7.2 Traditional EV Performance Indices & Late-Project Quirks

7.2 Traditional EV Performance Indices & Late-Project Quirks

AACE Professional Context: Earned Value Management is a widely used integrated performance method, but its traditional schedule metrics—specifically Schedule Variance ($SV = EV - PV$) and the Schedule Performance Index ($SPI = EV / PV$)—harbor profound mathematical and structural limitations. On the PSP examination, planning professionals must not only know how to calculate traditional performance indices, but also possess the advanced analytical capability to identify when and why traditional EV metrics fail, recognize the 'convergence trap' near project completion, and explain why earned value must never be evaluated in isolation from CPM network critical path float.

---

Traditional Performance Indices: SPI, CPI, and the Critical Ratio

Performance indices convert absolute dollar variances into dimensionless ratios representing operational efficiency. These metrics allow cross-project benchmarking and serve as the mathematical foundation for statistical forecasting.

1. Schedule Performance Index ($SPI$)

$\mathbf{SPI = \frac{EV}{PV}}$

• Measures the efficiency of work accomplished relative to the baseline schedule plan.
• $SPI > 1.00$: Favorable. Work is being earned at a faster rate than planned.
• $SPI = 1.00$: On plan. Work earned exactly equals scheduled work.
• $SPI < 1.00$: Unfavorable. Work is being earned at a slower rate than planned (e.g., $SPI = 0.80$ implies earning $0.80 of work for every $1.00 scheduled).

2. Cost Performance Index ($CPI$)

$\mathbf{CPI = \frac{EV}{AC}}$

• Measures the financial efficiency of work accomplished relative to actual expenditures.
• $CPI > 1.00$: Favorable. Project is experiencing cost underrun efficiency (e.g., $CPI = 1.25$ indicates $1.25 of work earned for every $1.00 spent).
• $CPI = 1.00$: On budget.
• $CPI < 1.00$: Unfavorable. Project is burning money faster than earning work (e.g., $CPI = 0.80$ indicates $0.80 of work earned per $1.00 spent).
• Durable Rule of Thumb: Empirical research across defense and infrastructure megaprojects demonstrates that cumulative $CPI$ stabilizes once a project reaches approximately 20% completion; it rarely improves by more than 0.10 from that point forward without formal re-baselining.

3. Critical Ratio / Composite Index ($CR$)

$\mathbf{CR = CPI \times SPI}$

• Provides a composite health indicator balancing cost and schedule efficiency.
• If a project is behind schedule ($SPI = 0.85$) but significantly under budget ($CPI = 1.20$), its composite index is $CR = 1.20 \times 0.85 = 1.02$, indicating overall manageable project stability.
• If both indices lag ($SPI = 0.80, CPI = 0.80$), $CR = 0.64$, indicating critical project distress.

---

The Fatal Flaw: The Late-Project SPI Convergence Quirk

The most dangerous limitation of traditional EVM is the behavior of $SV$ and $SPI$ as a project approaches and passes its planned completion date. This phenomenon is frequently tested on the AACE PSP exam.

                          THE LATE-PROJECT SPI CONVERGENCE TRAP

   SPI Value
      ▲
  1.20│
  1.00│───────Planned Performance (1.00)───────────────────────► CONVERGES TO 1.00!
      │                                                     ▲
  0.80│                       Actual SPI Drops               │
      │                             \                       │  False Recovery
  0.60│                              \                     /   (EVM Blind Spot)
      │                               \                   / 
  0.40│                                └─── Late Project ┘
      └───────────────────────────────────┬──────────────────┬──────────────► Time
                                       Target Finish      Actual Finish
                                          (PD)               (Late)

Mathematical Origin of the Quirk

1. By definition, the Planned Value curve terminates at project completion when $PV = BAC$. Once the planned duration ($PD$) elapses, Planned Value stops growing; it remains fixed at $PV = BAC$.
2. Even if a project is catastrophically delayed by months or years, physical work continues until all scope is finally completed. Therefore, Earned Value continues to rise toward $BAC$.
3. As physical completion nears 100%: $EV \to BAC$ $PV = BAC$
4. Look at what happens to the mathematical formulas: $\lim_{EV \to BAC} SV = EV - PV = BAC - BAC = \mathbf{\text{\textdollar}0}$ $\lim_{EV \to BAC} SPI = \frac{EV}{PV} = \frac{BAC}{BAC} = \mathbf{1.00}$

The Real-World Danger of SPI Convergence

Consider a $10 million hospital construction project planned for 12 months. At Month 12, the contractor has achieved only $7 million of work ($EV$ = $7M, $PV$ = $10M, $SPI = 0.70$). The project is severely delayed.

• During Months 13 through 18, the contractor finishes the remaining $3 million of work.
• At Month 15: $EV$ = $9M, $PV$ = $10M $\implies SPI = 0.90$, $SV$ = -$1M.
• At Month 18: $EV$ = $10M, $PV$ = $10M $\implies SPI = 1.00$, $SV$ = $0!
• The Absurd Reality: An executive reviewing the monthly EV dashboard sees $SPI$ rise from 0.70 to 0.90 to 1.00, giving the illusion of a heroic schedule recovery. In reality, the project finished 6 months late (50% schedule overrun)! Traditional $SPI$ and $SV$ completely lost their ability to signal schedule performance.

---

Dollar-Denominated Schedule Variance Confusion

Another fundamental defect of traditional EVM is that Schedule Variance is denominated in currency units (dollars, euros) or labor hours, NOT calendar time.

Why Dollars Cannot Measure Time

• If a project controls report states: $SV$ = -$250,000, what does this mean to the site superintendent?
  • Does it mean the project is 2 days late? 2 weeks late? 2 months late?
  • It is impossible to know without looking at the rate of expenditure (the slope of the PV curve) at that specific period.
• Furthermore, the magnitude of $SV$ is heavily skewed by the capital intensity of the work packages active during that window:
  • A delay on an activity involving the delivery of a $2,000,000 gas compressor generates an immediate negative $SV$ of -$2,000,000.
  • A delay on a $15,000 software integration logic activity on the critical path generates an $SV$ of only -$15,000.
  • Traditional EVM flags the $2,000,000 compressor as an existential schedule disaster, even if it has 90 days of float, while ignoring the $15,000 critical path activity that is actively delaying total project commissioning!

---

Critical Path vs. Non-Critical Path Earned Value Distortion

Traditional EVM treats every dollar of budget identically. It possesses zero network topology awareness. It cannot differentiate between a dollar earned on the zero-float critical path and a dollar earned on a low-priority task with 200 days of float.

                 CRITICAL PATH vs. NON-CRITICAL PATH EV DISTORTION

   Baseline Plan (Month 4): Target PV = \$500,000
   ┌───────────────────────────────────────────────┬───────────────────────────────┐
   │ Critical Path Foundation Work: \$250,000      │ Non-Critical Landscaping:     │
   │ (Zero Float - Drives Completion Date)         │ \$250,000 (120 Days Float)    │
   └───────────────────────────────────────────────┴───────────────────────────────┘

   Actual Field Execution at Month 4:
   ┌───────────────────────────────────────────────┬───────────────────────────────┐
   │ Critical Foundations: 0% Earned (EV = \$0)    │ Non-Critical Landscaping:     │
   │ STALLED DUE TO PERMIT DELAY                   │ Accelerated! 100% Earned      │
   │ Critical path delayed by 60 days!             │ EV = \$500,000 (Scope advance)│
   └───────────────────────────────────────────────┴───────────────────────────────┘

   Consolidated EVM Project Report:
   • Total Planned Value (PV):  \$250,000 + \$250,000 = \$500,000
   • Total Earned Value (EV):   \$0 + \$500,000       = \$500,000
   • Schedule Variance (SV):    EV - PV = \$500,000 - \$500,000 = \$0
   • Schedule Index (SPI):      EV / PV = \$500,000 / \$500,000 = 1.00

   STATUS REPORT: "Project is exactly on schedule (SPI = 1.00)!"
   FIELD REALITY: Project completion date has slipped 2 months on the critical path!

The "Cherry-Picking" / "Front-Loading" Vulnerability

Because contractors are often evaluated or paid based on $SPI$ and progress billing, unprincipled project managers can artificially inflate $SPI$ by deploying crews to execute non-critical, easy-to-earn work packages (such as bulk excavation or procurement deliveries with high float), while critical path engineering or permitting languishes.

The AACE Professional Mandate

earned-value management practice and AACE scheduling principles mandate that Earned Value must never be interpreted without cross-referencing the CPM schedule critical path and total float distributions:

1. Supplement aggregate SPI with CPM milestone and path analysis; a critical-path-only index may be useful if its scope and weighting are defined.
2. Never assume that $SPI \ge 1.00$ implies the project will finish on or ahead of the contract completion date.

---

Worked Calculation: The SPI Convergence & Distortion Demonstration

A $1,200,000 data center expansion project has a Planned Duration of 10 months ($PD = 10$).

• Monthly planned value is linear at $120,000 per month ($BAC$ = $1,200,000).
• By Month 8, execution difficulties occur. At Month 10 ($AT = 10$), the project has only earned $840,000 ($EV$ = $840,000).
• Work continues for 4 additional months until Month 14, when all scope is finished.

Table of Progression Across Months 10 through 14:

Observations for Exam Review:

1. Look at Months 11 through 14: Although the project is in deep contractual breach (operating past the contract completion date), $SV$ becomes less negative each month (from -$360k to -$90k to $0) and $SPI$ marches relentlessly upward toward $1.00$.
2. A project controls manager who relies solely on traditional EVM would report 'improving schedule performance' during the very period the project was suffering an uncontrolled schedule overrun.