Investment Screening

CIMCalc

Intent-Aware Investment Screening. Computation, Not Advice.

The Problem

Institutional investors review dozens of opportunities for every investment they make. Sovereign funds, pension funds, PE firms, and family offices each face the same structural challenge: screening financial documents — prospectuses, pre-investment due diligence evaluations (PDIEs), annual reports, CIMs, institutional research — is slow, subjective, and inconsistent. Sixty to eighty analyst hours per decision, and the output is a narrative opinion that cannot be reproduced or audited. When the Investment Committee asks “why did you pass?” the answer is judgement, not computation.

Existing AI tools in this space either operate as general-purpose document chatbots — answering questions you already know to ask — or as workflow automation platforms that require manual configuration. Neither provides an independent analytical verdict driven by your specific investment thesis.

The Product

CIMCalc applies an evidence-based disqualification methodology to investment documents — screening not to confirm your thesis, but to stress-test it against 33 forensic financial calculators and 252 risk signals calibrated to your specific investment intent. Upload any financial document. Declare your thesis. Receive a structured risk assessment with a clear verdict and precise source citations traceable to the page.

The critical differentiator: the same company receives a different verdict depending on your thesis. A company with 45% customer concentration is a critical risk for a yield investor — one customer loss breaks debt service — but a monitor-level finding for a strategic acquirer buying the IP, not the customer list. Same document. Same extraction. Different computation.

How Intent Changes Everything

Your Thesis Determines What Matters

Every investment thesis carries a different set of fears. A yield buyer fears margin erosion and customer concentration — because one customer loss can break debt service. A growth buyer fears velocity collapse — because stalled unit economics kill the scaling thesis. A distressed buyer fears hidden liabilities — because undisclosed liens destroy recovery value. CIMCalc encodes these fears into its analytical architecture.

Yield Acquisition

Harvesting stable cash flows

Primary Fear: Margin Erosion

Customer concentration, FCF conversion, revenue predictability, distribution sustainability

Growth Expansion

Scaling a proven model

Primary Fear: Velocity Collapse

Unit economics, churn rate, LTV:CAC ratio, technology scalability

Platform Consolidation

Merging firms for synergy

Primary Fear: Integration Failure

Tech debt, system compatibility, workforce overlap, combined leverage

Distressed Turnaround

Acquiring failing assets to repair

Primary Fear: Uncapped Liability

Hidden liens, litigation exposure, asset coverage, cash runway

Talent Acquisition

Acquiring the team

Primary Fear: Key Person Flight

Founder dependency, IP attribution, retention structures, non-compete enforceability

Strategic IP

Acquiring patents and moat

Primary Fear: Vaporware

IP validity, contamination risk, licensing revenue, competitive moat durability

Regulatory Arbitrage

Acquiring licenses and certifications

Primary Fear: Non-Transferability

License portability, regulatory approval timeline, person-held vs. entity-held

The same company with 45% customer concentration and an Altman Z-Score of 2.82 produces a CRITICAL verdict under a yield thesis (one customer loss breaks debt service) but a PASS under a distressed turnaround thesis (the buyer expects to restructure the revenue base). Same data. Same extraction. Different computation. The intent is not a filter applied after analysis — it recalibrates every threshold, reweights every signal, and shifts what constitutes a material risk.

Output

What the Output Looks Like

CIMCalc report — financial metrics dashboard with computed values and status indicators
CIMCalc report — leverage indicators and signal detail with threshold analysis
CIMCalc report — screening outcome with verdict basis and key observations

Real report sections from CIMCalc. Every number traces to a page in the source document. Every verdict includes the mechanism that triggered it.

Process

How It Works

01

Upload

Your financial document — prospectus, PDIE, annual report, CIM, institutional research, SEC 10-K/10-Q, investor presentation, debt memo, or business plan.

02

Declare

Your investment intent. CIMCalc supports yield acquisition, growth expansion, platform consolidation, distressed turnaround, strategic IP, talent acquisition, and regulatory arbitrage.

03

Receive

A structured screening report: a clear verdict (Pass, Monitor, Material, or Critical), domain-by-domain risk analysis across 9 categories, every conclusion traceable to the source document — plus prioritised verification questions that a deal professional should ask based on the specific risks found. The report doesn’t just tell you what’s wrong — it tells you what to ask next.

Capabilities

Key Capabilities

Evidence-Based Disqualification
CIMCalc screens to stress-test, not to confirm. The methodology preserves the abandonment option at its lowest-cost exercise point — surfacing fatal flaws before analyst time and capital are committed.
Intent-Aware Verdicts
The same document produces different risk assessments depending on your investment thesis. Intent recalibrates thresholds, reweights signals, and shifts what constitutes a material risk. Seven canonical intents, jurisdiction-aware threshold overrides.
33 Financial Calculators
Forensic-grade financial models including Altman Z-Score, Beneish M-Score, Piotroski F-Score, Dechow F-Score, Sloan Accrual Ratio, Revenue Manipulation Index, DuPont Decomposition, Quality of Earnings, DSCR, FCF conversion, and 24 more — computed automatically from extracted data.
252 Risk Signals
Organised across 9 risk domains: earnings quality, cash flow sustainability, leverage & solvency, growth credibility, operational efficiency, customer & revenue, management & governance, deal structure, and valuation.
Full Traceability
Every data point, calculation, and verdict links back to the source document with page references. Investment committees can independently verify any conclusion. Run it twice, get the same answer.
Multi-Layer Verification
Extraction consensus, unit reconciliation, cross-period anomaly detection, and dual-run validation ensure analytical integrity. A Data Foundation grade tells the IC how much data the engine had — and how much was missing.
Deep Forensic Analysis
Every line item — quantitative and qualitative — receives forensic-grade scrutiny. Financial metrics are not summarised; they are decomposed through models like DuPont (margins × turnover × leverage), Beneish M-Score (8 manipulation indicators), and Sloan Accrual Ratio (cash quality). Qualitative claims — management projections, market positioning, competitive advantages — are cross-referenced against the extracted data and flagged when unsupported.
Real-Time Claim Validation
CIMCalc goes beyond the document. Web-derived signals — peer valuation multiples, market positioning data, governance and regulatory indicators, geographic risk factors — are extracted in real time to simulate and confirm claims made in the source document. When a CIM claims "market-leading margins," the engine checks against sector benchmarks. When a prospectus projects growth, the engine compares against observable industry data.
Prioritised Verification Questions
Every report includes deal-specific questions derived from the actual findings — not a generic checklist. If CIMCalc detects a potential related-party connection between an investor and a top customer, it asks: "Confirm arm's-length basis for the 20% revenue relationship." If margins have been statistically flat for 5 years, it asks: "Request disaggregated monthly P&L to verify." These are the questions a senior deal professional would ask — computed from the evidence, not templated.
Any Financial Document
Prospectuses, PDIEs, CIMs, annual reports, SEC 10-K/10-Q (via EDGAR iXBRL), institutional research, investor presentations, debt memos, and business plans. Malaysian, US, and international document formats.
Jurisdiction-Aware
Threshold calibration adjusts for jurisdiction-specific norms. Malaysian market defaults (customer concentration, board independence, RPT thresholds, leverage norms) are built in. Shariah screening thresholds aligned to SC Malaysia SAC guidelines.

Use Cases

How Institutions Use CIMCalc

Pre-Investment Screening

Run any document through CIMCalc before committing analyst time — prospectuses, PDIEs, annual reports, institutional research. Surface earnings manipulation, governance red flags, and liquidity traps in hours, not weeks. For every investment backed, institutions review approximately 80 opportunities. CIMCalc compresses the time to a defensible "no" — preserving capital and analyst bandwidth for the deals that matter.

Portfolio Monitoring

Re-evaluate existing holdings against the original investment thesis. Has customer concentration increased? Has the margin profile degraded? The computation removes anchoring bias — the analyst who championed the buy does not grade their own homework. Research shows investors hold losing positions approximately 50% longer than winners, costing 3–6% annually in forgone returns.

Thesis Resolution

When two teams disagree on the same company — growth mandate versus income mandate — give each a verdict calibrated to their thesis. The data resolves the argument, not seniority. Same document, same extraction, different computation, different verdict.

Audience

Who It’s For

Sovereign Wealth & Pension FundsPrivate Equity & Search FundsInvestment BanksCorporate DevelopmentFamily Offices

DocScreen: The Engine Behind CIMCalc

CIMCalc is built on DocScreen — a general-purpose document screening engine that can evaluate any document against any checklist. CIMCalc is DocScreen with the investment knowledge graph pre-loaded. For organisations with screening needs beyond investment due diligence — lending, procurement, compliance, insurance underwriting, legal review — DocScreen can be configured with custom criteria and thresholds.

See CIMCalc in Action

Request a demo using your own documents. You judge the output.