Financial Fraud: The Patterns Behind Most Losses

Most financial losses run through a small number of patterns rather than an unlimited variety of stories. Impersonation borrows trust that belongs to somebody else. An advance fee collects a payment against something that will not arrive. A pump moves a price with attention and sells into it. The stories change constantly, and the shapes do not.

A good deal of what sits around this subject is already established. The buffer this household holds, and how many months of commitments it covers, is known. So is where the line between education and advice runs, and what the four questions are. So are the escalation ladder, the forums, and what a written record is worth. How a loss starts in the first place is the one thing none of that settles.

Here is the honest position most households are in. Nobody can evaluate every claim that reaches them. In one week a household is contacted by more voices, messages and screens than any person can check, and each of them arrives asking for a small amount of belief. The individual claims are built precisely to survive being evaluated, so sorting them one at a time cannot work. Evaluation is the skill everybody assumes is wanted. Evaluation is also the wrong tool, and reaching for it harder is not an improvement.

What can be sorted is the patternThe shape of an arrangement: who contacts whom, which way the money moves, and what is being taken on trust. Much harder to disguise than the story attached to it.. A pattern is the shape of the arrangement rather than its content: who reached whom, in which direction the money moves, what is being taken on faith and who benefits if it is. There are only a few shapes in wide use. The shapes are hard to disguise. Disguising a shape means changing what the arrangement actually does, and an arrangement whose workings change stops working. A household that knows the shapes is not trying to out-think anybody. The job is much smaller and far more reliable.

What patterns account for most financial losses?

Three, mostly, and they are not variations on each other. Each one borrows a different thing, and each one moves money in a different direction, and telling them apart matters because the moment of recognition sits in a different place for each.

The first is impersonationBorrowing credibility that belongs to somebody else, so that an ordinary and correct habit of trust gets pointed at the wrong person.. Somebody presents as an institution, an office or a person the household already has a reason to trust, and the trust that belongs to that institution gets used by whoever borrowed it. The money moves out of the household's account because the household sent it. An impersonation loss is therefore very hard to unwind afterwards.

The second is the advance feeA payment collected now against something larger described as arriving later. The payment is the whole arrangement; the larger thing was never on the other side of it.. Something large is described as waiting for the household, and a small payment is required before it can arrive. The small amount is paid. The large thing does not come, and frequently a second small amount is then required, and then a third. The money moves out in pieces, each of them modest, and the total only becomes visible once the pieces are added up afterwards.

The third is pump and dumpMoving the price of something thinly traded by directing attention at it, then selling into the movement the attention produced.. Attention is directed at something whose price moves easily. The price moves because the attention arrived, not because anything about the thing changed. Whoever assembled the attention sells into the movement, the attention leaves, and the price goes back. Here the money does not move out of the household's account to anybody in particular. The money moves through a purchase that looked, at the time, like an ordinary decision.

As a table, the three separate cleanly on four questions, none of which asks whether the story sounded true.

Not one of those four columns asks whether what was said was true. The table is built that way on purpose. Truth of content cannot be checked in the moment, and truth of content is exactly what is engineered to hold up against checking. The four columns are all about shape, and shape is visible from where the household is standing.

How Impersonation Fraud Uses Financial Trust: what is actually being borrowed?

Impersonation reaches more households than the other two, and the worked instance below belongs to it, so the first pattern deserves proper treatment.

Begin away from money altogether. A man in the uniform of the electricity supply comes to the door with a clipboard and says he needs to read the meter at the back, and the household lets him through. Not because anything was checked. A meter reader reading a meter is an ordinary event. A uniform is what a uniform is for. And treating every person at every door as a possible problem is an unliveable way to run a household. The trust is not a lapse. The trust is what makes ordinary life possible, and it is correct almost every single time.

Impersonation does not attack that trust. Impersonation borrows it. The borrowing takes something genuinely worth trusting, an institution, an office, a role, a familiar person, and points the trust the household already holds at somebody standing in a different place. Nothing about the household's judgement had to fail for that to work. The judgement worked exactly as designed and was aimed one degree off.

Now put money in it. An institution that holds a household's money does sometimes contact it about that money. Contact from an institution is not a suspicious idea but a true one, and a household needs the belief to function. The moment somebody borrows that institution's standing, the true belief becomes the mechanism. There is nothing to unlearn. The belief is correct, and a household that stopped holding it could no longer deal with its own accounts.

What happened on 14 September

At 11:38 on 14 September, Ashok Bhosale, an invented tailor, answered a call. The caller said there was a problem with the recurring deposit. He used Ashok Bhosale's name. He used the last four digits of an account number. He was calm and unhurried and did not push, and everything he said was correct right up until the moment it was not.

At 11:42, Rs 12,000/- left the household. At 11:47, another Rs 6,000/- left. Total Rs 18,000/-, in five minutes.

Naming what was borrowed precisely is the whole lesson. Nothing was seen, so no logo was borrowed. No voice was being imitated, so no voice was borrowed. The caller borrowed the ordinary and entirely reasonable belief that somewhere holding a household's money might contact it about that money. The belief is correct: correct on 13 September, and correct now.

Ashok Bhosale runs a tailoring counter, does his own accounts, and reconciles his own takings every single evening, and it worked on him anyway. Ashok Bhosale knows to the rupee what came in on a Tuesday. He is not distracted, he is not careless, he is not credulous and he is not old. He answered a call, and the call did what it was built to do.

The realisation came at 11:53, eleven minutes after the first transfer. Eleven minutes is fast, and most people take very much longer, for reasons that have nothing to do with intelligence, education or care. The bank was told at 12:20 and the national cyber-crime reporting arrangement at 12:35, both in writing where writing was possible, both with the time recorded. The money did not come back.

What did five minutes cost this household?

Figures tell this better than warnings do, so here is the arithmetic without any softening applied to it.

Before 14 September the buffer stood at Rs 31,320/-, and that figure was not luck. The buffer was built across nine months, an average of Rs 3,480/- a month set aside where Rs 42,770/- leaves in an ordinary month. Rs 3,480/- a month, for nine months, in a household with no spare money lying about. Nine months of that discipline is what the buffer was made of.

Rs 18,000/- of Rs 31,320/- is 57.47 per cent, or 57.5 per cent rounded. The buffer falls to Rs 13,320/-. Against the Rs 42,770/- the household commits each month, cover falls from 0.73 months to 0.31 months, a drop of 0.42 months. And if the household rebuilds at exactly the rate it built at before, Rs 3,480/- a month, replacing Rs 18,000/- takes 5.17 months, so a little over five months of the same discipline just to get back to where the morning of 14 September started.

Put the two rates side by side and the asymmetry is the point: money leaves at the speed of a conversation and returns at the speed of a household. Rs 3,480/- a month going one way, Rs 18,000/- in five minutes going the other. No amount of care at the counter, no evening reconciliation and no discipline about spending operates at that speed. Being told to be more careful with money is therefore no answer at all. This household already was.

Why does impersonation work on people who are careful?

Because two things were done to the situation that have nothing to do with the person in it, and both of them are documented.

Cialdini set out authority as one of the levers that reliably produces complianceDoing what is asked. It is what pressure produces, and it is not the same as being persuaded or agreeing with the reason given., along with urgency and scarcity, reciprocity, liking, consistency and social proof. His finding is the important part and it is easy to miss: these levers do not work by convincing anybody. The levers work by removing the moment in which checking would have come up at all. An authoritative frame does not persuade the listener that the request is right. The frame stops the request from registering as a request at all, and turns it into a step in a process already under way.

Kahneman set out what happens to deliberate thinking under time pressure, and the finding fits exactly on top of the first. Careful reasoning is slow, effortful and easily crowded out. Under pressure, people do not reason worse; they stop reasoning and fall back on the fast pattern-matching that runs constantly underneath. Compress the time and the deliberate step does not degrade, it is simply not there.

Put together, the two do something specific to a decision. An ordinary decision has three steps: something arrives, it is checked somewhere else, and it is acted on or it is not. Authority attacks the second step by making checking feel inappropriate, even rude. Time attacks the same step by removing the room it needs. Neither touches step one or step three. Both go for the middle, and the middle is where all the safety lives.

The mechanism matters more than the theory behind it. If the middle step is what gets removed, then the middle step is the only thing worth defending, and defending it has nothing whatever to do with being more suspicious. A household cannot make itself immune to authority, because responding to authority is how institutions work at all. A household can move the middle step outside the conversation instead, and that is a change of procedure rather than a change of character.

Notice too what this rules out. The levers do not select for people who are gullible, unschooled, elderly or inattentive. The levers operate on the architecture of deciding. Everybody has that architecture and nobody can switch it off. Somebody unusually careful with money has the same middle step and it goes the same way. None of that is an excuse. It is a mechanism.

Advance-Fee Fraud: what exactly is being paid for?

Move to the second pattern, the oldest of the three by a long way. The advance fee has survived every change in technology because it depends on none of them.

The shape is simple. Something large is described as sitting on the other side of a small payment, and paying the small amount releases it. Direction and order make it an advance fee rather than an ordinary transaction, and size has nothing to do with it: money goes from the household first, and the thing it pays for has never been in the room.

Out of finance, the shape is completely familiar. A man at a bus stand says the goods lorry is held at the checkpost and a small amount will release it, after which he brings a full load at half the going rate. Nothing needs to be known about lorries to see the shape: money leaves before anything arrives, and there is no lorry to walk over and look at. Put a document, an office and a company name around the same arrangement and it stops looking like a bus stand, but the shape has not moved.

The financial versions are variations on one sentence. A sum is held and releases on payment of a processing amount. A position is arranged and confirms on a registration amount. Credit is approved in principle and releases once a charge is settled. A prize is held and needs clearing. In every one, the household pays first and nothing at all sits on the far side.

Why is an advance fee always small next to what it promises?

Because the ratio is doing the work, and the ratio is chosen rather than incidental.

Suppose the amount described is Rs 4,00,000/- and the amount asked for now is Rs 4,000/-. One per cent, in other words. Now watch the sum a household naturally runs. The sum is the engineered part. If there were even a one in fifty chance that the Rs 4,00,000/- were real, the arithmetic would say pay. Rs 4,000/- multiplied by fifty is Rs 2,00,000/-, half of what is described, so on that reasoning the small payment is worth making. The arrangement is not asking to be believed. The arrangement is inviting an expected-value sum in which one of the inputs is fabricated, and that sum can be run perfectly and still arrive in the wrong place.

The fee is always small for that reason, and a household that feels it is being sensible about a modest amount is having exactly the reaction the design intends. Being sensible about small amounts is a good habit, and here the habit is pointed at a number that was chosen to receive it.

Then the second half of the design, rarely described. An advance fee does not stop at one payment. A second amount is named, then a third, each with a reason attached: a clearance, a verification, a charge that came up unexpectedly. Take the first illustration through steps of Rs 4,000/-, Rs 6,500/- and Rs 11,000/-, and the total reaches Rs 21,500/-, still only 5.375 per cent of the Rs 4,00,000/- described.

Each step is defended by the ones before it, and that is not a coincidence either. Kahneman, with Tversky, set out how heavily people weigh what they have already put in, and how much harder walking away becomes once money is behind a position. The first payment makes the second easier to justify and the second makes the third easier still, and no single step ever looks like the moment to stop. By construction there is no such moment. The staircase exists to remove it.

One practical consequence changes behaviour, so it is worth stating on its own. The question is never whether the amount asked for is affordable. The amount was chosen to be affordable, and it always is. The question is whether anything exists on the other side that can be looked at or confirmed somewhere the request did not come from, and the answer to that is the same whatever the amount.

Pump and Dump: what actually moves the price?

The third pattern differs in kind, and the difference matters. In the first two, somebody makes contact and asks for something. In this one, frequently nobody makes contact at all. The household arrives on its own, having seen something, and every decision from there is its own.

Here is the shape. Something thinly traded is chosen, meaning something whose price moves a great deal on a small amount of buying. Attention is assembled and pointed at it: a run of posts, a stream of messages, a room where names are given out. The attention brings buyers. The buyers move the price. The thing is thin, so a small amount of buying is enough. The price moving brings more attention, the attention brings more buyers, and for a while the whole thing feeds itself while every claim being made about it is true at the moment it is made. Then whoever assembled the attention sells into what the attention produced, the buying stops, and the price goes back to roughly where it started, or below it.

Notice what did not happen anywhere in that description. Nothing about the underlying thing changed. The thing became neither more valuable nor less valuable. The movement was manufactured, and the movement was the product being sold rather than a consequence of anything. The household did not buy the thing. The household bought the movement, and the movement has no way of continuing once the attention that made it leaves.

Everyday version, and it is a real one. A vegetable seller notices that when four people cluster at his cart, more people stop. A crowd reads as a reason. So he asks two cousins to stand there each morning. The crowd is real and the people stopping are real. Nothing about the vegetables changed. Once the cousins go home, the crowd disperses and the last person who stopped because there was a crowd is standing at an ordinary cart.

Work the arithmetic on the illustration above. The figures make one thing visible that prose does not. The price starts at Rs 40/-, runs up to Rs 96/- as the attention arrives, and falls to Rs 34/- inside four days once it leaves. A household that arrived on day nine, when the price was moving and the claims about it were all currently true, bought 200 units at Rs 90/- for Rs 18,000/-. Fifteen days later those units are worth Rs 6,800/-. The fall is Rs 11,200/-, or 62.2 per cent of what went in.

There is an echo in that number. The 14 September call took Rs 18,000/- out of a household in five minutes. Here is Rs 18,000/- again, leaving over a fortnight, with nobody calling anybody, no impersonation, no lie told at the moment it was told, and no conversation to point at afterwards. The same money leaves by completely different doors, so one pattern of vigilance cannot cover all three.

One more thing separates this pattern from the others and is why it survives: there is nothing false in it to catch. Every statement made while a price is rising is correct. The price is up. The price went up yesterday too. Somebody did make money and the screenshot is genuine. A household applying the ordinary test of is this true gets yes, repeatedly, all the way in. The only checkable question is structural: who assembled the attention, and what are they doing while everybody arrives.

Registered Professional vs Unregistered Impersonator: where does the difference actually sit?

The unregistered impersonator sits across the first two patterns, and households meet one more often than they realise.

Registration is required for certain financial activities, and where a register is looked up is already established. One further fact about the register is both uncomfortable and useful. Put a genuinely registered professional and an unregistered impersonatorSomebody presenting as a registered professional who is not on the register. Everything visible can match; only the register entry does not. side by side and work down the list of everything a household can actually observe, and the two match on every single line.

Both give a name clearly. Both are calm, competent and unhurried. Being calm and competent is neither difficult nor scarce. Both may have an office a household can sit in, and an office costs a month's rent. Both may have something framed on the wall, and framing is not regulated. Both give a detailed explanation that answers the questions put to them. A good explanation is a skill anybody can acquire, and there is every incentive to acquire it.

There is no observable difference, there was never going to be one, and the entire reason a register exists is that the difference cannot be seen from inside the room. None of that is a failure of the household's perception. The invisibility is the definition of the problem the register was built to solve. Seen that way, checking stops feeling like an insult and starts feeling like the only instrument available.

Two things follow from that figure and both change what a household does.

The first is that presentation carries no information in either direction. A shabby office does not indicate a problem and a splendid one does not indicate its absence. Confidence carries none. Detail carries none. A story that fits together perfectly is the ordinary output of somebody who has told it many times, and that is as true of the registered professional as of the person presenting as one.

The second is that checking is not an accusation. A registered professional expects to be looked up and is not injured by it, and somebody put out by being checked has shown a shape rather than a personality. The household does not need that signal anyway. The check answers the question directly. In India the securities-market registers sit with the Securities and Exchange Board of India at sebi.gov.in, insurance intermediaries with the Insurance Regulatory and Development Authority of India at irdai.gov.in, and banking and payments entities with the Reserve Bank of India at rbi.org.in. Each authority's own site states what its register holds and how it is searched.

What do all three patterns have in common?

Now put the three back together, because one thing sits underneath all of them and naming it makes the set usable rather than a list to memorise.

All three run on borrowed trustCredibility taken from an institution, a role, a relationship or a crowd rather than earned by whoever is using it.. Not trust created and not trust earned, but trust that already existed somewhere else and got pointed in a new direction. Impersonation borrows an institution's standing. An advance fee borrows a future, and a future is trust extended forward in time. A pump borrows other people's attention, and attention is trust that a crowd is a reason.

Borrowed trust is genuinely reassuring rather than frightening. None of the three creates trust from nothing, because creating trust is slow, expensive and difficult. All three redirect something that already exists instead. Redirecting is fast and cheap, and that is why the whole area has this shape. The consequence is direct. If the trust is always borrowed, one question finds it every time: where does this trust actually come from, and can that source be reached another way?

One question, then. Not a checklist, not a set of warning signs to memorise. One question with three words at the centre of it: another way. Reach the institution another way. Look at what sits behind the fee another way. Find out who assembled the attention another way. All three patterns depend on the household staying inside the channel they opened, and in all three the same question breaks it.

Why is recognising a pattern easier than recognising a lie?

Because of a selection effect. Being exact about the selection effect is worth the trouble, and it is the most useful idea in the whole subject.

Whatever reached this household reached a great many other households first. The versions that did not work on people were expensive to run and produced nothing, so they were abandoned quietly and without anybody being told. The surviving version is the one that works. The refinement is survivorshipThe versions a household encounters are the ones that already worked. Everything that failed was discarded before it ever arrived, so what does arrive is systematically the convincing kind. operating on the arrangements themselves, and it means the arrangement in front of the household is not a random attempt at deceiving somebody but the current best output of a long process of refinement that was never shown to anybody.

What should a household actually change, and what should it leave alone?

Most writing on the subject ends on be careful. Be careful is not a change anybody can make, and this household had already made it.

Start with what should not change: the level of suspicion. Suspicion is not a dial that turns up without cost, and the cost is high. A household that suspects everybody deals badly with its own bank, delays things that matter, and recognises nothing any better. Nor is suspicion the answer. The 14 September call would have passed a suspicious listener just as well. The content was correct. There was nothing for suspicion to find.

One habit changes, and it is a procedural habit rather than a personal quality. Nothing gets acted on inside a conversation that somebody else started. The habit is the entire change. Not never trust anybody, not check everything, not be more careful. Just: the acting happens somewhere other than where the asking happened.

The habit costs very little. Separating the asking from the acting does not require judging anything correctly, or being right about who is calling, or being suspicious at all. The habit works identically whether the person on the other end is genuine or not, and that is what makes it usable every day rather than only on the days that feel alert. A genuine matter survives being dealt with ten minutes later through a channel the household found for itself. Surviving that is the whole test.

The last row admits no softening. Reporting is worth doing because a record is worth having and a report that exists can be acted on, not because it produces money. Most of the time the money does not come back.