Methodology

Scope: 81 markets across cert set 35583, 873 (Pinnacle Financial Partners).

1. Unit of analysis

One row per footprint CBSA (Metropolitan or Micropolitan Statistical Area), and one row per footprint county that belongs to no CBSA (market_type = 'Rural county'). The market key is the CBSA code, or the 5-digit county FIPS for rural markets. * Metropolitan Statistical Area: 65 * Micropolitan Statistical Area: 12 * Rural county: 4

2. Size tier

Major Metro: MSA with population >= 1,000,000. Secondary Metro: 250,000-999,999. Small Metro: below 250,000. Micropolitan: any micropolitan CBSA. Rural: non-CBSA county.

3. Growth indicators

1. pop_cagr — 2020->2025 population CAGR (Census PEP)
2. netmig_rate — cumulative net domestic migration 2021-2025 / population 2020
3. gdp_cagr — 2019->2024 real GDP CAGR (BEA CAGDP9)
4. emp_cagr — 2019->2024 QCEW annual-average employment CAGR
5. pcpi_cagr — 2019->2024 per-capita personal income CAGR (BEA CAINC1)

4. Standardization

CBSA markets are z-scored against all US metro+micro CBSAs; rural county markets against all US counties. Moments are computed AFTER winsorizing the national distribution at the 1st/99th percentiles (ddof=0); the market's own UNWINSORIZED value is then scored against those robust moments.

emp_cagr has no national CBSA distribution in v1.0.0: its z-score is (area_emp_cagr - US_emp_cagr) / sd(footprint emp_cagr). This is a WEAKER benchmark than the other four.

5. Composite and growth tier

composite_z = 0.225*z_pop_cagr + 0.225*z_netmig_rate + 0.225*z_gdp_cagr + 0.225*z_pcpi_cagr + 0.100*z_emp_cagr

Weights are renormalized over whichever indicators are present; any renormalization is recorded in data_flags. High: composite_z >= +0.33. Low: composite_z <= -0.33. Medium: otherwise. * High: 47 * Medium: 24 * Low: 10

6. Deposit-growth context — the symmetric endpoint guard

A branch reporting deposits in 2020 and a literal $0 in 2025 has had its book moved to another office of record. Dropping it from the base is correct only if the receiving office is dropped from the current numerator too. Where an office in the SAME market absorbed at least the whole vanished book, the two sides are PAIRED and compared as one continuing unit; where none did, the survivor-only reading is stated on the row.

7. Franchise stage — ordered gates, first match wins

1. Limited-Service Foothold — Zero FULL-SERVICE branches today: presence is a loan-production / limited-service office only, so the bank files no Summary of Deposits record and holds no measurable deposit share. Checked first because every deposit-based rule below is undefined here.
2. New Entrant — entry_year >= 2021 (first pro-forma SOD branch in the last 5 surveys) OR (tenure < 8 yrs AND branches_current <= 2 AND deposit_share_current < 2%). Young and sub-scale. A market with at least one full-service branch today but NO SOD history at all is also a New Entrant: the office was established after the survey date and has not yet filed a Summary of Deposits.
3. Entrenched Leader (consolidating) — BOTH tests fire: deposit_share_current >= 8% AND the Pulling-Back branch/share test (rule 4). Evaluated BEFORE plain Pulling Back and plain Entrenched Leader, so a dominant franchise thinning overlapping branches is never mislabelled as a retreat.
4. Pulling Back — net_branch_change_5yr <= -2 AND share_delta_5yr_bps < 0; OR branch_change_pct <= -25% with a measured non-positive share delta (share_delta_5yr_bps <= 0). Physical and share retreat together.
5. Entrenched Leader — deposit_share_current >= 8%. A market where the bank is a structural incumbent.
6. Expanding — net_branch_change_5yr >= +2 OR share_delta_5yr_bps >= +50 OR (tenure < 12 yrs AND share_delta_5yr_bps > 0 AND branches grew).
7. Established - Mature — Everything else: long presence, share and branch count broadly stable, below the entrenchment cutoff.

• Entrenched Leader: 29
• Established - Mature: 17
• Expanding: 10
• Entrenched Leader (consolidating): 10
• New Entrant: 8
• Limited-Service Foothold: 4
• Pulling Back: 3

8. Share construction

Numerator (the bank) and denominator (all institutions) are aggregated over the SAME county set in the same year. Counties where the bank holds only a limited-service office are excluded from BOTH sides — an LPO files no Summary of Deposits, so it contributes a structurally zero numerator against a full denominator. Branch COUNTS use the wider CBSA county map, because a branch count is a bank-side census rather than a share. branch_effectiveness_index = deposit share / branch share; it is null, never zero, where branch share is zero.

### 8.6 Office-level book relocation An office whose absolute move between the 2020 and 2025 surveys clears $1.0bn, or 25% of its market's own base book (floor $25m), AND is anomalous in its own terms (source retained <= 20% of its base book, or recipient ended at >= 5x it) is reported as a MATERIAL BOOK MOVEMENT. It changes no label; it is a diagnostic.

### 8.6b Main-office zero reports * Pinnacle Bank (uninum 82012, cert 35583) reported $0 in 2025. Cert deposits in Davidson county went 10,452,141k -> 15,440,535k: the book was reallocated within the same county, so the market-level share is unaffected.

9. Data flags observed in this run

• entered_via_recent_merger: 45 market(s)
• limited_service_offices: 24 market(s)
• mixed_legacy_and_merger_branches: 13 market(s)
• book_relocation_in: 6 market(s)
• no_sod_history: 5 market(s)
• book_relocation_out: 4 market(s)
• no_full_service_office: 4 market(s)
• branches_2020_includes_cbsa_county_since_exited: 3 market(s)
• deposit_book_drop_-36pct: 2 market(s)
• nashville_main_office_uninum82012_reported_$0_in_june2025_sod;cert35583_davidson_county_deposits_2020=10,452,141k->2025=15,440,535k(book_retained_in_county_share_unaffected): 1 market(s)
• deposit_book_drop_-29pct: 1 market(s)
• deposit_book_drop_-26pct: 1 market(s)
• deposit_book_drop_-37pct: 1 market(s)
• deposit_book_drop_-22pct: 1 market(s)
• deposit_book_drop_-52pct: 1 market(s)
• deposit_book_drop_-40pct: 1 market(s)
• deposit_book_drop_-44pct: 1 market(s)