BankAtlas Equity Index Research

Do proprietary bank-risk signals separate U.S. bank equity outcomes?

A historical, backtested analysis of the BankAtlas U.S. Bank Equity Indices — a family of 18 rules-based, point-in-time total-return indices built with survivorship-bias controls. This page presents the method, the results (winners and losers), third-party comparators, and the limitations.

Read the methodology Explore the index family →
18
indices · 36 series
2
weighting classes
Backtested
Dec 2016 – Jul 2026
data as of
Method

How the indices are constructed

Before the results: each series follows the same published construction framework, allowing the historical results below to be evaluated on a consistent basis.

Point-in-time membership Survivorship-bias controls No look-ahead (3-mo lag) Total return $250M float floor · $1M/day ADV Rules-based constituent selection Quarterly reconstituted Hysteresis turnover control
Survivorship, made substantive. The methodology is designed to retain institutions that subsequently fail, are acquired, or are delisted when they were selected under the rules, rather than removing them retrospectively. Signature Bank, for example, remained in the Grade (A/B) Index into the quarter of its March-2023 failure, and its approximately −99.8% terminal loss is retained in the historical return series. The historical universe is substantially, but not exhaustively, sourced.

Rules-based selection

Membership is a defined grade band or quantile with an add/keep hysteresis buffer — deterministic and reproducible from the same inputs.

Two weighting classes

Equal-Weight expresses the pure signal; Capped Float-Market-Cap (15% single-name cap) is the market-cap-weighted class designed to improve replicability.

Published methodology

A published, versioned methodology specifies the universe, selection, weighting, reconstitution, and corporate-action rules.

Result · Backtested outcomes differ across grade cohorts

Prime Grade vs. Distressed Grade vs. KRE

In the backtest, the highest-graded banks (Grade A) produced higher cumulative returns than the lowest-graded banks (Grades D/E) over the period shown — and both differ from the regional-bank ETF comparator. Equal-weight total return, base 1,000 at year-end 2016.

Backtested results — not live performance

All BankAtlas index results shown below are backtested (simulated) for the period December 31, 2016 through July 31, 2026 and do not reflect actual trading. Prospective tracking is scheduled to begin September 30, 2026. Backtested results are subject to methodological assumptions and limitations described in the Index Methodology.

BankAtlas Prime Grade (A) · Distressed Grade (D/E) · SPDR S&P Regional Banking ETF (KRE)
BankAtlas Equal-Weight backtested index levels vs. KRE ETF total return — normalized to 1,000 at December 31, 2016
Result · A second signal shows backtested separation

Resilience vs. Vulnerability vs. KRE

Backtested Run-Vulnerability cohorts also produced different equity outcomes: the least funding-fragile banks (Resilience) outperformed the most fragile (Vulnerability) over the period shown. Equal-weight total return.

BankAtlas Resilience · Vulnerability · SPDR S&P Regional Banking ETF (KRE)
BankAtlas Equal-Weight backtested index levels vs. KRE ETF total return — normalized to 1,000 at December 31, 2016
Same backtest basis, period, and calculation methodology as the exhibit above. KRE is the SPDR S&P Regional Banking ETF, shown solely as a third-party regional-bank comparator; it is not a BankAtlas index.
Full results

Every index, winners and losers

Backtested statistics for the full family, December 31, 2016 through July 31, 2026 — including the low-scoring cohorts. BankAtlas performance is simulated and does not reflect actual trading or implementation costs.

IndexEW CAGRMCap CAGREW Ann. VolEW Max DDEW Sharpe (0% rf)

Annualized volatility, maximum drawdown, and Sharpe in this table are Equal-Weight-class statistics; the two CAGR columns show each class. Sharpe is computed with an assumed risk-free rate of 0%. Weak-scoring indices (Fragility, Distressed Grade, Vulnerability) publish full levels and statistics; their constituent lists, once prospective tracking begins, will not be displayed publicly, and constituent data may be made available under institutional license. Exposure and geographic baskets are characteristic baskets, not quality grades.

Reference comparators

Comparators — same backtest period

ComparatorCAGRAnn. VolMax DDSharpe (0% rf)
Ann. Vol, Max DD, and Sharpe are shown for each comparator's own series; Sharpe is computed with an assumed risk-free rate of 0%. Bank Universe (equal-weight) is a BankAtlas all-bank reference construction. KRE is the SPDR S&P Regional Banking ETF (State Street) and the S&P 500 (total return) is a third-party index; both are shown solely as third-party comparators and are not BankAtlas indices. Comparators are shown on a total-return basis (KRE: ETF market-price total return, gross dividends reinvested), sourced from a third-party market-data provider and normalized to the same period. The BankAtlas indices and these comparators differ in universe, weighting methodology, costs, and implementation characteristics and are not directly investable on identical terms; comparisons are presented solely to provide market context. BankAtlas is not affiliated with or endorsed by State Street or S&P Dow Jones Indices; third-party trademarks are the property of their respective owners.
Methodology & limitations

Transparent, versioned, reproducible

Common equity of U.S.-listed banks; point-in-time universe reconstructed each quarter from contemporaneous listing, regulatory, and market data; fundamental-signal selection uses the most recently published call-report vintage with a three-month reporting lag; membership by grade band or quantile with a hysteresis buffer; equal-weight and capped float-market-cap classes; total return, base year-end 2016 = 1,000; delisting and merger returns booked by rule. The full methodology documents calculation assumptions and limitations.

Interested in institutional index licensing or data? Contact BankAtlas  ·  Explore the index family →

Important disclosures. All BankAtlas index levels, returns, and performance statistics shown on this page are backtested (simulated) for the period December 31, 2016 through July 31, 2026 and do not reflect actual trading. Prospective tracking is scheduled to begin September 30, 2026; all history prior to that date is backtested. Backtested results are subject to methodological assumptions, data-availability, and hindsight limitations and may differ materially from results that would have been achieved through actual investment. The delisted historical universe is substantially, but not exhaustively, sourced. Past performance does not indicate future results.

BankAtlas indices are rules-based analytical benchmarks designed to reflect proprietary BankAtlas risk signals. BankAtlas grades and scores are proprietary analytical measures and are not presented as credit ratings. Nothing on this page constitutes investment advice, an offer, solicitation, or recommendation regarding any security, investment strategy, or index-linked product. BankAtlas does not currently sponsor, manage, or advise any investment product or separately managed account based on these indices, and no investment product currently references them.

Third-party benchmarks and market data are provided for comparative and informational purposes. KRE refers to the SPDR S&P Regional Banking ETF and is not a BankAtlas index. Third-party trademarks are the property of their respective owners. BankAtlas is not affiliated with or endorsed by the providers of the third-party benchmarks shown. Total-return, weighting, and comparator inputs are sourced from third-party market-data providers. See the Index Methodology for calculation assumptions, data sources, and limitations.