In many industries, such as engineering and construction, the majority of a project is implemented through a series of contracts with independent vendors and contractors. A general contractor might have many employees, but most would be supervisors, estimators, and managers. Very few of them would perform any physical work on a construction site.
Such projects might be termed high-procurement projects. If the contracts are firm fixed price agreements, then we might also consider these to be one with a high degree of certainty (low risk) to the purchaser. The combination of those two factors points towards the use of contract management as the best means for assessing progress and expenditures to date, in relation to the project scope, schedule and budget – the project plan.
With contract management, the project may be seen as primarily an assembly of contracted work packages. Great effort is taken to ensure that the project design allows for the separation of the work into packages that can readily be assigned to specialist consultants, suppliers, installers and other vendors.
In building construction and land development, budgets may be structured so that the cost categories map directly to the trade groups and suppliers that will perform the work. In that way, as quotations and bids are received, they can be readily compared with the corresponding amount in the cost budget.
In this environment, cost and schedule status is determined largely by comparing:
- The agreed contract amount with the approved budget amount.
- The actual contract start date with the planned start date for that work.
- The actual contract end date with the planned end date for that work.
- The total of all contract amounts with the total of all budget amounts for that work.
- The degree of progress on a contract with the expected percent complete according to the plan.
The contract management approach is not appropriate for all project types, and would be totally inappropriate for one that is being implemented fully with internal resources, with no vendor contracts.
Procurement and EVM
EVM measures only the value of work that has been completed up to a given point in the project. EVM establishes the earned value of all work, both internal activities and external activities that have been procured or purchased for the project. But EVM assesses the value of vendor contracts only if the work has been completed, or is at least underway.
Procurement practices allow organizations to obtain a degree of certainty regarding packages of work (contracts) that vendors agree to provide within specific constraints. Normally, those contracts are established well in advance of the execution of the work. Many organizations use the cost of those contracts as a means of confirming the expected total cost of the project, well before it is completed.
One might therefore expect that some published material on EVM would have addressed the use of procurement to provide assurance on future costs, and thereby reduce risk and improve cost forecasting accuracy. Although there is a significant body of literature related to EVM and to project procurement, no paper that links the two methodologies was discovered in the preparation of this paper. It therefore presents a number of concepts that build on both EVM and procurement practices, but otherwise appear to be original.
The central proposition is that contracts that have been signed, but not completed should be carefully considered in the calculation of the expected total project cost. The value of fixed-price contracts represents work that will be completed (with a high degree of certainty) for a predetermined cost by a future point in time. Other contracts, such as cost-plus or design-build types, offer less certainty, but should still be considered in forecasting the EAC.
The anticipated value of those contracts will be termed Assured Value. The associated concept and process will therefore be Assured Value Analysis (AVA).
This paper will demonstrate that Assured Value Analysis, in combination with EVM, allows the calculation of indices, variances and estimates at completion that are more meaningful than those available from earned value techniques alone. Before introducing AVA, a brief review of earned value methodology is provided, with comments related to its ability to provide useful indicators of future performance.
Earned Value Methodology
Earned Value Measures:
Earned value management established three measures (earned value, actual cost and planned value) and uses those to calculate variances and indices for both cost and schedule, and to forecast the total estimated cost at completion for the project, before it is completed. These are well known, established and accepted, (PMI, 2004).
Planned Value (PV): What did we plan to achieve by now?
In the past this was called the Budgeted Cost of the Work Scheduled (BCWS). This is defined as the portion of the approved cost estimate planned to be spent on the activity during a given period. Planned Value includes both direct and indirect costs. The total Planned Value for a project is termed the Budget at Completion (BAC). This can include a portion of the budget for activities that should have been partly completed.
Earned Value (EV): What have we achieved so far?
This has been called the Budgeted Cost of the Work Performed (BCWP). This is a portion of the total budget equal to the portion of the work actually completed. This an include an estimated value for partly completed activities
Actual Cost (AC): What have we spent so far?
Previously called the Actual Cost of the Work Performed (ACWP), this is the total of direct and indirect costs incurred in accomplishing work on all activities that have been started in a given period.
These can be used to measure the value and cost of both internal work performed by staff, and external work performed by vendors. Many organizations find it easier to tabulate vendor costs rather than internal costs, and to determine the budget amounts for vendor contracts rather than internal work packages. This is due to the fact that vendor contracts typically attract more scrutiny than internal work, with their associated bidding, negotiations and legal agreements. In many organizations, the cost of the work undertaken by internal staff, even that related to projects, is not estimated in advance nor tracked accurately as it is expended.
An extremely simple project will be used illustrate the established EVM measures, variance, indices and calculation of EAC. Assured Value Analysis will then be applied to the same situation.
Assume a project to install 100 computers with a Planned Value of $2,000 each, over a period of a year. The total project budget or BAC is $200,000.
After three months, it is found that 20 computers have been installed, at a cost of $44,000. According to the plan, 25 should have been installed by now. Therefore PV=50,000, EV=40,000 and AC=44,000; all are expressed in dollars. This simple situation is shown in the “Sample Project” diagram below.
One can readily calculate the EVM cost variances and indices:
CV = EV − AC = 40,000 − 44,000 = − 4,000
CPI = EV / AC = 40,000 / 44,000 = 0.9091
So this project is over budget, with a cost variance of negative $4,000 and a CPI that is tracking below 1. It is not performing well.
EVM theory holds that if this project continues without any significant changes to its efficiency or rate of progress, then it will eventually complete as shown in the diagram below. The project will end when all activities are finished, that is when EV = PV = 200,000. Applying conventional EVM theory, the Estimate at Completion (EAC) can be calculated using the “realistic” formula, as follows:
EAC3 = BAC / CPI = 200,000 / 0.9091 = 220,000
This projection is illustrated in the “EVM Estimate at Completion” diagram below.
So EVM predicts that this project will finish late, and $20,000 over budget. That is not a certainty, just a forecast based on the notion that the project will continue to unfold much as it has to date, and that future work is not all that different from the work that has taken place so far. It also tacitly assumes that we have no other useful information on which to base our forecast.
Introducing Assured Value Analysis
This section introduces Assured Value Analysis as a new concept, and demonstrates how it may be used to generate meaningful projections.
New Assured Value Measures:
The accepted EVM measures (EV, PV and AC) are effective in assessing what has taken place so far in a project in relation to its performance measurement baseline. However, to determine with greater certainty where a project is headed, two additional measures are proposed for the first time in this paper:
Assured Value (AV): What is the value of signed future contracts?
This might be called the Assured Value of Future Contracts. This is the total of all of the budget amounts for project work that will be performed under future contracts. It includes the balance of the value of contracts that have commenced but not been completed.
Expected Cost (EC): What will we spend on signed future contracts?
This might be called the Expected Cost of Future Contracts. This is the total cost of all vendor contracts for project work that will be performed in the future. It includes the balance of the cost of contracts that have commenced but not been completed.
Both of these measures relate to work that will be preformed by vendors in the future, such as consulting services, equipment supply and installation, construction of facilities, and similar purchased goods and services.
The choice of words here deserves some explanation. In both cases, the term “signed” is used rather than “executed” to avoid confusion – since “executed” could also mean completed or implemented. The phase “future contract” means any agreement with an external supplier, vendor, contractor, etc. that covers the provision of a portion of the project activities and deliverables to the project owner in the future. It should not be confused with “futures contract”, a term used in finance to describe rights or options to purchase a quantity of goods at a point in the future. In the balance of this paper, the term “future contract” implies one that has been negotiated, carries a firm fixed price, and has been approved by the purchaser as the project owner.
Assured Value (AV) represents the value of signed contracts that will be completed in the future. It is “assured” because the contract provides the project manager with a high degree of assurance (but not total certainty) that the budgeted value will be delivered according to the agreement. The “value” of the work is equal to the budget amount (the Planned Value) contained in the project cost baseline.
Expected Cost (EC) represents anticipated actual costs for vendor contracts. It is “expected” because the agreed contract price is what the project owner as buyer expects to pay the vendor for the specified work or material.
In selecting these new terms, care was taken to avoid abbreviations that might be confused with existing terms in project management or accounting. For example, PV is now used as an abbreviation for Planned Value, but it also represents Present Value, a term used in discounted cash flow calculations.
Why is there no AVA measure to represent planned value of a future contract? That is because the Planned Value is already established in the project budget. The Planned Value of a future contract is the sum of the Planned Values of all of the work packages that it comprises.
Once the work under a future contract has been completed, the work package(s) that it covers will be evaluated using the standard earned value measures. The Assured Value of that contract becomes the Earned Value, and the Expected Cost becomes the Actual Cost.
What if there are changes to the project that will affect these future contracts? Of course, the final cost of future vendor work might turn out to be different than the initial contract price, due to agreed changes to the contract. In the event of such changes, the Expected Cost for those contracts will need to be adjusted. At the same time, the project's change control administration should ensure that the Planned Value of the related work packages, are also adjusted to reflect those changes. Changing the Planned Value of a work package will automatically change the Assured Value of any contract for work in that package.
Calculating the Assured Value and the Expected Cost
It is proposed here that the following methods be used for determining the Assured Value (AV) of a vendor contract:
- Contract not started: AV is equal to the PV of the work package(s) that comprise the future contract.
- Contract completed: AV is zero, as the work packages covered by that contract are measured as EV.
- Contract in progress: AV is calculated as a portion of the PV for a contract in progress, based on the remaining percentage after the EV has been allocated.
For a contract in progress, a “percent complete” figure is typically calculated and reported for the purpose of determining contractor draws. In those cases, the AV is the “percent incomplete”. The key consideration is that EV + AV = PV for any contract underway.
The following methods are proposed in Assured Value Analysis for determining the Expected Cost (EC) of a vendor contract:
- Contract not started: EC is equal to the total agreed contract price of the contract, agreement, purchase order, etc.
- Contract completed: EC is zero, as the cost of the work packages covered by that contract (as amended) is measured as AC.
- Contract in progress: EC is calculated as a portion of the total agreed contract price for a contract in progress, based on the remaining percentage after the AC has been allocated.
The key is to calculate the AC first, then subtract that from the Total Agreed Contract Price (TACP). For any contract underway, the TACP = EC + AC.
Proposed AVA Variances and Indices
These two new Assured Value Analysis measures – Assured Value (AV) and Expected Cost (EC) – prove to be advantageous in calculating new variances and indices to evaluate the performance of projects.
Future Cost Variance (FCV):
This proposed new AVA measure for future cost variance compares the value of the work covered by future contracts with the expected cost of those contracts. This simple formula results in a positive value when the Assured Value of future contracts exceeds their Expected Cost.
- Future Cost Variance is Assured Value less Expected Cost FCV=AV − EC
Total Cost Variance (TCV):
The standard EVM cost variance formula ignores the value of any vendor contracts that have been successfully negotiated, but not completed. Those contracts could conceivably cover the majority – or almost all – of the remaining work. It would be useful to take them into account when reporting variance from the budget – particularly if the total of all those contracts is significantly less than the amounts budgeted for them.
Using Assured Value Analysis, it is possible to combine the existing CV formula with the new FCV formula (described above) to create a more effective assessment of the project cost performance:
- Total Cost Variance is Earned Value and Assured Value less both Actual Cost and Expected Cost, or
- Total Cost Variance is Cost Variance to date plus Future Cost Variance
TCV = EV + AV − AC − EC
Since EV − AC = CV and AV − EC = FCV, then TCV = CV + FCV
This slightly more complex variance formula results in a positive value when the value of achievements plus future contracts exceeds the cost of work to date plus the cost of those future contracts.
It can also be seen that poor performance to date can be compensated by the assurance of good results in future vendor contracts. This is a significant advantage of this proposed TCV formula over the accepted CV formula. It would be rather counterproductive to alert management and to take action to “rein in costs” based on a negative CV to date, if that variance is more than compensated by a positive FCV for future contracts. By the same token, it would be unfortunate if one was lulled into a false sense of security by a positive or nil CV value, without taking into account the expected cost of future contracts in comparison with the budget line amounts for that work.
Future Contract Performance Index (FCPI):
It is proposed that the new AVA Future Contract Performance Index (FCPI) will provide a ratio between the value of the work under future contracts, and the expected cost of achieving those results.
- Future contract performance index is the assured value of confirmed future contracts, divided by their expected costs FCPI = AV / EC
This simple formula results in a value below 1.0 when the Expected Cost of future work exceeds the Assured Value, and over 1.0 when those Expected Costs are less than the Assured Value.
At this point, it would be useful to return to the simple example that was used in the EVM review. Additional information on future contracts will be considered.
What is the result if we know that half of the computers will be supplied and installed by a reputable vendor during Q3 and Q4, and that a contract has been signed that requires them to perform that work at a cost of $1800 per computer? That means that the current crew will install only a quarter of the computers – which is good news, given their poor performance to date.
Calculating the Assured Value measures:
EC = 50 x 1800 = 90,000 and AV = 50 x 2000 = 100,000
With these two AVA measures, the cost variances can be further defined:
FCV = AV − EC = 10,000 and TCV = CV + FCV = − 4,000 + 10,000 = 6,000
These AVA variances are encouraging indicators – both are positive. The total cost variance (TCV) indicates that the project might actually be completed below budget. In contrast, the CV alone had indicated that performance to date was over budget.
Calculating the Future Contract Performance Index:
FCPI = AV/EC = 100,000/90,000 = 1.11
As this is greater than 1.0, it is evident that the vendor-performed work in the balance of the project will tend to compensate for the poor performance of the internal staff to date.
Forecasting with Assured Value Analysis
EVM variance indicators and indices are useful in identifying project situations that may require the attention of the organization's leadership – assuming that senior executives can interpret those indicators properly. There is no substitute, however, for a new and higher estimate of the project cost in attracting executive attention. Simply put, management wants to know: How will it all end? This seems to put the project manager in the position of predicting the future, which is a risky proposition at best.
Of course, a forecast of future results is not a prediction that they will necessarily occur. It is a warning that those results might occur if action is not taken to change the project course.
EVM theory holds that the project's direction can be revealed by the results from the early stages with enough accuracy that the final total cost can be reliably predicted early in the project. Final results are actually determined not by the estimate at completion, but by a number of other factors.
The quality of the project plan will have a major impact on the success of the project team. Spurred by the dire predictions of negative cost variances, a team might revisit their plan and make improvements to it. The actual performance results of the team may well improve after the early EVM results are published. Finally, the management team may decide, given the indicators, to rescue the project by providing more highly qualified staff, more effective equipment, or other improvements that will increase efficiency and results.
EVM forecasting takes three variables into account: the value of the work remaining, a performance efficiency factor, and the total actual costs to date. AVA adds the missing ingredient: the value and cost of future signed contracts.
Calculating the EAC with Assured Value Analysis
Calculating the Estimate at Completion with Assured Value Analysis will result in greater accuracy, since it takes into account the value of work packages that will occur in the future, for which the cost is already know with a good degree of certainty.
AV EAC 1: Actual and expected costs plus a new estimate
This approach assumes that the original estimating was seriously flawed, and would be an appropriate choice if the Actual Costs have consistently exceeded the Planned Value for work to date. In simple terms, the approach is:
- Take your actual costs to date, plus the expected costs for all future contracts, plus a new estimate for any remaining work that is not part of a future contract.
AV EAC1 = Actual Cost + Expected Cost + New Estimate for Balance of Scope
This expression is not a formula; one must prepare a new cost estimate for the remaining work on the project that is not covered by any future contracts. The new EAC could well be significantly higher than the original BAC.
This approach would not produce results that are very different than the corresponding pessimistic EVM EAC1 approach. In both cases, the project manager would take signed future contracts into account when preparing a new estimate for the balance of the work. The key difference is that this expression identifies those expected costs as being significantly different than other future non-contracted costs – over which there is much less certainty.
AV EAC 2: Actual and expected costs plus the remaining budget
This optimistic approach assumes current cost variances (if any) are atypical, and that the cost estimates for the balance of the project are valid and reliable.
- Take your actual costs to date, plus the expected costs for future contracts, then add the budget for remaining work that is not part of a future contract.
Since Earned Value and Assured Value represent the budgeted costs for the work that either has been or will be achieved, then the budget for the remaining work is the total budget less those two values. This leads to a formula that is easy to calculate: AVEAC2 = Actual Costs + Expected Costs + (Total Budget – Earned Value – Assured Value)
AVEAC2 = AC + EC + BAC − EV − AV
At first, this seems rather complicated; however, since CV = EV − AC and FCV = AV − EC then the above formula can be written: A VEAC2 = BAC − CV − FCV
This might be slight more logical to remember. Basically it states that the total project cost will be the original budget less the cost variance to date, and also less the future cost variance.
Since TCV = CV + FCV then the above formula can also be written: AVEAC2 = BAC − TCV
Put another way, with this approach the total expected project cost is the original budget less the total of the current and future cost variances. This is an optimistic approach because it assumes variances are atypical, and that the existing budget amounts are still valid.
This approach will produce results that are different from the standard EVM optimistic approach (EAC2) because it takes into account any variances that will exist between the value and the cost of future contracts.
Assured Value EAC 3: Assured Value CPI EAC
This approach builds on the Cumulative CPI EAC method, but also takes into account the value of contracts that have been negotiated and formalized, but not completed. We begin with:
- Add your cost to date to the expected cost of future contracts, and then add the remaining work divided by performance to date.
AVEAC3 = Actual Costs + Expected Costs + (Total Budget - Earned Value – Assured Value) / CPI AVEAC3 = AC + EC + (BAC − EV − AV) / CPI
This formula can also be further simplified. Since CPI = EV/AC then AC = EV/CPI
Substituting, we obtain: AVEAC3 = EV/CPI + EC − BAC/CPI − EV/CPI − AV/CPI
By simplifying and rearranging the results, we obtain: AVEAC3 = (BAC − AV) / CPI + EC
This final formula can be restated in words as:
- Take the total budget, subtract the planned value of future contracts, divide that by the cost efficiency to date, and then add the expected costs for those future contracts.
Does this formula pass a reality check? Yes, logically, it makes sense. We are saying that the total project cost will be composed of two items:
- The first one – the total budget less planned amounts for any future contracts – either has been or will be affected by the cost efficiency to date.
- The second one is simply the expected cost of those contracts. They are not affected by the cost efficiency factor because we already know what the agree cost is for the contract, and what value we had already assigned to it in the budget.
This approach should be more reliable than the conventional EVM approach for CPI EAC, which assumes that future work will be implemented just as efficiently as past work. That might be the case, or it might not, depending on the stage at which the EAC is prepared. In any event, it seems reasonable to consider the known cost and value of future contracts as reliable information, rather than to ignore them as the EVM approach suggests.
What is the Estimate at Completion using Assured Value Analysis? Two different formulae can be applied to our example:
AVEAC2 = BAC − TCV = 200,000 − 6,000
AVEAC2 = 194,000
This assumes that the internal crew have gotten over their difficulties, and will install their remaining 25 computers according to plan. Therefore, the project will come in $6,000 under budget – an amount equal to the total cost variance.
AVEAC3 = (BAC−AV) / CPI + EC = (200,000 − 100,000) / 0.9091 + 90,000
AVEAC3 = 110,000 + 90,000 = 200,000
This assumes that the internal crew will continue to under-perform, but the project will get back on budget due to the positive future cost variance produced by our external contract, per the “Assured Value Analysis” diagram below.
Of course, outsourcing is not always good news. If the vendor contract was for $2,400 per computer, then the situation becomes even worse than predicted by the use of the internal forces. But that is precisely the point; the more information is available on the procurement arrangements for the balance of the contract, the better prepared we can be to accurately predict project total costs at completion.
Calculating the Estimate to Complete with AVA
Conventional EVM provides two methods for preparing a cost estimate to complete (ETC) a project.
The EVM optimistic approach assumes that the cost to complete the project will be covered by the budget for the remaining work.
Optimistic EVM ETC = BAC − EV
The EVM realistic approach assumes that the cost to complete the remaining work will be subject to the same efficiencies (or inefficiencies) experienced with the completed work to date.
Realistic EVM ETC = (BAC − EV) / CPI
With Assured Value Analysis, these Estimate to Complete approaches become slightly more involved.
The AVA optimistic approach assumes that the cost to complete the project will be covered by the budget for the remaining work that is not covered by a future contract, plus the expected cost of those contracts.
Optimistic AVA ETC = BAC − EV − AV + EC
The AVA realistic approach assumes that the cost to complete the remaining work that is not covered by a future contract will be subject to the same efficiencies experienced to date.
Realistic AVA ETC = (BAC − EV − AV) / CPI + EC
Applying these ‘Estimate to Complete’ approaches to our simple example, one finds the following results:
Optimistic EVM ETC = BAC − EV = 200,000 − 40,000 = 160,000
Realistic EVM ETC = (BAC − EV) / CPI = 160,000/0.9091 = 176,000
Optimistic AVA ETC = BAC − EV − AV + EC = 200,000 − 40,000 − 100,000 + 90,000 = 150,000
Realistic AVA ETC = (BAC − EV − AV) / CPI + EC = (60,000)/0.9091 + 90,000 = 66,000 + 90,000 = 156,000
In both cases, the AVA approach estimates a lower “to complete” figure, as it takes into account the positive performance figures generated by the vendor arrangements for half of the work.
Calculating the Variance at Completion
Conventional EVM provides this formula for forecasting the cost variance at the completion of the project:
VAC = BAC − EAC
With Assured Value Analysis, the same formula is valid – though different values would be entered for the EAC.
In presenting the EAC for a project to its sponsor, executive or client, it would be useful to indicate the level of certainty that can be attached to that EAC. Forecasts and projections are just that, and some are rightly viewed with a fair degree of skepticism – or outright disbelief. Conventional EVM theory does not include an indicator of the level of certainty that can be attributed to an EAC figure.
The degree of certainty for a conventional earned value management EAC calculation is largely dependent on the percentage of the project that is completed. That is, as the earned value approaches the planned value, there is less likelihood that variances for the remaining work will be significant in relation to the total costs. In simple terms: the more that gets done, the less there is that can go wrong.
This paper proposes an expression be created to express level of confidence, called the Certainty Factor (CF), based on the EV divided by the BAC. As the project nears completion, the Certainty Factor associated with the EAC forecast will approach 1.0. A figure of 0.23 indicates less certainty, as the project is only 23% complete – and there is a great deal of opportunity for costs to get out of control.
Certainty Factor for a conventional EVM EAC: EVCF = EV / BAC
With Assured Value Analysis, greater certainty can be indicated at an equivalent point in the project. We obtain assurance not only from the work that has been completed, but also from future work that is covered by procurement contracts.
Certainty Factor for Assured Value Analysis EAC: AVCF = (EV + AV) / BAC
It should be noted that the certainty of an EAC calculation can also be dependent on many other factors, including the degree of variation in the CPI values to date, the track record of the project team in estimating costs for similar projects, and the extent of risk evident in the project.
Applying the proposed Certainty Factor measure of confidence to our initial EVM projections, it is found that CF = EV / BAC = 40,000 / 200,000 = 0.20. Therefore one would have a very low level of certainty concerning the EVM forecasts. How confident should one be about both of the AVEAC calculations? Applying the proposed AVA Certainty Factor, one finds that:
AVCF = (EV + AV) / BAC = (40,000 + 100,000) / 200,000 = 0.70
This shows that a much higher level of confidence can be justified for either of the two AVEAC forecasts, than was possible with the EVM EAC forecast.
Contract Risk and Expected Cost
This paper has proposed that the total agreed contract price be used to represent the Expected Cost of the work package(s) that contract comprises. This is based on the notion that a firm fixed price (FFP) or lump sum contract provides a high degree of certainty on those future project costs. This leads to the question: What if any Expected Cost should be assigned to the work package(s) covered by a contract that carries a higher degree of purchaser risk, such as a cost plus fixed fee (CPFF) contract? Similarly, what Assured Value should be ascribed to work packages covered by such contracts?
These are valid questions for study, but beyond the scope of this document, which seeks to introduce Assured Value Analysis, rather than overly complicate the discussion at this time.
That said, the following approach to different contract types – and therefore differing levels of purchaser risk – might be suggested.
- For a firm fixed price contract with a highly reputable vendor, or any lump sum vendor agreement secured by a performance bond, completion insurance or similar security, use the agreed contract price as the Expected Cost.
- For a unit price or cost-plus contract containing a maximum upset price provision, signed with a reputable vendor, use the maximum upset price as the Expected Cost.
- For all other contracts, do not ascribe any Expected Cost or Assured Value, as the degree of uncertainty is excessive.
Earned Value Management has been demonstrated an effective tool, particularly in the management of large, complex projects possessing a high degree of uncertainty and change. On the other hand, it has not been accepted in many other industries, such as construction, where projects are typically delivered though the use of fixed price contracts.
Assured Value Analysis builds on that concept by identifying two new measures – Assured Value and Expected Cost – that can be employed to develop very useful extensions to EVM. Those measures permit the development of cost variance, cost performance index and estimate at completion formulae that recognize the certainly that is achieved through the use of firm contracts for future project deliverables.
The use of AVA can provide not only greater certainly in calculating an estimate at completion, but also reduce the likelihood that senior management might receive false warnings of a potential cost overage in a situation that did not, in fact, represent a significant risk.
The potential benefits of AVA should be investigated further, and demonstrated thought the study of suitable projects possessing a mix of internally-delivered results and a significant share of subcontracted deliverables.
Anbari, F. T. (2003) Earned value project management method and extensions, Project Management Journal, 34(4), 12-23
Christian, D. S. & Ferns, D. V. (Spring, 1995) Using earned value for performance measurement on software development pojects, Acquisition Review Quarterly, (Spring) 155-171.
Fleming, Q. W. & Koppelman, J. M. (1996) Earned value project management, Project Management Institute, Newtown Square, PA.
Kim, E., Wells, W. G. & Duffery, M. R. (2003) A model for effective implementation of earned value management methodology, International Journal of Project Management, 21(5) 375-382.
PMI (2000) A Guide to the Project Management Body of Knowledge, Newtown Square, PA: .Project Management Institute, Inc.
PMI (2004) Practice standard for earned value management exposure draft, Newtown Square, PA: .Project Management Institute.
Prentice, J.-P. (2003) Earned value in the construction and facilities maintenance environments, AACE International Transactions, CS171.
True, L. (2003) Do you really know your construction costs?, Journal of Construction Accounting & Taxation, 13, 5-10.
© 2004, Douglas Bower
Originally published as a part of 2004 PMI Global Congress Proceedings – Anaheim, California