Run every investor in parallel — and defend every basis point.
Correspondent operations live and die by speed of price discovery and quality of documentation. RateStack runs the full ladder against every investor in milliseconds, surfaces BEST_EX, and ships a per-rule trace your compliance team can read out loud.
Correspondent shops feel two pressures at once: speed (the LO needs the price now) and defensibility (compliance needs to know how the price was built). Most pricing engines give you one or the other. RateStack gives you both because the rule engine emits the trace as a side-effect of doing the math.
Two-stage eligibility means you stop running the full ladder against programs the borrower can't take. When a 1% LTV move would knock out two investors, the LO sees it before they pick up the phone — not after the lock comes back repriced.
The audit story is uncomplicated: every quote, every adjustment, every rate-lock state change writes to an append-only log linked by SHA-256 hash chain. When the regulator asks 'why did this loan price this way on this day,' you replay the quote against the as-of ratesheet and ship the trace. Done.
Before vs. after
The shape of a day.
The same operating model, rebuilt around explicit pricing and a single audit log.
Before
LOs pull individual investor portals to run scenarios. Three quotes per borrower means six portal logins.
After
One pricing call returns BEST_EX across every active investor. The trace explains the order of finish.
Before
Compliance asks 'how was this price built' and gets a one-page Excel with no links to source rules.
After
Every quote ships with a per-rule trace including ratesheet version, rule id, condition matched, and combine strategy.
Before
Lock-day surprises when an LO didn't realize the borrower's DTI knocked out two of the three investors.
After
Stage 1 eligibility runs as the LO types. The ineligible panel explains the why for every rejected program.
Before
Ratesheet ingestion is a manual 7am chore that breaks when a vendor changes a column header.
After
Email-in + portal automation + learning header-mapping templates. Vendor reorders columns; pipeline keeps running.
Capabilities, framed for you
The platform pieces you'll touch first.
BEST_EX across all investors
Single call, every active investor priced in parallel, top-K returned with the trace.
Two-stage eligibility
Cheap pre-flight discards programs the borrower can't take before the engine runs the full ladder.
Full per-rule trace
Every adjustment is a row in the trace — rule id, condition, value, combine strategy.
Encompass-native import
Standardized Report CCL profile lands 400 fields on day one; AI mapper handles the rest.
Smart 1003
LOs see roughly the 40% of fields that matter for the loan type; QC mode unhides everything.
Lock subsystem + Exception Inbox
Lifecycle, sell-side, lock-desk policy, exception triage with AI suggestions.
Onboarding
What week one looks like.
A pragmatic sequence — from sandbox to first signed quote.
- 1
Day 1: sandbox + sample ratesheets
We seed your sandbox with a sample ratesheet pack so you can run scenarios immediately.
- 2
Day 3: your real ratesheets
Email-in or portal automation pulled in by the ingestion service. We tune the mapping templates against your investors.
- 3
Week 2: scenario migration
Common LO scenarios migrate to saved scenarios; saved-scenario quick-rerun fits muscle memory.
- 4
Week 3: lock-desk policy
Configure your desk's cutoff time, extension fee curves, free-extension grace, and allowed lock periods.
- 5
Week 4: webhooks + downstream BI
Subscribe pricing.computed and locks.* events; we wire them to your BI / hedge stack.
- 6
Production cutover
Run parallel with your current engine for one week; cut over once the trace and the prices reconcile.
Two-stage eligibility cut our lock-day surprises to almost zero. The drill-down is the first time a pricing engine has actually answered 'why' for our compliance team.
Frequently asked
Specific to your operating model.
Do we have to switch off our current pricing engine to start?
No. Most correspondents run RateStack in parallel for one to two weeks, reconcile the trace, then cut over. The platform is read-only with respect to your investor data; you can leave the old engine in place indefinitely if you want.
How do you handle our investor-specific overlays?
Overlays are rules. Investor-level overlays apply across the investor's programs; program-level overlays apply to one program. Both sit in the same eligibility table. No code changes for new overlays.
Can we control which investors are visible to which entities?
Yes. The capability catalog includes per-investor visibility scopes; entities (BRANCH / REGION / TEAM / DIVISION) see only the investors assigned to them. Org-level admins can override. Cross-tenant grants are TTL-bounded and audit-chained.
What about rate-lock fees?
Configured per-org as a fee curve (rate vs. lock period). The lock endpoint applies the curve and writes a journal entry with the fee that fired. Free-extension grace windows are first-class.
How fast is the BEST_EX call?
Sub-200ms p95 on the warm path with 50 active investors. Cold-cache cold-ratesheet lands around 800ms p99. The three-tier cache is invalidated by NATS, not TTL drift, so cold paths are rare in steady state.
Ready when you are
See correspondent lenders on RateStack.
Live demo with your real ratesheets, your real scenarios, and an honest read on whether the platform fits your team.