What Is a Cost Variance Report? Formulas, CPI, and Forecasts

A cost variance report is a project financial document that compares the budgeted value of work actually completed against the money actually spent to complete it, using the formula Earned Value minus Actual Cost. A positive result means the project is under budget for the work done so far. A negative result means it is over budget. The same report usually pairs that figure with schedule variance and a set of forecasting indices so readers can see not just where the project stands today but where it is heading.

The Three Inputs Every Report Depends On

Three numbers do the heavy lifting. If any one of them is wrong, nothing downstream can be trusted.

Planned Value (PV) is the authorized budget for the work that was scheduled to be complete by the reporting date. On a $500,000 project where the baseline says 40% should be finished by March 31, planned value on that date is $200,000. It comes from the approved baseline and does not move unless the baseline is formally revised.

Earned Value (EV) is the budgeted cost of the work that has actually been performed, regardless of when it was supposed to happen. If the team has completed 35% of the scope by March 31, earned value is $175,000. EV converts physical progress into dollars so it can be compared directly against both the plan and the spending.

Actual Cost (AC) is every dollar spent on the completed work through the reporting date: labor, materials, equipment, subcontractor invoices, and allocated overhead. If the team spent $190,000 to reach that 35%, actual cost is $190,000.

How finely these numbers are measured depends on the project’s Work Breakdown Structure. A detailed structure with many small work packages produces tighter variance data. A coarse structure with large packages can hide problems until they are expensive to fix.

The Core Calculation

Cost Variance answers the spending question: is the project over or under budget for the work completed?

CV = Earned Value − Actual Cost

Schedule Variance answers the progress question: is the work ahead of or behind the plan?1U.S. Department of Energy. Earned Value Management Module 6 – Metrics, Performance Measurements and Forecasting

SV = Earned Value − Planned Value

Both anchor on earned value. A positive CV means less was spent than budgeted for the completed work. A negative CV means more was spent. A positive SV means more work has been finished than scheduled. A negative SV means the project is behind.

A Worked Example

Take the $500,000 software implementation above. At the end of month three:

  • Planned Value: $500,000 × 40% = $200,000
  • Earned Value: $500,000 × 35% = $175,000
  • Actual Cost: $190,000

The variances fall out directly:

  • Cost Variance: $175,000 − $190,000 = −$15,000 (over budget)
  • Schedule Variance: $175,000 − $200,000 = −$25,000 (behind schedule)

Both figures are negative, which is the worst combination. The project has done less than planned and spent more than budgeted for what it has done. That calls for immediate investigation, not a wait-and-see.

Expressing Cost Variance as a Percentage

Raw dollars can understate or overstate severity. A $15,000 overrun on a $500,000 project is a different problem than the same overrun on a $50,000 one. The percentage form normalizes it:

CV% = (CV ÷ EV) × 100

In the example, CV% = (−$15,000 ÷ $175,000) × 100 = −8.6%. Many organizations treat variances within ±5% as healthy, flag anything beyond ±10% for formal review, and treat variances past ±15% as requiring immediate corrective action such as pausing work or reallocating resources.

Reading Cost Variance Alongside Schedule Variance

Cost variance in isolation misses half the picture. A project can be under budget and behind schedule at the same time, or over budget while running ahead of plan. The pairing tells the real story.

  • Positive CV, positive SV: under budget and ahead of schedule. The best scenario, though worth confirming the quality of completed work.
  • Positive CV, negative SV: under budget but behind schedule. The team is working efficiently but slowly. If the schedule is critical, adding resources may be justified even though spending is currently favorable.
  • Negative CV, positive SV: over budget but ahead of schedule. Often the sign of aggressive spending to hold pace, such as overtime or expedited materials. On contracts where delay penalties exceed the overspend, that trade-off can be deliberate.
  • Negative CV, negative SV: over budget and behind schedule. The most serious combination, typically requiring a formal corrective action plan and possibly scope or baseline adjustments.

Schedule variance has a known limitation worth flagging. It is measured in dollars, not time, so a negative SV does not tell you whether the delayed work sits on the critical path. A two-week slip on a non-critical task barely matters. A two-day slip on a critical-path task can push the completion date. Pair schedule variance with a critical-path analysis for the full picture.1U.S. Department of Energy. Earned Value Management Module 6 – Metrics, Performance Measurements and Forecasting

Turning Variances Into Forecasts

Variance figures describe what already happened. Performance indices describe where the project is heading if the current pattern holds. That forward look is where the report earns its keep, because catching a trend at month three is far cheaper than discovering a blowout at month nine.

Cost Performance Index and Schedule Performance Index

The Cost Performance Index measures how efficiently the project converts dollars into completed work:

CPI = Earned Value ÷ Actual Cost

A CPI of 1.0 means perfect efficiency. Above 1.0, every dollar buys more work than planned. Below 1.0, every dollar buys less. In the running example, CPI = $175,000 ÷ $190,000 = 0.92, meaning each dollar spent is producing only 92 cents of planned value.

The Schedule Performance Index works the same way for progress:

SPI = Earned Value ÷ Planned Value

An SPI below 1.0 means work is being completed more slowly than planned. In the example, SPI = $175,000 ÷ $200,000 = 0.875, or about 88% of the scheduled work rate.1U.S. Department of Energy. Earned Value Management Module 6 – Metrics, Performance Measurements and Forecasting

Estimate at Completion and Variance at Completion

Estimate at Completion projects the total cost of the project if current spending efficiency continues:

EAC = Budget at Completion ÷ CPI

With a $500,000 budget and a CPI of 0.92, EAC comes to roughly $543,478. That is the number leadership tends to care about most, because it answers the question everyone asks: how much will this project actually cost when it is done?

Variance at Completion extends the picture to the finish line: VAC = Budget at Completion − EAC. Here, VAC = $500,000 − $543,478 = −$43,478. The project is on track to finish roughly $43,000 over budget unless something changes.

To-Complete Performance Index

The To-Complete Performance Index flips the question. Instead of asking where the project is heading, it asks what efficiency the remaining work must achieve to hit the budget target:

TCPI = (Budget at Completion − Earned Value) ÷ (Budget at Completion − Actual Cost)

In the example, TCPI = ($500,000 − $175,000) ÷ ($500,000 − $190,000) = $325,000 ÷ $310,000 = 1.048. The team would need to perform about 4.8% more efficiently on all remaining work just to break even on the budget. Whether that is realistic depends on why the overrun happened in the first place. If CPI has been drifting steadily downward, a TCPI above 1.0 is a signal that the original budget may be unrecoverable.

Separating Direct and Indirect Cost Variances

Not every variance shows up the same way. Direct costs tie to a specific project task: the salary of an engineer assigned to the project, the lumber delivered to a construction site, the software license bought for a particular deliverable. Indirect costs benefit multiple projects or the organization broadly and get allocated rather than charged directly, such as office rent, administrative staff, and general IT infrastructure.2FEMA. Indirect vs. Direct Costs

Direct cost variances are usually easier to trace. If a work package budgeted 200 labor hours and the team logged 240, the overrun is visible line by line. Indirect variances are murkier because they depend on allocation methods. A spike in company-wide utility costs flows into the project’s overhead allocation even though nobody on the team caused it. Splitting the two in the report prevents misleading conclusions about project-level efficiency.

Federal grants require tracking both. Under federal cost principles, the total cost of a federal award is the sum of allowable direct costs and allocable indirect costs, minus any applicable credits.2FEMA. Indirect vs. Direct Costs

Why Variances Happen and What to Do

A variance number is a symptom. The report becomes useful when it drives investigation into the cause. Common culprits fall into recognizable patterns:

  • Scope changes. Uncontrolled additions inflate costs without matching increases to the baseline.
  • Labor rate differences. Assigning more senior (and more expensive) staff than planned, or incurring unplanned overtime.
  • Material price fluctuations. Market changes between the time the budget was set and the time materials were purchased.
  • Productivity shortfalls. Tasks taking more hours than estimated due to rework, unfamiliar technology, or coordination problems.
  • Estimation errors. The original budget was simply wrong, which is more common than most project managers want to admit, especially on novel work where historical data is thin.

Once the cause is identified, a corrective action plan should address both the immediate variance and the conditions that produced it. Fixing the symptom without fixing the process guarantees the same variance next month. Effective plans include a clear root cause statement, specific actions with owners and deadlines, and a way to verify the fix worked.

Corrective action should be proportional. A 3% overrun on a single work package might need nothing more than a conversation with the task lead. A 15% overrun across multiple packages probably calls for a formal review with resource reallocation and possibly a revised baseline.

Getting the Inputs Right

The math is straightforward. Clean input data is where most of the real work sits.

Planned Value comes from the approved baseline, usually documented in a signed contract or statement of work that maps budgeted costs to scheduled work packages across the timeline. If the baseline has been revised through formal change control, use the current version, not the original.

Actual Cost pulls from several sources: labor time-tracking and payroll for staff hours, vendor invoices and purchase orders for materials and subcontracted work, and general ledger reports for allocated overhead. Reconciling those sources against each other catches duplicate entries and missed charges. Every line item should trace back to a purchase order or employment agreement.

Earned Value is the hardest to measure honestly. It requires assessing physical progress on each work package and translating that progress into dollars using the baseline budget. Objective methods, such as milestone completion or units delivered, produce more reliable data than subjective percent-complete estimates. The temptation to round progress upward is constant, and it produces a pattern of favorable variances early in the project that reverse sharply near the end.

Most organizations feed these inputs into enterprise resource planning systems or specialized project management software that calculates variances and indices automatically. The software eliminates arithmetic errors but not garbage-in-garbage-out problems. Someone with financial expertise, whether a project controller or a CPA, should review the inputs and outputs before the report goes out.

Federal Contract Thresholds

If the project is a federal contract, particularly in defense, the report is not just a management tool. It is a compliance obligation. The governing standard is EIA-748, which contains 32 guidelines organized into five categories: organization; planning, scheduling, and budgeting; accounting; analysis and management reporting; and revisions and data maintenance.3Department of Defense. Earned Value Management System Interpretation Guide

As of February 2026, the dollar thresholds that trigger EIA-748 compliance were revised upward through a class deviation to the Defense Federal Acquisition Regulation Supplement:

  • $50 million or more: the contractor’s earned value management system must comply with EIA-748.
  • $100 million or more: the system must be formally validated by the cognizant federal agency as EIA-748 compliant.
  • Under $50 million: applying a formal earned value management system is optional, and the contracting officer makes a risk-based decision documented in a cost-benefit analysis kept in the contract file.
  • Firm-fixed-price contracts: applying earned value management is discouraged regardless of dollar value and requires a waiver.4Department of Defense. Class Deviation 2026-O0011 – DFARS Part 234 Earned Value Management

Among the 32 guidelines, Guideline 22 requires generating cost and schedule variance data at least monthly at the control account level, and Guideline 23 requires analyzing significant variances monthly.3Department of Defense. Earned Value Management System Interpretation Guide

Where Overrun Funding Comes From

When a report shows unfavorable results, the next question is where the money to cover the shortfall comes from. Two budget mechanisms exist for that purpose, and mixing them up is a common error.

Contingency reserves, sometimes called contract reserves, are set aside for specific identified risks within individual work packages or project segments. If the risk materializes, the reserve is drawn down and the funds move into the performance measurement baseline. If the risk does not materialize, the unused amount is returned to the project-level pool. These reserves sit inside the cost baseline.

Management reserve sits above the cost baseline and covers events that genuinely could not have been anticipated when the budget was approved. It is not tied to any specific work package. Accessing it typically requires formal approval, and drawing from it changes the project’s budget at completion. On federal contracts, management reserve is tracked and reported separately to maintain the integrity of the original baseline.

The distinction matters for variance reporting because drawing on contingency does not change the baseline, while drawing on management reserve does. A project that repeatedly taps management reserve is signaling that the original scope and budget were fundamentally misaligned, and that is a different kind of problem than individual work packages running hot.