Concept Explainer

Earned Value Management: Formulas, EAC, Examples

Earned value management turns three numbers, planned value, earned value and actual cost, into a forecast of what a project will finally cost. This guide carries the formula set, a worked week-6 readout on a $120,000 build, and a matrix showing which of the four EAC formulas your situation calls for.

Project Management 9 min read Updated Aug 2026

Earned value management converts three numbers into a cost forecast: what the work was budgeted to cost by now, what the finished work was worth, and what it actually cost. In February 2026, SAE International published EIA-748 Revision E, cutting the governing standard from 32 guidelines to 27.

EVMS Guidelines
27
EIA-748-E, down from 32
Compliance Trigger
$20M
DFARS 234.201 cost contracts
High-Risk Guidelines
16
Named in DFARS 234.201(7)

The arithmetic did not change. The compliance scaffolding around it did, and most published explainers still describe the 32-guideline version retired this year.

What is earned value management?

Earned value management, usually shortened to EVM, is a performance measurement method that integrates scope, schedule and cost into a single set of indicators. It answers one question that a budget report alone cannot: is the money spent buying the work it was supposed to buy?

A conventional cost report compares budget to spend. If a team has burned 60% of a $200,000 budget, that report says the project is on track. It says nothing about whether 60% of the work exists.

EVM closes that gap by assigning a dollar value to completed work. That value, earned value, is the pivot of the whole method. Everything else is a ratio or a difference built from it.

The technique originated in US defence contracting. The Department of Defense's 35 Cost/Schedule Control Systems Criteria were rewritten by an industry committee into 32 guidelines and formalised as ANSI/EIA-748 in 1998. That standard is now stewarded by SAE International with the NDIA Integrated Program Management Division.

What are EVM's three core inputs?

Four terms carry the entire system. Three are measured each reporting period. One is fixed at baseline.

Budget at completion (BAC) is the total authorised budget for the work. It is the denominator behind most percentage calculations and it does not move without a formal baseline change.

Planned value (PV) is the budgeted cost of the work scheduled to be finished by the status date. On a $120,000 project planned to run evenly across 12 weeks, PV at week 6 is $60,000.

Earned value (EV) is the budgeted cost of the work actually finished. If only 42% of the scope is complete at week 6, EV is $50,400, regardless of what was spent getting there.

Actual cost (AC) is what the completed work genuinely cost, taken from the accounting system rather than from a progress claim.

The most common student error is treating EV as a spend figure. It is not. EV is priced at budgeted rates, which is exactly why comparing it against AC isolates cost efficiency from scope change. Both the ENGG5203 cost baselines chapter and most PMBOK-aligned units teach this distinction before any formula is introduced.

How is earned value calculated?

Every EVM formula falls into one of four families: measurement, variance, index, and forecast. The full set fits on one page.

Metric Formula What it tells you
Planned value BAC × planned % complete Work the baseline promised by now
Earned value BAC × actual % complete Budgeted worth of finished work
Schedule variance SV = EV − PV Dollars of work ahead or behind
Cost variance CV = EV − AC Dollars over or under budget
Schedule performance index SPI = EV ÷ PV Rate of progress vs plan
Cost performance index CPI = EV ÷ AC Work bought per dollar spent
Percent complete EV ÷ BAC Scope delivered to date
Percent spent AC ÷ BAC Budget consumed to date
Estimate to complete ETC = EAC − AC Money still required
Variance at completion VAC = BAC − EAC Projected final overrun
To-complete index (to BAC) TCPI = (BAC − EV) ÷ (BAC − AC) Efficiency needed to still hit budget
To-complete index (to EAC) TCPI = (BAC − EV) ÷ (EAC − AC) Efficiency needed to hit the revised forecast
Two subtractions, six ratios, one forecast family. Source: ANSI/EIA-748 common terminology, EIA-748-E, February 2026.

Signs are consistent across the whole set. Negative variance is bad. An index below 1.0 is bad. A CPI of 0.80 means every dollar spent purchased 80 cents of budgeted work.

TCPI is the least-taught formula and the most useful in an exam. It inverts the question: instead of reporting how efficient you have been, it states how efficient you must become. When TCPI exceeds CPI by more than about 10%, the original budget is usually unrecoverable.

From AskSia's online library
The PMGT5889 Project Controls Course Bible in AskSia's library maps how EVM is actually assessed rather than defined. Across its 12 chapters, earned value sits at Chapter 9, deliberately placed after work breakdown structure, cost baselines and network logic. Chapters 10 and 11 then cover estimate at completion and variance diagnosis as separate topics. That ordering is the practical lesson: units that assess EVM almost never ask for a bare CPI. They give you a baseline, a status date and a change request, and mark you on whether your forecast method matches the situation. Running the chapter through AskSia's Concept Map shows which upstream chapter each formula depends on.

What does an EVM readout look like?

Take a 12-week office fitout with a $120,000 budget, scheduled to spend evenly. At the end of week 6 the project manager reports 42% of scope complete and $63,000 invoiced.

Metric Week 6 value Reading
Planned value $60,000 Half the budget, half the schedule
Earned value $50,400 42% of scope, at budgeted rates
Actual cost $63,000 53% of budget consumed
Schedule variance −$9,600 Roughly one week of work missing
Cost variance −$12,600 Overspend already exceeds the delay
TCPI to BAC 1.22 Needs 53% better efficiency than achieved
Variance at completion −$30,000 Using the CPI-based EAC of $150,000
The 1.22 TCPI is the number that ends the argument. Source: worked example, BAC $120,000, status week 6 of 12.
Cost · CPI
0.80
$50,400 ÷ $63,000 · 80c of work per dollar
Schedule · SPI
0.84
$50,400 ÷ $60,000 · 84% of planned pace

Read the gap between the two, not each in isolation. CPI sits below SPI, which means the cost problem is larger than the delay. Adding resources to recover schedule would make the worse of the two problems worse.

That diagnostic pattern is what postgraduate units assess. Monash's project management unit, covered in our FIT5057 study guide, awards every mark through coursework rather than a final exam, so the reasoning behind a number carries more weight than the number.

Which EAC formula should you use?

Estimate at completion is where most students lose marks. There are four standard formulas and they are not interchangeable. Each encodes a different assumption about the future.

Formula Assumption Worked result Fails when
AC + (BAC − EV) Overrun was a one-off $132,600 The cause is systemic, such as a rate error
BAC ÷ CPI Current efficiency persists $150,000 Early periods carry mobilisation cost
AC + [(BAC − EV) ÷ (CPI × SPI)] Delay drives further cost $166,571 Schedule and cost are independent
AC + bottom-up ETC Baseline no longer valid Requires a re-estimate you may not have time for
A $34,000 spread from the same inputs. The bottom-up figure is shown as "—" because it cannot be derived from indices. Source: worked example above.

The spread matters. Four defensible methods produce forecasts between $132,600 and $166,571 from identical data. Choosing one is a judgement, and examiners mark the justification.

Default to BAC ÷ CPI when nothing in the case suggests otherwise. It is the neutral choice and the one PMI-aligned syllabi treat as the base method.

Use the CPI × SPI variant only when the case describes a genuine coupling, such as extended site hire or standing labour. Applying it reflexively inflates the forecast and is a common marking penalty.

Where is EVM actually required?

EVM is not optional advice on US federal contracts. DFARS 234.201, current as of the 7 May 2026 change, sets three tiers by contract value.

Cost or incentive contracts at $20,000,000 or more must use an EVMS complying with ANSI/EIA-748. At $50,000,000 or more the system must be formally determined compliant by the cognizant federal agency, though DoD Class Deviation 2015-O0017 raised the compliance-review threshold to $100,000,000 and remains in force. Below $20,000,000 the decision is risk-based and must be documented.

One detail rewards close reading. DFARS 234.201(7) lists 16 high-risk guidelines by number, including guidelines 27, 28, 30 and 32. EIA-748-E now contains only 27 guidelines, so the regulation still cites a numbering scheme the standard retired in February 2026. Expect remapping guidance before the citations are corrected.

For students, EVM surfaces in three places: project management coursework such as the University of Sydney's School of Project Management units, professional certification, and graduate roles in project controls.

AskSia is one workspace rather than a stack of five. For a controls unit that means the lecture recording, the cost baseline spreadsheet, the textbook chapter and the practice questions sit in the same place, so a formula you meet in week 9 links back to the WBS you built in week 3 without reopening four tools.

Both the PMP and CAPM exams test EVM through situational items rather than bare calculation. Mock Exam mode reproduces that format and grades the reasoning, not just the arithmetic.

Frequently Asked Questions

How is EVM calculated?

EVM is calculated from three measured values and one baseline figure. Planned value is the budget at completion multiplied by planned percent complete. Earned value is the budget at completion multiplied by actual percent complete. Actual cost comes from the accounting ledger. From those, schedule variance is EV minus PV and cost variance is EV minus AC, while the two indices are EV divided by PV and EV divided by AC respectively. On a $120,000 project at week 6 of 12 with 42% delivered and $63,000 spent, that produces PV of $60,000, EV of $50,400, SPI of 0.84 and CPI of 0.80. The single most common error is calculating earned value at actual rates rather than budgeted rates, which collapses CPI to 1.0 in every period and makes the whole system report nothing. Check your percent-complete method against your unit's assessment brief before running any figures.

What is a good EAC formula?

There is no universally good EAC formula, only a formula that matches the situation. BAC divided by CPI is the default and assumes current cost efficiency continues, giving $150,000 in the worked example above. AC plus the remaining budgeted work assumes the overrun was a one-off and gives $132,600. The CPI multiplied by SPI variant assumes delay drives additional cost and gives $166,571. A bottom-up re-estimate applies when the baseline is no longer credible. That is a $34,000 spread from identical inputs, which is why exam markers award credit for the justification rather than the number. Pair the forecast with TCPI: in the worked case TCPI to BAC is 1.22 against a CPI of 0.80, meaning the team would need to run 53% more efficiently for the rest of the project. When that gap exceeds roughly 10%, state plainly that the original budget is unrecoverable.

What is an example of earned value analysis?

A 12-week office fitout budgeted at $120,000 illustrates the full cycle. At week 6 the baseline expected $60,000 of work complete, but only 42% of scope existed, worth $50,400 at budgeted rates, and $63,000 had been invoiced. Schedule variance is negative $9,600 and cost variance is negative $12,600. Because CPI of 0.80 sits below SPI of 0.84, the cost problem is larger than the delay, so adding crew to recover schedule would worsen the dominant issue. That reading, rather than the raw indices, is the deliverable. Real assessment items go further and introduce a scope change or a rate correction mid-period, which forces a decision about whether to rebaseline. Work through the cost baselines and earned value chapter in the AskSia library to see how a baseline change alters every downstream figure.

How do you get certified in earned value management?

Three routes exist. AACE International's Earned Value Professional credential is the specialist option, requiring 8 years of industry experience, or 4 years plus a four-year related degree, and it is built on the current version of ANSI/EIA-748. It stays valid for 3 years and is renewed with 12 continuing education units or by re-sitting the exam. PMI's PMP covers EVM as one component of a broader project leadership exam aimed at experienced practitioners. CAPM is the entry-level alternative for students and recent graduates with no project management experience requirement of that scale. For an undergraduate, CAPM is usually the realistic first step, with EVP a mid-career target. Check the current eligibility rules on the AACE and PMI sites directly, since experience requirements are revised periodically.

What is the difference between SPI and CPI?

Both divide earned value by something, but the denominators answer different questions. CPI divides EV by actual cost and measures purchasing efficiency: a CPI of 0.80 means each dollar spent bought 80 cents of budgeted work. SPI divides EV by planned value and measures pace against the baseline: an SPI of 0.84 means the project delivered 84% of the work it promised by that date. CPI is the more reliable of the two. Published research on defence programs has long observed that the cumulative CPI stabilises early and rarely improves materially thereafter. SPI carries a structural flaw: as a project finishes, earned value converges on planned value, so SPI drifts to exactly 1.0 even on a project delivered months late. Cross-check SPI against critical path float, or use the earned schedule method, before reporting schedule health late in a project.


Three limits are worth stating honestly.

SPI becomes useless near completion. Because both EV and PV converge on BAC, SPI returns to 1.0 on any project that eventually finishes, however late. The earned schedule method, developed by Walt Lipke in 2003, addresses this by measuring progress in time units instead of dollars, and it is increasingly taught alongside classic EVM.

EVM is also blind to quality. Work that is 100% complete and 100% defective earns 100% of its budgeted value. The system measures quantity of scope delivered, nothing more.

And it fits some contract types badly. DFARS explicitly discourages applying EVM to firm-fixed-price contracts of any value, because the cost risk sits with the contractor. Agile teams face a related mismatch: EVM presumes a stable scope baseline that iterative delivery deliberately keeps open.

None of this makes the method obsolete. It makes the method a diagnostic rather than a verdict. The full AskSia library carries the controls chapters for the units that assess it, and Sia Note compresses a 33-page Course Bible into the formula set and the decision rules that sit behind it.

Recommended

Study faster with AskSia

Turn course materials into clear notes, practice questions, and review plans.

Try AskSia