# Why Money Launderers Love $100 Bills

> Notes from an Odd Lots conversation on money laundering: the paradox of record cash in circulation, the scale of trade-based laundering, and a compliance system that costs a fortune and doesn't work. Educational, not investment advice; no stock picks or price targets.

Published: 2026-09-07
Locale: en
Tags: money laundering, cash, seigniorage, compliance, trade finance


![An expensive wristwatch catching the light at an airport customs desk, a queue of travelers receding into the distance](/covers/oddlots-2026-09-07-why-money-launderers-love-100-bills-cover.png)

> Counterfeit coin multiplies by the day, while the five grains do not… the law cannot forbid it, the decree cannot stop it.
>
> — Jia Yi, "Memorial Against Private Minting," recorded in the *Book of Han, Treatise on Food and Money* (Western Han, c. 175 BCE; translation mine)

A memorial written twenty-two centuries ago, about private coinage nobody could stamp out. My first thought after this episode: the form has changed, and that last clause needs no editing at all.

## What This Episode Is About

On the September 7, 2026 episode of Bloomberg's Odd Lots, the hosts talk to journalist and author Oliver Bullough about money laundering. The opening is domestic: they recall touring the Chicago Fed's cash facility, the peculiar smell of banknotes, the pinks and oranges you only notice once dollars are stacked. Then the number lands — of the $2.4 trillion in circulation, 85% is in hundred-dollar bills.

Which raises the obvious problem. One host says he takes twenties and tens out of the ATM and has no hundreds in his pocket. So who is holding all of them?

## The Main Points

**1. The banknote paradox: demand is falling, supply is exploding.** Cash is now roughly 9% of transactions in the UK and 13% in the US, yet circulation keeps setting records — nearly $2.5 trillion in the US, around €1.6 trillion in the eurozone. Andrew Bailey named the puzzle formally in 2009 when he was the Bank of England's chief cashier. His candidate explanations then were low rates, more ATMs, and post-crisis distrust of banks. All three have since reversed, and the pile keeps growing; sterling in circulation has doubled since that speech. Bullough's analogy: if Netflix had won and VHS production were hitting annual records, we'd all be asking where the tapes were going. Because it's banknotes, everyone says "store of value" and moves on.

![Two lines run in opposite directions: the share of transactions made in cash falls year after year, while the volume of banknotes in circulation climbs year after year, and the two cross near the middle.](/figures/cash-paradox-two-lines-en.svg)

**2. "Store of value" doesn't survive contact with the data.** Central banks reason that if cash isn't a medium of exchange, it must be a store of value. But their own surveys put the average American adult at about $430 on their person and at home, while circulation works out to more than $7,000 per man, woman and child — an order of magnitude apart. Granting that roughly 65% of dollar notes sit outside the US only moves the question one step: abroad, they still have to be doing something.

![Two bars of very unequal length: the short top bar is the roughly 430 dollars an ordinary person keeps on hand, and the much longer bottom bar is the 7,000-plus dollars in circulation per person.](/figures/wallet-vs-per-capita-cash-en.svg)

**3. The "2–5% of global GDP" figure is a thirty-year-old guess.** It traces back to Michel Camdessus at the IMF in the late 1990s, built on two economists' work at the time. Bullough is upfront that nobody knows — criminals don't file honest returns. The uncomfortable implication sits inside the number: if the criminal share of the global economy is the same now as in the 1990s, three decades of effort have at best kept it from growing faster than everything else.

**4. Two hundred billion dollars a year of compliance spending, feeding a black hole.** That's the LexisNexis estimate for global anti-money-laundering compliance. Bullough notes the same money would solve world hunger and provide clean water and sanitation to everyone on earth, with $50 billion left over. Banks employ tens of thousands of compliance officers filing millions of suspicious activity reports a year, and no country resources its law enforcement to read them. This was the part that stuck with me: the incentives are all downside. Miss something and the fine starts at a billion; get it right and nothing happens. Under that structure people file more, until they're covered.

**5. The big flows never touch the banking system.** Danske Bank was accused of moving about $130 billion over several years for suspicious Russian clients. Global cash smuggling runs into the hundreds of billions annually. Bigger than both is trade-based laundering, which Global Financial Integrity puts near a trillion dollars a year. The mechanism is to move goods rather than money: fentanyl north, tractors and farm equipment south; European handbags to China, cocaine from South America to Europe, closing a triangle. Bullough points out the Medici did the same thing — deposit in Venice, collect in Bruges, with no bullion crossing the Alps, only silk going north and wool coming south, and the money movement buried in the paperwork of the trade.

![Three bars of different lengths: the top bar, the big bank cases, is the shortest and covers several years combined; the middle bar, cash smuggling, is longer; the bottom bar, trade-based laundering, is the longest and is per year.](/figures/laundering-scale-three-bars-en.svg)

**6. A metric economics rarely uses: value per unit of volume.** One host raises this at the end and it's the sharpest line in the episode. Carry more than $10,000 in cash through an airport and you declare it; wear a million-dollar watch and nobody looks twice. On that axis, a hundred-dollar bill, a Rolex, a designer handbag and a kilo of cocaine belong to the same family. The counterexample is a large couch. Crypto — specifically stablecoins, in this conversation — hasn't replaced cash; it stacks on top. Cash works at street level, gets converted, and the value crosses borders instantly.

![A scatter chart with volume on the horizontal axis and value on the vertical axis: a watch sits top-left with small volume and high value, a sofa sits bottom-right, and a stack of hundred-dollar bills lands near the watch.](/figures/value-per-volume-en.svg)

![Four places are linked by arrows into a loop, each arrow labelled with a kind of goods, and the centre of the loop notes that no money moves.](/figures/trade-laundering-loop-en.svg)

**7. Killing the big note is a collective action problem.** Bullough says plainly he wishes the US would stop printing hundreds, and that nobody but the bad guys would be inconvenienced. But if it did, Europe keeps printing €100s and €200s, criminals switch currency, and the US hands the seigniorage to the ECB. How does seigniorage work? The simple-but-wrong version: a hundred-dollar bill costs about nine cents to print, so you make $99.91. The correct-but-complicated version: you can redeem it at face value any time, so it's an interest-free loan to the government. Against nearly $40 trillion of federal debt, $2.5 trillion is about 6% the government pays no interest on. Europe did retire the €500 note after the 2016 Paris attacks; its nickname was "the Bin Laden," because everyone had heard of it and nobody had seen one.

## Going Further

### "There are more rules every year — why does nobody seem to get caught?"

This is the first thing that clicked for me, and it goes well beyond laundering.

A system has two separate properties: **input** and **effectiveness**. We habitually use the first as a proxy for the second — $200 billion spent, tens of thousands of staff, millions of filings, surely something is working. This episode pulls them apart and the answer is no. The break is at the last mile: the filing side has legal obligations and penalties behind it, the reading side has no headcount. Throughput of the whole chain is set by its narrowest segment, and nobody owns that segment.

![A pipe that starts wide and narrows sharply: the entrance is millions of filings a year, the middle is the staff who can actually read them, and the exit is so thin it is barely visible.](/figures/narrowest-segment-funnel-en.svg)

I've been applying this to companies and it holds up. When a company reports research spending, a new system, a certification — those are inputs. My follow-up is: where is the narrowest link, and did it widen too? A firm expanding production without adding the test equipment that qualifies the line has capacity on paper only. The same question inverts usefully: when an industry is legally compelled to spend, who receives the money? Compliance software vendors are selling into spending that will happen regardless — a real business, but one driven by regulatory density rather than by how much crime it prevents. Separating "there is demand" from "it solves the problem" is the most portable tool I took from this.

### "A number has been quoted for thirty years — should I believe it?"

That "2–5% of global GDP" sounded weighty to me at first. The IMF, three decades, everyone uses it. The provenance in the interview takes it apart: an estimate by one managing director in the late 1990s, resting on two economists' contemporaneous work, described by him at the time as a guesstimate.

I've been burned by this shape before. Reading research, I'd see a figure cited by three reports and feel it had been verified three times; chasing the source, all three were quoting one paper. **Citation count is not verification count.** Now I ask three questions before a number is allowed to drive a decision: who computed it first, by what method, and has anyone recomputed it by a different method and landed nearby. The third one carries the weight. This laundering estimate has been attacked from macro and micro directions and keeps returning to the same band, which earns it some credit — but it still can't tell you whether this year is higher than last, because its error bar dwarfs the annual change.

![Two panels side by side: on the left, arrows from three reports all point back to the same single source; on the right, three different methods each arrive at a similar answer on their own.](/figures/citations-are-not-verifications-en.svg)

A number can be roughly right and useless for trend at the same time. I used to run those together.

### "Everyone knows it doesn't work — so why doesn't it change?"

Bullough gets visibly annoyed near the end; he says looking at the situation makes him furious most of the time. His explanation is the cleanest account I've heard of why bad systems survive: **profits concentrate, harm disperses.**

The hundred-dollar bill is the cartels' favorite instrument. That is a catastrophe for Colombia and Mexico and a cheap funding line for the United States. The City of London has done well for decades as a conduit, with the damage spread elsewhere. The beneficiaries are named, quantifiable, and lobby; the injured are scattered across dozens of countries, uncountable, unrepresented. Nobody in the picture has to do anything wrong for it to stay put. Kenneth Rogoff has argued this case for years; what defeats it isn't the logic, it's getting everyone into one room.

![On the left a single large concentrated circle stands for the gains, on the right dozens of scattered small dots stand for the harm, and the two areas are comparable while only one side is visible.](/figures/concentrated-gains-diffuse-harm-en.svg)

I've tried turning this on myself. For something I've been leaving alone, I ask: who bears the cost of not dealing with it? If it lands on someone else and disperses until it's invisible, it will still be there next year, and no amount of waiting for clarity will change that. An untracked decision, a position whose original reason expired, a process nobody owns — most of them have this shape.

## Worth a Look

- Odd Lots, "Why Money Launderers Love $100 Bills," September 7, 2026, with Oliver Bullough
- Oliver Bullough's book, *Everybody Loves Our Dollars*, referenced throughout
- Andrew Bailey's 2009 remarks on the banknote paradox, the formal starting point
- Fed economist Ruth Judson's research on dollar banknotes held abroad
- Global Financial Integrity's estimates of trade-based laundering
- Kenneth Rogoff's long-running case for retiring high-denomination notes

## One Thing to Take Away

One idea: **to judge whether a system works, look at its narrowest segment, not at what it cost.**

Spending is visible, so we let it stand in for results. But actual output is set by the link nobody is watching — here it's "nobody reads the reports." On you it will be something else.

Here's a thing I've tried, if you want it: pick something you're putting effort into but can't name a result for, break it into three to five links end to end, and next to each one write "who receives the output of this step?" One link will have no name next to it. That's your answer, and it's rarely the link you've been adding effort to.

I ran this on my own fitness log. What I kept adding to was "record it more carefully." The link with no name was "nobody, including me, ever reads it afterward" — and no amount of effort at the front makes the back end appear.