TRC fatal exceptions, explained

Three severities, two outcomes

Every TRC exception lands in one of three buckets. Fatal (F) exceptions must be corrected. Full stop: an uncorrected fatal prevents the SACS Web System from completing an Official Export. Warning (W and WC) exceptions must be corrected or explained; an unexplained warning blocks the export just as hard as a fatal. Informational (O) exceptions ask for review and, in some cases, an explanation, but they are the only tier that will let you through untouched.

The operative rule in CDE's user manual is blunt: every check must carry a status of Passed or Explained before an official submission can occur. There is no third state. A fatal cannot be explained away, which is what separates it from a warning, and is also why fatal counts are the number to watch when a deadline is close.

The four check groups

TRCs run in four groups, and knowing which group produced an exception tells you where the fix lives.

• Import checks validate account strings and amounts as data enters: invalid codes, invalid combinations, malformed records. Fixes live in your ledger coding.
• General ledger checks test balancing: interfund ties, intrafund ties, contributions netting to zero, components of ending fund balance. Fixes are usually journal entries.
• Supplemental checks test form completion and consistency between forms and GL data.
• Export validation verifies form dependencies and requirements at the moment you try to leave.

The expensive ones are the first two groups, because their fixes happen outside the SACS Web System entirely. A coding correction means a journal in your financial system, a fresh extract, and a reimport before the check will read Passed.

Why early beats often

Each fatal you discover at export time costs a full cycle: post the correction in your ledger, re-extract, reimport, rerun the checks. Discover twelve fatals on the afternoon of the deadline and the cycles stack. The same twelve, surfaced against the raw ledger export a week earlier, are one batch of journals and one clean import.

That is the whole argument for checking the file before the system does: not different findings, earlier findings, with time left to act on them.