In the basement of the Antwerp Diamond Centre at 9/11 Schupstraat, over the weekend of 15–16 February 2003, a small team of Italian thieves opened a vault that the industry had treated as unbreakable and emptied 109 of its 189 safe-deposit boxes. The man who organised the operation was Leonardo Notarbartolo, a 51-year-old jeweller from Turin who had rented an office in the building three years earlier and held a box in the very vault he would later loot. The amount taken has never been fixed with certainty, but insurers and the diamond trade put it at more than US$100 million in loose stones, gold and cash, which makes it, by most reckonings, the largest diamond theft in history.
The outcome is not in doubt, and this file states it plainly: the principals were caught and convicted, but the diamonds were not. Within days of the break-in, investigators from Belgium’s Diamond Squad had a suspect; within two years a court had a verdict. Notarbartolo was sentenced in 2005 to ten years in prison; three named accomplices received five years each. He was paroled in 2009. The stones, by contrast, were gone — fenced, recut or scattered through channels that no recovery effort has unwound, and the overwhelming majority remain missing more than two decades later.
What makes the case a permanent teaching text is the asymmetry between the two halves of the crime. The entry was a near-flawless exercise in patience, reconnaissance and the quiet defeat of layered technology — ten distinct security systems neutralised without triggering an alarm. The exit was undisciplined. A partially eaten salami sandwich, envelopes and a half-burned bag of refuse, dumped along a rural roadside outside the city, carried the DNA and paper trail that tied Notarbartolo to the scene. The vault was beaten by craft; the gang was beaten by garbage.
The episode also exposed a complacency at the centre of the world diamond trade. The Antwerp Diamond Centre vault was rated among the most secure in Europe, yet it had been denied an insurance policy on account of its flaws, and the human and procedural gaps around the hardware proved far easier to exploit than the hardware itself.
At about 6:40 on the morning of 26 November 1983, six armed men walked into Unit 7 of the Heathrow International Trading Estate in west London, into a warehouse run by the security firm Brink’s-Mat, expecting to leave with perhaps £3 million in cash. They left instead with 6,800 bars of gold bullion weighing roughly three tonnes, together with diamonds and traveller’s cheques, a haul valued at the time at £26 million — equivalent to several hundred million pounds today. It remains one of the largest robberies in British history, and it was made possible by a man on the inside: security guard Anthony Black, the brother-in-law of one of the raiders.
The outcome can be stated without suspense. The men were caught, and several were convicted, because the inside man broke within days. Black confessed in December 1983, named his brother-in-law Brian Robinson as one of the gang, and was sentenced to six years for his part. In December 1984 Robinson and the raid’s other organiser, Micky McAvoy, were each jailed for 25 years. The investigation then widened from the robbery to the gold, and in 1986 Kenneth Noye, who had melted and recast the bullion to launder it back into the market, was convicted of handling and sentenced to 14 years. The case closed in court many times over. What never closed was the recovery: the overwhelming majority of the gold was never found.
The robbery is a case study in two failures of security, one belonging to the victim and one to the criminals. Brink’s-Mat lost the gold because it trusted an employee who supplied a copied key and the layout of its defences; an alarm and a strongroom are worth little when a guard opens the door from within. The gang lost its freedom because the same insider who made the raid possible was the most obvious thread for detectives to pull, and he unravelled almost immediately.
The episode left a long and bloody wake. Disposing of three tonnes of traceable gold drew in launderers, financiers and fences across two decades, and an unusual number of those connected to the proceeds met violent ends — a pattern the British press called “the curse of Brink’s-Mat.” The bars themselves, recut and remixed, dissolved into the legitimate gold supply and could not be retrieved.
Shortly after midday on 12 July 1987, two well-dressed men walked into the Knightsbridge Safe Deposit Centre on Cheval Place in central London and asked to rent a box. Admitted to the strongroom, they drew handguns, overpowered the manager and guards, hung a “closed” sign at the street door, and let in accomplices. Over roughly two hours the crew forced open about 114 of the centre’s safe-deposit boxes and left with a haul commonly estimated at £40 million to £60 million in cash, jewellery and valuables — by some reckonings the most lucrative robbery in Britain to that point. The operation was led by Valerio Viccei, an Italian fugitive already wanted at home for a string of armed robberies, and it depended on a man on the inside: Parvez Latif, the centre’s manager, who was in debt and let the raiders through the door he was paid to guard.
The case is filed as closed, and as recovered, though both descriptions carry qualifications worth stating up front. The crew was identified almost at once and dismantled within a month, because Viccei left a bloody fingerprint in the looted vault. Rather than arrest him immediately, police placed him under surveillance, mapped his associates, and swept them up in coordinated raids on 12 August 1987. In 1988 the courts handed down long sentences across the gang — Viccei himself receiving 22 years — and a substantial quantity of cash and valuables was recovered, though not the whole haul. The closure was complete; the financial restitution was partial.
What lifts the raid above an ordinary armed robbery is the contrast between its planning and its single, fatal lapse. The entry was disciplined: a legitimate-seeming request for a box, an inside man to wave it through, hostages taken quietly, the premises sealed behind a closed sign so the looting proceeded unseen from the street. Yet the same scene held an injury and a smear of blood carrying Viccei’s fingerprint, and that one mark converted an anonymous, near-perfect crime into a named one. The vault was beaten by an insider; the gang was beaten by a print left in their own blood.
The capture of Viccei himself added a final, almost theatrical detail. Having fled abroad with the gang’s success behind him, he could not leave his Ferrari Testarossa in England, and returned to arrange its shipment; police, already on him, took him in a violent street arrest. The robbery that had turned on patience and an inside man ended on the opposite quality — the vanity of a fugitive who came back for a car.
On the night of Friday 12 September 1997, six men entered the Dunbar Armored regional depot on Mateo Street in downtown Los Angeles and left roughly half an hour later with US$18.9 million in cash. No shot was fired. The operation was organised by Allen Pace III, a Dunbar regional safety inspector from Compton who had walked the building legally for months and who had been fired the day before for tampering with company vehicles. By volume of cash carried out the door, it remains the largest cash robbery in United States history; only the 2024 Easter Sunday burglary of a Los Angeles money-storage facility, estimated at more than US$20 million, is sometimes ranked above it, and even that comparison is contested.
The outcome is settled, and this file states it without suspense. Pace and his five childhood friends were caught, and the principals were convicted in federal court. Pace was sentenced on 23 April 2001 to 24 years in prison and was released on 1 October 2020. The recovery of the money, however, failed almost entirely. Less than a third of the haul — roughly US$5 to 7 million, depending on the source — was ever traced, leaving in the region of US$12 to 14 million unaccounted for more than two decades later.
The case is studied because its two halves point in opposite directions. The theft itself was a near-textbook inside job: a trusted employee who knew the camera arcs, the guard routine and the vault schedule, paired with a small crew and an alibi built at a house party before the raid. The crew left almost no forensic evidence at the scene. What it could not control was what happened to the cash afterward. The investigation that broke the case did not begin with the robbery at all; it began two years later, when one robber paid a real-estate broker with banknotes still bound in their original Dunbar currency straps.
The Dunbar robbery is therefore less a story about defeating a vault than about the gap between stealing money and keeping it. The crew solved the first problem with insider knowledge and discipline, and was undone by the second through ordinary impatience and the simple, traceable fact of branded cash.
On the evening of 17 January 1950, seven masked men walked out of the Brink’s, Inc. armored-car depot at 165 Prince Street in Boston’s North End carrying roughly US$2.775 million — $1,218,211.29 in cash and $1,557,183.83 in checks, money orders and securities — in what was then the largest robbery in American history. The operation was organized by Anthony “Tony” Pino, a Boston career criminal who had spent some two years casing the building, and executed by a crew of eleven that included Joseph “Specs” O’Keefe, Stanley “Gus” Gusciora, Joseph McGinnis and Henry Baker. For a time it looked like the title the press gave it: the perfect crime.
It was not. This file states the outcome plainly. The case broke almost exactly six years later, and it broke from the inside. On 12 January 1956 — five days before the federal statute of limitations on the robbery would have expired on 17 January — the FBI arrested the core of the gang. The decisive cause was not forensics or stakeouts but a defection: O’Keefe, embittered over money and nearly killed by a hired gunman, agreed to talk. Eight of the conspirators were convicted later in 1956 and sentenced to life imprisonment; two others died before they could be tried. Of the more than $2.7 million taken, only about $58,000 was ever recovered.
What makes the Brink’s job a durable teaching text is the gap between its operational brilliance and its human fragility. The entry was a masterpiece of patient reconnaissance: the crew had let themselves into the building repeatedly at night, removed and copied the lock cylinders on five successive doors, and rehearsed the raid until it took minutes. Dressed in near-identical Navy pea coats, chauffeur caps and rubber Halloween masks, they moved as an interchangeable unit that left witnesses with nothing to describe. The plan that defeated the building, however, could do nothing about the eleven men who knew the secret, the money they could not safely spend, and the six years they had to keep quiet. Brink’s, a firm whose entire business was the secure movement of cash, had left a count-room reachable through a sequence of ordinary locks that a patient crew could defeat one at a time, with no one positioned to stop five employees being surprised on a quiet weeknight — proof that a hardened target can be undone by the routine around it.