ARC-CECL models your allowance from every loan you hold, reconciled to your general ledger every month. When it moves, you see the portfolio, the pool, the loans, and the assumption that moved it. Implementation, reporting, the data warehouse, and your annual review are included.
Built on the same reconciled loan record as the rest of ARCSys.
Illustrative example · sample institution
The CECL number is not the hard part. Explaining why it changed is.
When the board, the auditor, or the examiner asks why the number moved, the answer is already traced to the loans behind it.
Loss history comes from your own loans first, loaded as far back as your data goes. Where a pool’s history is thin, outside data can fill the gap, and you’ll see exactly where it entered.
We measure each loan’s prepayment every month, so the life of every loan in your estimate reflects how your borrowers actually pay.
With third-party data, the forecast is built once, and every institution gets fitted to it. With your loan-level data, we forecast every single one of your pools.
Segmentation by delinquency, risk rating, and credit score, so each pool groups loans that behave alike, not Call Report categories.
More than one methodology and economic scenario run on your loans, backtested against actuals, with management’s choice documented.
Drill from the total to the portfolio, the pool, and the loan, and see which assumption moved it.
Loading, field mapping, and GL tie-out are part of the relationship, not a data project for your team. Board, auditor, and examiner packets and GAAP disclosures are produced from the same loans as the allowance.
Losses fell for three straight quarters. The allowance rose anyway. Here’s why, loan by loan.
▸ See it on your loansAs rates rose, borrowers prepaid more slowly, so loans stay on the books longer and expected lifetime losses rise.
Three loans were downgraded after their annual reviews.
The unemployment path was revised upward.
The reports your committee, your auditor, and your examiner ask for, produced from the same loan-level data as the allowance itself.
The quarter in one page your board can read.
What moved, by pool, with the loans behind it.
Methods side by side, with management’s documented choice.
Complete, automated, with prior-period model inputs.
Per-loan speeds tied back to pool forecasts.
Coverage beyond the loan book.
Your own reports for ALCO, the board, or the exam file.
The work other approaches leave with your team, or put on a separate invoice, is part of the ARC-CECL relationship.
Your allowance is modeled from your institution’s own loss history. Where a pool’s history is thin, outside data can fill the gap, and you’ll see exactly where it entered. It’s never quietly blended in.
We load, format, and analyze your historical loan data, often many years of it, reconcile it to your GL, and recommend pools and models based on what we find.
Segmentation reflects delinquency, risk rating, and credit score, not Call Report categories that treat all CRE alike.
Every method you run sits side by side, so you can backtest, compare, and change methods with documentation behind the decision. You’re never locked into one answer.
Drill from the total to a pool to a single loan, and tie it back to forecasts and reversion rates. No black box when an examiner asks.
GAAP disclosures, board reporting, the ARCAnalytics data warehouse, your annual ARC-CECLReview, and accounting support, all inside the fee.
Implementation is done with you, not handed to you. Once the history is loaded and tied out, every quarter-end follows the same documented path.
What you hold, how your teams use the data, and how much history is available.
Core files from every legacy system are consolidated, with history reaching back as far as your data goes.
Fields are mapped, reconciled, and matched to the general ledger before any model runs.
Pools and methodologies are chosen for your portfolio and documented as they are built.
The model is validated, the methodology is written down, and your team runs it with ARCSys beside you.
The quarter's loan file loads, maps, and ties out to the GL.
Loans are pooled, prepayment runs per loan, and the selected methodology is refreshed.
Every movement in the allowance is traced to its cause, down to a single loan.
The packet is assembled with the explanation attached, ready for the board and the examiner.
Some tools look simpler because they show less. ARC-CECL keeps every input open to inspection, so when scrutiny arrives, you answer from your own loans instead of rebuilding the answer in a spreadsheet.
▸ See it on your loansSpend less time assembling the answer and gain a clearer path to the data, drivers, reporting, and support needed to explain it.
Last quarter’s hardest question, answered at the loan level.
Operational burden, reporting, documentation, governance readiness, and specialist support, weighed together rather than as a software license alone.
Every assumption is traceable and every output is ready for review, with the methodology written down where a reviewer can find it.
A 30-minute walkthrough on the allowance question that’s hardest to answer today, with a specialist who works inside the same loan-level data you do.
