Part V · Chapter 13 · Predators · lesson 3

Where the money comes from

14 min

You are going to run one. Not because operating a scheme is an interesting skill, but because the collapse date is computable from the first day, and computing it yourself is the only way the number stops being abstract.

Then we turn it around, because a reader who has only ever sat in the operator's chair has learned the wrong half.

The setup

Ten people give you 10,000 each. You have 100,000 and you have promised 20% a year, paid out annually in cash. You are not investing any of it — that is what makes this the structure rather than a bad fund.

New deposits arrive at 50,000 a year, every year, from people who have heard that the payments are reliable. They are reliable. That is the product.

Here is your cash position, year by year. Follow it rather than reading it.

Year 1. You owe 20,000 on the original 100,000. New deposits bring 50,000. You pay, and hold 130,000. Total deposits taken: 150,000.
Year 2. You owe 30,000. Cash after the new 50,000 and the payout: 150,000. Deposits taken: 200,000.
Year 3. You owe 40,000. Cash: 160,000. Deposits taken: 250,000.
Year 4. You owe 50,000, which is exactly what came in. Cash: 160,000, unchanged for the first time.
Year 5. You owe 60,000. Cash: 150,000, and falling.
Year 6. You owe 70,000. Cash: 130,000.
Year 7. You owe 80,000. Cash: 100,000.
Year 8. You owe 90,000. Cash: 60,000.
Year 9. You owe 100,000. Cash: 10,000.
Year 10. You owe 110,000 and you have 60,000. It ends here.

Predict

In which year did the scheme become arithmetically doomed?

Year 10, when the cash ran outYear 4, when the payout first equalled the new depositsYear 1, since there was never any investmentIt depends on when people ask for their capital back

Where the money on the statements comes from

The cash arithmetic above assumes everyone takes their payment in cash. Most do not — most reinvest, because a return like that is worth compounding, and the operator encourages it because a reinvested payment costs nothing to make.

That version does not collapse in year 10. It collapses later and much larger, because the balance on the statements is now compounding while the money in the account is not.

The model below runs exactly that. The starting deposits, the yearly new deposits and the promised rate are the three numbers from the setup. What it calls growth is money that has never existed.

Compound growth

Hypothetical model, not a forecast

Grows a starting balance plus a fixed yearly contribution at a constant rate.

It assumes

Deposits arrive at exactly 50,000 a year and never slow down or accelerate.
Every participant reinvests, so no payment leaves the account in cash.
Nobody withdraws their capital at any point during the whole period.
The promised rate is credited once a year and never varies from it.

It ignores

Tax on growth, income or withdrawal
Platform fees, spreads and trading costs
Inflation — every figure shown here is nominal
Any year in which the return happens to be negative

Find the year in which more than half of what participants believe they own has never existed.

Years5

Final balance: 620,912
Total paid in: 350,000
Growth: 270,912
Growth as a share of the final balance: 0.436

Shown on statements
Money actually received

Two lines from the same point. Money received rises in a straight line; the statement balance curves away above it and keeps widening.
Period	Shown on statements	Money actually received
0	100k	100k
1	170k	150k
2	254k	200k
3	355k	250k
4	476k	300k
5	621k	350k

Not there yet — keep moving the controls.

By year seven, more than half of the wealth on those statements is a number somebody typed. The participants can see the balance, can request a withdrawal and receive it, and can recommend the arrangement to a friend in complete good faith. Everything they can observe is working.

The defining feature is not dishonesty about returns. It is that inflows are the source of outflows, which makes the whole structure a claim on future recruitment rather than on any productive activity.

Two consequences follow, and both are counterintuitive. The scheme is at its most convincing shortly before it ends, because the payment record is longest then. And the participants who lose everything are overwhelmingly the ones who joined last, which means the number of victims grows right up to the final day.

Why growth is the only escape and also the trap

Look again at the year-4 crossover. The scheme survives exactly as long as new deposits exceed the payments owed, and the payments owed grow with everything ever deposited.

So the deposits must grow at least as fast as the promised rate, forever. That is a requirement for exponential growth in recruitment, which fails for a reason that has nothing to do with anybody's diligence: there are a finite number of people, and each round needs more of them than the last.

This is why the structure has an end date rather than a risk of ending. It also explains a feature that looks like greed and is actually necessity — schemes raise the promised rate over time, or add a recruitment bonus, because the only way to postpone the crossover is to make the inflow grow faster.

Now turn it around

What would you do

A friend of a colleague runs a small private arrangement for about forty people. It has paid 1.5% a month for four years without a missed payment, including through a bad year for markets. The stated mechanism is short-term lending to small businesses that banks turn down, at rates that make the returns work. Withdrawals have always been honoured, usually within a week. Two people you know have taken money out and been paid. New participants are accepted by introduction only.

You have been offered a place. There is no deadline and nobody is pushing you.

Optimising for: Working out what would have to be true for the payments to be coming from something other than deposits — not deciding whether the person is honest.
You cannot know: Whether the loans exist, and who could confirm it independently of the operator
What proportion of current payouts is funded by money that arrived this quarter
Whether anyone has asked for a large amount back at the same time as others

Ask for the audited accounts and a list of the underlying loans.Ask what happens if a third of the participants ask for their capital back in the same month.Ask to speak directly to two of the borrowers, chosen by you from the loan list.Put in a small amount for a year and see whether the payments arrive.

The uncomfortable part

Nothing in that scenario permits a verdict, and you should be suspicious of any version of this lesson that hands you one. The arrangement described might well be a real lending business, and small private lending at those rates does exist.

What the exercise produces is not a judgement but a list: here is what would have to be true, here is which of those things I can check, and here is what I would be relying on the operator's word for. That list is the deliverable, and in this case it is short and the items on it are load-bearing.

The last lesson of this track turns that into a reusable set of questions.

Worth remembering

A scheme paying returns out of new deposits has a collapse date fixed by arithmetic, not a risk of collapsing.
It becomes doomed at the point payouts first equal inflows, which can be many years before anything visible goes wrong.
The payment record is longest and most convincing immediately before the end, and most participants join in that window.
Where returns are reinvested, the balance on the statements compounds while the money does not, so most of the reported wealth never existed.
Receiving payments and successful small withdrawals are produced by both the real and the fraudulent version, so neither is evidence.