Running pattern detector against the corpus
The dataset shows circular money flows between charities — reciprocal pairs, triangular cycles, and longer chains. Most loops are structurally normal; the signal is the deviation from norm.
Glassbox surfaces correlations and observations, not findings of misconduct. Severity is a calibrated read on how strongly the dataset matches the pattern definition — a starting point for review, not a verdict.
The match meets the pattern threshold but the signal is the lower band. Normal review queue.
The match exceeds the routine threshold. Worth a closer look — pull the source rows, walk the citation.
The match is in the highest band of the calibrated scale. Recommend: prioritise for active review.
Severity is computed from the detection signal · app.gift_loops: simple cycles ≤ 4 hops in the ≥ $5K qualified-donee gift graph, hop fiscal periods non-decreasing and the whole loop within 12 months, hubs (degree ≥ 50 or DAF/foundation/federated/diocese names) never interior; edge weight = gift ÷ donor's gifts that year; score = min edge weight × hops penalty (2:1.0, 3:0.8, 4:0.6); aggregated per BN, best loop ≥ 0.05 · fallback: cra.loop_universe score ≥ 12
The dataset shows THE CANADIAN CHINESE ALLIANCE CHURCHES ASSOCIATION (BN 887658599RR0001) participating in 2 time-ordered circular gift flows totalling $35,100; strongest loop (THE CANADIAN CHINESE ALLIANCE CHURCHES ASSOCIATION → WESTWOOD ALLIANCE CHURCH) is 2 hops with score 0.10 — at least 10% of each donor's giving that year travelled around the loop; 2 of the loops originate at a hub (DAF / community foundation / federated body).
Why this may be benign · Donor-advised-fund platforms and gift-fund sponsors (e.g. Canada Gives, Aqueduct, Charitable Impact, Benefaction, Strategic Charitable Giving, Private Giving Foundation) are pass-through intermediaries; loops and high degree through them reflect donor instructions.
Once a match is identified, the dataset suggests these review actions. Glassbox does not prescribe outcomes — these are calibrated options for funders, auditors, and program officers, framed as correlations to investigate.
For each entity in the cycle, examine reciprocal gifts, board overlap, and program-vs-overhead ratios. The cycle may reflect denominational hierarchy or federated structure (normal) or reciprocal-revenue inflation (review-worthy).
Glassbox is not yet hard-coded with denominational hierarchy or federated-charity exclusion lists. Before flagging externally, confirm the loop is not an explainable structural pattern.
If the same entities also receive federal grants for similar program purposes, the loop becomes a duplicative-funding signal as well — see /follow/duplicative-funding once that detector ships.
If the loop persists across review cycles, escalate to the relevant audit body. If it resolves (entity exits, program ends, structure clarified), close with reasoning recorded.
These are calibrated suggestions, not directives. The dataset shows the pattern; the funder decides the response.
Aggregated from the matches below. Severity bands are calibrated per pattern; counts and dollar totals reflect this snapshot only.
The dataset shows WESTWOOD ALLIANCE CHURCH (BN 894238757RR0001) participating in 1 time-ordered circular gift flow totalling $21,000; strongest loop (THE CANADIAN CHINESE ALLIANCE CHURCHES ASSOCIATION → WESTWOOD ALLIANCE CHURCH) is 2 hops with score 0.10 — at least 10% of each donor's giving that year travelled around the loop; 1 of the loops originate at a hub (DAF / community foundation / federated body).
Why this may be benign · Donor-advised-fund platforms and gift-fund sponsors (e.g. Canada Gives, Aqueduct, Charitable Impact, Benefaction, Strategic Charitable Giving, Private Giving Foundation) are pass-through intermediaries; loops and high degree through them reflect donor instructions.
The dataset shows THE ROMAN CATHOLIC EPISCOPAL CORPORATION OF PRINCE RUPERT (BN 119122786RR0023) participating in 9 time-ordered circular gift flows totalling $346,439; strongest loop (THE ROMAN CATHOLIC EPISCOPAL CORPORATION OF PRINCE RUPERT → THE ROMAN CATHOLIC EPISCOPAL CORPORATION OF PRINCE RUPERT) is 2 hops with score 0.10 — at least 10% of each donor's giving that year travelled around the loop.
Why this may be benign · Denominational hierarchies (dioceses, synods, conferences, councils, presbyteries) routinely move funds between the parent body and congregations; structural relatedness here is expected.
The dataset shows BOW ISLAND CONGREGATION, CHURCH OF GOD IN CHRIST, MENNONITE (BN 870063195RR0001) participating in 3 time-ordered circular gift flows totalling $301,418; strongest loop (CENTRAL CHARITIES BOARD,CHURCH OF GO IN CHRIST, MENNONITE CANADA → BOW ISLAND CONGREGATION, CHURCH OF GOD IN CHRIST, MENNONITE) is 2 hops with score 0.10 — at least 10% of each donor's giving that year travelled around the loop; 3 of the loops originate at a hub (DAF / community foundation / federated body).
Why this may be benign · Denominational hierarchies (dioceses, synods, conferences, councils, presbyteries) routinely move funds between the parent body and congregations; structural relatedness here is expected.
The dataset shows FOURSQUARE GOSPEL CHURCH OF CANADA (BN 107396905RR0047) participating in 1 time-ordered circular gift flow totalling $45,911; strongest loop (FOURSQUARE GOSPEL CHURCH OF CANADA → FOURSQUARE GOSPEL CHURCH OF CANADA) is 2 hops with score 0.10 — at least 10% of each donor's giving that year travelled around the loop; 1 of the loops originate at a hub (DAF / community foundation / federated body).
Why this may be benign · Donor-advised-fund platforms and gift-fund sponsors (e.g. Canada Gives, Aqueduct, Charitable Impact, Benefaction, Strategic Charitable Giving, Private Giving Foundation) are pass-through intermediaries; loops and high degree through them reflect donor instructions.
The dataset shows VICTORY LIFE CHURCH ST. ALBERT (BN 859881179RR0001) participating in 2 time-ordered circular gift flows totalling $98,839; strongest loop (VICTORY CHURCHES OF CANADA INTERNATIONAL ASSOCIATION → VICTORY LIFE CHURCH ST. ALBERT) is 2 hops with score 0.10 — at least 10% of each donor's giving that year travelled around the loop; 2 of the loops originate at a hub (DAF / community foundation / federated body).
Why this may be benign · Donor-advised-fund platforms and gift-fund sponsors (e.g. Canada Gives, Aqueduct, Charitable Impact, Benefaction, Strategic Charitable Giving, Private Giving Foundation) are pass-through intermediaries; loops and high degree through them reflect donor instructions.
The dataset shows HILLSIDE COMMUNITY CHURCH (BN 141091652RR0001) participating in 2 time-ordered circular gift flows totalling $125,224; strongest loop (EVANGELICAL MISSIONARY CHURCH OF CANADA → HILLSIDE COMMUNITY CHURCH) is 2 hops with score 0.10 — at least 10% of each donor's giving that year travelled around the loop; 2 of the loops originate at a hub (DAF / community foundation / federated body).
Why this may be benign · Donor-advised-fund platforms and gift-fund sponsors (e.g. Canada Gives, Aqueduct, Charitable Impact, Benefaction, Strategic Charitable Giving, Private Giving Foundation) are pass-through intermediaries; loops and high degree through them reflect donor instructions.
The dataset shows ALZHEIMER SOCIETY OF CANADA - SOCIETE ALZHEIMER DU CANADA (BN 118784925RR0001) participating in 32 time-ordered circular gift flows totalling $18,843,981; strongest loop (ALZHEIMER SOCIETY OF CANADA - SOCIETE ALZHEIMER DU CANADA → ALZHEIMER SOCIETY OF ALBERTA AND NORTHWEST TERRITORIES) is 2 hops with score 0.10 — at least 10% of each donor's giving that year travelled around the loop; 32 of the loops originate at a hub (DAF / community foundation / federated body).
Why this may be benign · Donor-advised-fund platforms and gift-fund sponsors (e.g. Canada Gives, Aqueduct, Charitable Impact, Benefaction, Strategic Charitable Giving, Private Giving Foundation) are pass-through intermediaries; loops and high degree through them reflect donor instructions.
The dataset shows ALZHEIMER SOCIETY OF ALBERTA AND NORTHWEST TERRITORIES (BN 129690343RR0001) participating in 3 time-ordered circular gift flows totalling $5,810,762; strongest loop (ALZHEIMER SOCIETY OF CANADA - SOCIETE ALZHEIMER DU CANADA → ALZHEIMER SOCIETY OF ALBERTA AND NORTHWEST TERRITORIES) is 2 hops with score 0.10 — at least 10% of each donor's giving that year travelled around the loop; 3 of the loops originate at a hub (DAF / community foundation / federated body).
Why this may be benign · Donor-advised-fund platforms and gift-fund sponsors (e.g. Canada Gives, Aqueduct, Charitable Impact, Benefaction, Strategic Charitable Giving, Private Giving Foundation) are pass-through intermediaries; loops and high degree through them reflect donor instructions.
The dataset shows SHABBAT SHEL SHALOM MESSIANIC FELLOWSHIP (BN 813906971RR0001) participating in 2 time-ordered circular gift flows totalling $25,760; strongest loop (BRITISH COLUMBIA BAPTIST CONFERENCE → SHABBAT SHEL SHALOM MESSIANIC FELLOWSHIP) is 2 hops with score 0.10 — at least 10% of each donor's giving that year travelled around the loop; 2 of the loops originate at a hub (DAF / community foundation / federated body).
Why this may be benign · Donor-advised-fund platforms and gift-fund sponsors (e.g. Canada Gives, Aqueduct, Charitable Impact, Benefaction, Strategic Charitable Giving, Private Giving Foundation) are pass-through intermediaries; loops and high degree through them reflect donor instructions.
The dataset shows BOW VALLEY BAPTIST CHURCH (BN 121687727RR0001) participating in 9 time-ordered circular gift flows totalling $1,675,310; strongest loop (CANADIAN NATIONAL BAPTIST CONVENTION → MIDWEST BAPTIST ASSOC → BOW VALLEY BAPTIST CHURCH) is 3 hops with score 0.10 — at least 10% of each donor's giving that year travelled around the loop; 9 of the loops originate at a hub (DAF / community foundation / federated body).
Why this may be benign · Denominational hierarchies (dioceses, synods, conferences, councils, presbyteries) routinely move funds between the parent body and congregations; structural relatedness here is expected.