Data pipeline and deadlines

How data reaches both origins

1. The administrator, custodian or pricing vendor drops the daily file (holdings, NAV, market price, bid-ask, MMF liquidity and flows).
2. The publish job validates it: as-of date is the prior business day, weights sum to about 100%, no blank CUSIPs, premium/discount within a plausible range. A failed check stops publication and alerts; it never posts stale data under today's date.
3. The job renders static HTML, CSV/JSON and chart images, and writes them to both origins.
4. The job reads each origin back directly and records the result in the evidence log (GEN-12).
5. Cloudflare serves whichever origin is healthy; the cache is purged on publish so neither origin shows yesterday's data.

Deadlines

Item	Data as of	Live on both origins by (internal target)	Legal deadline
ETF holdings, NAV, price, premium/discount, 30-day spread	Prior business day close	8:30 a.m. ET	Before market open (9:30 a.m. ET)
MMF liquid assets, flows, NAV	Prior business day	9:00 a.m. ET	Each business day (see Q-05)
ETF 2% statement	Day 8 of the run	Morning after the 8th trading day	Immediately after the 8th trading day
MMF N-CR Part C posting	Event	Same hour as the filing	Same business day as the filing
Prospectus, SAI, supplements, shareholder reports	Effective date	Same time as primary	On or before use
ETF premium/discount table and graph	Quarter-end	5 business days after quarter-end	Current through the latest quarter
MMF monthly schedule of investments	Month-end	3rd business day	5th business day
N-CSR Items 7–11; Q1/Q3 holdings	Period end	50 days after period end	60 days after period end

Pipeline checks

OPS-01 Required

The publish job validates every daily file before release, and a failed validation stops publication and alerts named owners at Pointpath and the client.

OPS-02 Required

Each item in the deadline table has a named owner at the client and at Pointpath, with an escalation path when an internal target is missed.