Atrium's risk intelligence engine combines machine learning models, servicer data, county filings, and market signals to produce risk grades, predicted LTV, estimated NOI, and DSCR projections on every monitored property. Then it watches — alerting you when conditions change.
Risk grades, ML-driven valuations, and real-time distress monitoring — updated continuously as new data arrives.
Every monitored property carries a risk grade from 1 (pass) to 9 (distressed), derived from loan performance, DSCR, maturity status, watchlist codes, and market conditions. The score updates as new data arrives.
Machine learning models trained on comparable sales, tax assessments, rent trends, and property characteristics estimate current market value — even between appraisals. See LTV drift before the next valuation cycle.
Models estimate net operating income and debt service coverage using rent comps, occupancy data, expense ratios, and loan terms. Identify loans where coverage has deteriorated below underwritten levels.
Court filings, lis pendens, tax liens, foreclosure notices, maturity defaults, and bankruptcy petitions — monitored continuously and matched to properties, borrowers, and lenders in your coverage universe.
CMBS and agency servicer watchlist status changes with underlying trigger codes. See when a loan moves to the watchlist and why — DSCR decline, occupancy drop, maturity default, or borrower request.
Configure surveillance across any portfolio: your own book, a bank's CRE exposure, a REIT's holdings, or a market. Get risk-stratified views with drill-down to individual properties and loans.
The data behind the edge.
From early distress detection to portfolio stress testing, Atrium's risk engine gives you the predictive layer that backward-looking data can't provide.
See deterioration before it surfaces in quarterly filings. Our models flag properties where estimated DSCR has dropped below 1.0x, where LTV has drifted above underwritten levels, or where market comps suggest value impairment.
Run scenario analysis across your portfolio: what happens to DSCR if rates rise 100bps? Which loans mature into a down market? Where is your LTV cushion thinnest? Data-driven answers, not assumptions.
Atrium's research team publishes 2,800+ credit signals with loan-level analysis — from bank earnings previews to borrower bankruptcy deep dives. Each signal includes the underlying data and property-level evidence.
When a loan hits the servicer watchlist, Atrium connects it to the property, borrower, county filings, and comparable sales. Turn a watchlist code into an actionable credit assessment in minutes.
"Our ML model flagged a 14% estimated value decline on a 300-unit multifamily six months before the annual appraisal confirmed it. The DSCR had dropped from 1.35x to 0.98x. We restructured the loan before it defaulted."
Risk grades, ML-driven analytics, and real-time distress surveillance across the entire capital stack.