Benchmarking data against the market: absorption, cap rate, and rent indices

Limitation: NPI reflects only unlevered, institutionally held stabilized assets. It excludes most private, leveraged, or opportunistic deals, and skews toward large core markets.

CoStar and rent/vacancy indices

CoStar is the dominant US commercial real estate data provider, tracking leasing comps, asking and effective rents, vacancy, and under-construction pipelinepipelineAll active sales opportunities across the stages of the sales process, together with their combined potential value and probability of closing.Voir la définition complète → down to the submarket level. It is the most-cited source for "market rent" in US broker and appraisal work. Europe has fragmented equivalents: MSCI Real Capital Analytics for transaction data, PMA and CBRE/JLL/Savills research desks for rent and vacancy series by city.

Absorption (a data metric, not a financial ratio) measures the net change in occupied square footage over a period. Positive net absorption means more space got leased than vacated. Trackers from CoStar, JLL, and Cushman & Wakefield publish this quarterly by market and property type, it is the primary leading indicator of supply-demand balance before rents move.

Cap rate surveys

The capitalization rate (net operating income divided by property value) is a pricing metric, but the data question here is sourcing: where does the "market cap rate" you compare against actually come from? Key survey sources:

• RCA (Real Capital Analytics/MSCI): transaction-based cap rates from actual closed sales
• PwC/ULI Real Estate Investor Survey: opinion-based, from investor and appraiser surveys (published as "Emerging Trends in Real Estate")
• CBRE Cap Rate Survey: broker-reported, twice yearly, by market and asset class

These three methods (transacted, surveyed, broker-estimated) can diverge meaningfully. As of early 2025 estimates, US CBD office cap rates were widely cited in the 7 to 9% range while stabilized multifamily sat closer to 5 to 5.5%, but treat any single figure as directional, not a hard fact, since methodology drives the number as much as the market does.

A worked comparison

Say your industrial property's internal rent roll shows average net rent of $8.50/sq ft. CoStar's submarket index shows average asking rent of $7.20/sq ft for comparable Class A industrial.

Gap = ($8.50 − $7.20) / $7.20 = 18% premium to market.

Next step is not to celebrate, it's to investigate the gap:

• Check lease start dates: is this an old, above-market lease about to roll down?
• Check comp set: is CoStar including older Class B buildings that drag the average down?
• Check units: net vs. gross rent mismatches are the single most common benchmarking error.

If the premium survives these checks, you likely have a genuine outperforming asset. If not, you've caught a data qualitydata qualityThe degree to which data is fit for purpose: accurate, complete, consistent, timely, valid and unique. Poor quality data undermines analytics, reporting and AI.Voir la définition complète → problem before it fed into a valuation or investment memo.

Data qualityData qualityThe degree to which data is fit for purpose: accurate, complete, consistent, timely, valid and unique. Poor quality data undermines analytics, reporting and AI.Voir la définition complète → checks before you benchmark

Benchmarking is only as good as the inputs on both sides. Run these checks first:

1. Definitional alignment: does "vacancy" mean physical vacancy or economic vacancy (including free rent periods)? NCREIF, CoStar, and your property manager may define it differently.

2. Timing lag: NPI is quarterly and reported with a lag; CoStar updates more continuously but revises historical periods as new leases are recorded. Compare same-vintage data, not your Q4 actuals against a stale Q2 index.

3. Sample size: a submarket index built on 5 comparable buildings is noisier than one built on 50. CoStar and RCA both let you check underlying sample counts, always do.

4. Currency and geography consistency (Europe): a Paris CBD office yield is not comparable to a secondary French regional city; benchmarks must match geographic granularity.

Simple governance metric: benchmark variance threshold

A practical internal control many asset managers use:

This kind of variance threshold turns benchmarking from a one-off analysis into a repeatable data governancedata governanceData governance is the set of policies, roles, and processes that ensure data is accurate, secure, well-defined, and used responsibly across an organization.Voir la définition complète → control, catching both data errors and genuinely newsworthy outperformance.

Choosing the right benchmark for the right question

| Question | Best benchmark source |

|---|---|

| Is my building's total return in line with institutional peers? | NCREIF NPI |

| Is my asking rent competitive? | CoStar / JLL / CBRE rent index |

| Is demand in this submarket strengthening? | Net absorption trackers (CoStar, C&W) |

| Is my exit cap rate assumption realistic? | RCA transaction data (most reliable) over broker surveys |

| Is my European asset priced in line with peers? | MSCI RCA Europe, national statistical offices, PMA |

Matching the question to the source matters more than defaulting to whichever dataset your firm already has a subscription to.

🎬 [VIDEO: "Understanding Cap Rates and NOI in Commercial Real Estate" - youtube.com/results?search_query=understanding+cap+rates+noi+commercial+real+estate - a practical walkthrough of how cap rate and NOI data are actually calculated and sourced, useful before comparing against survey benchmarks]

Key Takeaways

• Never evaluate a property's occupancy, rent, or yield in isolation. Anchor every internal metric to an external market benchmark (NCREIF, CoStar, RCA) before drawing conclusions.
• Know what each benchmark actually measures: NPI is unlevered institutional total return, CoStar/JLL indices are rent and vacancy by submarket, RCA cap rates are transaction-based while CBRE/PwC surveys are opinion-based. Methodology drives the number.
• A benchmark gap is a starting point for investigation, not a conclusion. Check definitions (physical vs. economic vacancy), timing lag, and sample size before trusting a variance.
• Build a simple variance-threshold rule (e.g., flag any metric more than 15% off benchmark) into your reporting process as a repeatable data governancedata governanceData governance is the set of policies, roles, and processes that ensure data is accurate, secure, well-defined, and used responsibly across an organization.Voir la définition complète → control.
• In Europe, match benchmark geography and currency carefully. National markets and even city submarkets can diverge sharply, so a single "European cap rate" is rarely meaningful.