Earned Value Formula: Worked Example in ₹ Crore and Programme Roll-Up

• The earned value formula needs three numbers at a status date (planned value, earned value and actual cost) and a budget at completion. With a ₹800 crore budget, 70% done and ₹640 crore spent, CPI is 0.875 and the cost forecast is ₹914.29 crore.
• For a programme, divide total earned value by total actual cost for CPI, and by total planned value for the schedule performance index (SPI). In our three-project example, averaging the project CPIs gives 1.058 while the true programme CPI is 1.00.
• Forecast each project's final cost separately and add the results (₹810.6 crore in our example), because one project's overrun does not shrink when another underspends.
• Report CPI against both the original and the revised budget. On the 1,981 MoSPI projects we calculated 1.04 against original costs and 1.28 against revised costs.

The earned value formula compares three ₹ amounts at a status date: planned value (PV), the budgeted value of the work scheduled by now; earned value (EV), the budgeted value of the work actually done; and actual cost (AC), the money spent. Add the budget at completion (BAC), the total approved budget. Everything else is arithmetic on these four.

PV is planned percent complete × BAC. EV is actual percent complete × BAC. Spend never enters EV: it measures value of work done at budgeted rates, whatever it cost.

Multiply budget by percent complete to get EV, then compare it with spend. With the defaults above, EV is ₹560 crore against ₹640 crore spent, so the CPI is 0.875: each rupee spent has bought 87.5 paise of budgeted work. The core formulas follow the NSF earned value gold card (published 15 October 2024, updated 5 June 2026).

The gold card gives CV, SV, CPI, SPI, EAC = BAC ÷ CPI, VAC and EAC = AC + ETC. The BAC-based TCPI above is our working from standard earned value practice: work remaining (BAC − EV) divided by cost remaining (BAC − AC), where the card's own TCPI divides work remaining by (EAC − AC).

Use BAC ÷ CPI as your default forecast, because it assumes the cost efficiency you have achieved so far continues. Choose another formula only when you can state a different assumption about the remaining work. Each formula below is EAC = AC + ETC, the card's definition, with a different ETC. The card writes the CPI-based and composite forms with work remaining, which we read as BAC − EV; the one-off variant AC + (BAC − EV) is standard earned value practice, shown here as our working. All figures are on the defaults.

In our judgement, the "one-off" row is the one people abuse. Use it only if you can name the cause and show it has ended. Where risks are logged, add their exposure as the last row does; our guide to programme risk management covers how programmes size that exposure.

A CPI or SPI of 1.00 means on plan, below 1.00 means behind or over, above 1.00 means ahead or under. The traps are in the units and in the late stages of a job.

• CV and SV are in rupees, not time. An SV of −₹40 crore means ₹40 crore of planned work is not done. Dividing it by the plan's spend per month gives a rough months-behind figure, but only where the plan is close to a straight line.
• SPI drifts to 1.00 at the end. When all work is complete, EV equals BAC and PV equals BAC, so SPI reads 1.00 even for a project that finished late.
• TCPI is a credibility test. Our example needs 1.50 on the remaining work to hit the original budget, against 0.875 achieved so far. When TCPI sits far above the CPI you have delivered, the original budget is a hope, not a forecast: reforecast.
• Set thresholds in advance. Which CPI or SPI triggers escalation is a governance decision; write it into the programme management plan and the portfolio performance scorecard so nobody argues about it after the number turns red.

Sum the EV, PV and AC of the projects first, then divide. The programme CPI is total EV ÷ total AC, and the programme SPI is total EV ÷ total PV. These are our derivations: the card's CPI and SPI definitions applied to programme totals. The figures below are illustrative, in ₹ crore.

Three things follow:

1. Averaging misleads. The mean of the three CPIs is 1.058 and the mean of the SPIs is 0.983, which tell the sponsor the programme is under budget and nearly on schedule. The true figures are 1.00 and 0.897.
2. Totals hide the problem project. Programme CV is zero, but project B is ₹30 crore over on work done, offset by A and C.
3. Forecast per project, then add. Programme BAC ÷ programme CPI gives ₹800 crore. The sum of project forecasts gives ₹810.6 crore, so the variance at completion is −₹10.6 crore. B keeps costing 1.25 times its budget whatever A and C do.

Applied to the 1,981 ongoing central-sector projects in the April 2026 MoSPI list, earned value gives a portfolio CPI of 1.04 on original costs and 1.28 on revised costs, so the choice of baseline changes the headline. We worked from Table 6 of the MoSPI Flash Report for April 2026, which covers central-sector infrastructure projects costing ₹150 crore and above. MoSPI does not publish a CPI; the figures below are ours.

How we calculated it (October 2026). This is a snapshot of the April 2026 report, the issue analysed, and later issues will differ. We extracted all 1,981 rows. Our totals match the report's headline: ₹37,12,662 crore original cost, ₹42,78,402 crore revised cost, ₹20,36,107 crore expenditure. For each project, EV = reported physical progress % × cost and AC = cumulative expenditure. The portfolio CPIs, the median cost and the median progress use all 1,981 projects. The median project CPI, the mean of project CPIs and the share below 1.00 use the 1,638 projects with both spend and progress above zero; the other 343 are excluded from those three figures. The median spend share uses the projects reporting spend above zero. The progress figures are reported to MoSPI by the ministries and implementing agencies, so these figures test method, not individual projects.

What this tells a programme controller:

• Never average indices. A handful of projects with almost no spend push the mean CPI to 79. Weighted by money, the same projects give 1.04.
• Re-baselining flatters. Moving from original to revised costs lifts the portfolio CPI from 1.04 to 1.28 without any work being done. Take Guwahati Airport's new terminal: ₹1,712 crore original, ₹2,520 crore revised, ₹2,627.39 crore spent at 98% progress. Against the original budget its CPI is 0.64 (CV −₹949.63 crore); against the revised budget it is 0.94 (CV −₹157.79 crore), with an EAC near ₹2,681 crore.
• A CPI above 1.00 is not always efficiency. Median spend (43%) trails median progress (70%). In our reading, that points to payments and expenditure booking lagging physical work, so check when spend is recorded before you celebrate a high CPI.

Our recommendation: report CPI against both the original approved baseline and the baseline after approved changes, and let the programme management plan say who may approve a change.

At programme level, cost control is one measure among several, and the failures that sink programmes are mostly outside the EV sheet: benefits that were never owned, risks shared across projects, stakeholders who moved. That is the ground our Program Management Essentials course covers: programme risk, stakeholder engagement, benefits planning and strategic alignment, across 10 modules, 12 weeks and 15+ hours of guided learning, with ready-to-use templates. As of October 2026 it is US$99 per enrolment, priced in US dollars; pricing varies by market.

We have mentored or trained 6,000+ professionals. If you are deciding whether to stay on the project path or take a programme seat, our free discovery consultation is available before enrolment through support@augmentconsultancy.com.