"Follow the money" is advice so old it has become wallpaper, yet the working method behind it is rarely explained. The core insight is simple: money leaves tracks in documents that exist for other reasons. Companies must register, ships must fly flags, courts must publish judgments, customs must record what crossed a border. An investigator is someone who reads those ordinary records with an unusual question in mind: who benefits here?
Start with companies, because structures end there
Almost every cross-border arrangement, legitimate or not, is held in companies. So the first tool is the corporate registry, and the skill is reading one properly.
The United Kingdom's Companies House is the most-used free registry in the world and a fine place to learn. A full company profile tells a story in small details: incorporation date, registered office (often an accountant's address serving hundreds of firms), directors and their other appointments, the filing history, and whether accounts are full or abbreviated. Charges, the registry's record of loans secured against the company, are among the most under-read documents in finance: a small firm with a large secured lender has revealed exactly who its money actually answers to.
Most of the world is now searchable in some form. Registry quality varies enormously, and that variance is itself information. A structure that routes through jurisdictions known for minimal disclosure, layered two or three deep, is not illegal and is not automatically suspicious. It is, however, a deliberate design choice, and the honest question is what the design achieves. Sometimes the answer is tax efficiency or privacy from competitors. Sometimes the answer is that no single registry shows the full picture, which is the point of layering.
Read the connections, not just the entities
One company tells you little. Networks tell you a lot. Three habits turn isolated records into a picture:
Common registered addresses. When dozens of unrelated companies share one maildrop, the address is functioning as a service, and whoever operates that service knows something all those companies have in common.
Overlapping directors and shareholders. The same name appearing across supposedly independent firms, as director, secretary or beneficial owner, links them. Registry search tools and the free datasets collected by investigative groups such as the Organized Crime and Corruption Reporting Project make this practical rather than theoretical.
Timing around events. Filings made days before a court case, a sanction or a divorce are rarely coincidence. Registers timestamp everything, and timestamps line up with the news.
The sanctions layer
Sanctions lists are public, free and updated constantly, which makes them the rare investigation tool that is also an official primary source. The United States Treasury's OFAC lists, the United Kingdom's and European Union's equivalents, and consolidated views built by projects such as OpenSanctions let anyone check whether a name, vessel or company appears on any major list.
Two cautions keep this rigorous. First, names collide; identical names are common across the world, so a hit requires corroboration before it means anything. Second, absence from a list proves absence from that list only. Much of what investigations care about, from fronts for designated parties to politically exposed wealth, never appears on any list at all.
Ships, planes and physical goods
Money moves as value, and value moves physically. Vessel movements are publicly trackable through automatic identification system data, and the mismatch between a ship's signal and its physical position has become one of the defining evidence types of this decade's sanctions enforcement. Gaps in a vessel's track, signals switching off near sanctioned ports, and ship to ship transfers in quiet waters are each individually innocent sometimes. Patterns of them are what enforcement teams, and journalists, build cases on.
Customs and trade data work the same way from the goods side. Bill of lading records for imports into the United States are searchable through commercial and some public sources, and trade statistics from official agencies let you test claims: if a country reports exporting a product to a partner at volumes the partner never reports importing, someone is misdeclaring something, and the discrepancy itself is the finding.
Court records, the most underused free source
Litigation is where hidden structures surface, because lawsuits force disclosure under penalty. Chancery and commercial courts in major financial centres publish rulings describing, in judges' flat prose, arrangements that participants spent decades keeping quiet. Free services aggregate judgments from many jurisdictions, and insolvency notices, another public record, frequently list the full creditor and ownership picture of a failed venture.
Before any leaked database, before any anonymous tip, a serious investigation asks: has anyone sued anyone about this yet? The answer is more often yes than you would expect.
Know the honest limits
Three boundaries keep this kind of work truthful. Bank records are closed; nobody outside the financial system and its regulators sees account flows, so any narrative built without them is inference from surrounding records. Beneficial ownership registers have exceptions, and the biggest structures are built inside those exceptions, trusts above all. And correlation in registries is not wrongdoing; the overwhelming majority of layered structures exist for mundane reasons, which is precisely why the interesting ones can hide among them.
What the public records can do, done patiently, is narrow the world. Not prove the story, but shrink it to a small set of people, companies, ships and dates that the official record itself connects. That shrinking is the real craft. Everything dramatic in published financial investigations stands on this quiet, clerical foundation.