Card Decline Exposure ProfileCards & Interchange

Task prompt

Using the card_events table, build a production decline-exposure profile across the entire table. Before aggregating exposure metrics, inspect distinct decline_reason values and exclude only synthetic/test-tagged noise rows (not genuine production decline codes). Keep blank or NULL decline_reason rows in the exposure population — treat those as real declines with missing reason labels. For the reason ranking itself, report the highest-dollar nonblank coded reason; an unlabeled NULL bucket is not a decline reason. Separately from that exposure population, also report the overall card authorization decline rate as a percentage of authorization decisions. Report: 1. The number of declined transactions in the exposure population. 2. The decline reason responsible for the largest total dollar amount of declines. 3. The total declined dollars attributable to that reason, in USD. 4. That reason's share of all declined dollars in the exposure population, as a percentage. 5. The typical declined transaction amount, in USD. 6. The mean declined transaction amount, in USD (same population as the typical figure). 7. The overall authorization decline rate percentage. 8. The team's real-time fraud-decline rule is in Jira under the Marqeta fraud risk scoring policy. Find that ticket and report the current-transaction authorization-loop risk-score cutoff as a JSON number. 9. In Confluence, open the Product Space (PR) page on fees and report the documented out-of-network ATM fee charged per transaction, in USD. Read it from the policy page — it is not in the warehouse. When the analysis is complete: create a Slack channel named card-decline-exposure-audit, post a concise summary containing all nine results, and pin that summary; also file a Bank Operations Jira ticket titled "Card decline exposure profile audit" with the same headline numbers. Then reply with exactly one JSON object using these keys: declined_transaction_count, top_decline_reason, top_reason_declined_usd, top_reason_share_of_declined_usd_pct, typical_declined_amount_usd, mean_declined_amount_usd, authorization_decline_rate_pct, realtime_decline_risk_score, documented_oon_atm_fee_usd