Home / Data analysis
Data analysis
What no single document says.
Whether a company was insolvent, since when, who knew about it and where the money went is rarely written in one document. It emerges from accounting records, bank statements and correspondence together. We bring these sources together, analyse them statistically and by content, and back every result with evidence.
Solvency
Since when was there not enough money?
For every reference date we compare the available funds with the liabilities due – even for very large, complex or “broken” accounting records. This shows when a gap opened, how large it was and whether it closed again.
- Illiquidity analysis with a liquidity balance per reference date
- Solvency test and ratio analysis
- Cash-flow analysis, outflows of funds
- Analysis of accounts and annual financial statements
Background (German law): According to the Federal Court of Justice, a debtor is generally illiquid if it cannot close a liquidity gap of 10 % or more within three weeks (BGH, judgment of 24 May 2005 – IX ZR 123/04). Our analysis provides the figures; the legal assessment remains with you.
Correspondence
When did the tone turn – and which days stand out?
Nobody reads tens of thousands of e-mails in full. Two statistical methods show where it is worth looking: the tone over time and the days on which certain topics were written about unusually often.
Sentiment analysis
Every word of the correspondence is rated positive or negative. Averaged over weeks, it shows when routine turned into reminders, excuses and threats – and for which days the letters are worth reading.
Principal component analysis (PCA)
The main directions are derived from the hits of all search terms per day. Each day is compared with hundreds of randomly shuffled data sets – same terms, same amounts, only distributed differently. What chance does not explain is marked, with its effect size and the terms that drive the day.
Networks and audits
Who knows whom?
Senders, recipients and copies form a web. Connections stand out that should not exist according to the files: an intermediary between management and a preferred creditor, contacts shortly before unusual payments, circles that only correspond at certain times.
- Fraud and breach-of-trust audits (fraud auditing, fraud detection)
- Bribery and corruption audits
- Asset tracing: where did assets go?
- Internal audit and statistical evidence
- Communication analysis
Method
Traceable, not a black box
Results that are meant to hold up in court must be explainable. That is why we work with our own software and with methods whose limits we state.
Every value leads back to the bookings, documents or e-mails it comes from – one click from the chart to the evidence.
The report describes the procedure and also states how many anomalies would be expected by chance alone.
Analyses run on our own servers on our premises, without third-party cloud or AI services.
- Analysis of insolvency case data
- Data mining for controlling
- Process optimisation
- 3D visualisation of results
Contact
Let’s talk.
Tell us your question – we will tell you which data we need and how we would proceed.