The incentives are. When a market gets harder, the cost of a bad quarter rises for the people who were already running thin, and for some of them the cheapest available source of liquidity is a shipment that is never going to be paid for. That is not a moral change, it is an arithmetic one, and the arithmetic affects everyone trading in the same market whether they intend to be dishonest or not.

If the honest companies are also at risk, how do I tell them apart?

You are looking for the shape of the company rather than the character of the person. A company with real volume, real documentation and a name that appears in more than one place is structurally harder to run away from than one that exists only on your screen. The character question becomes much easier once the structural risk is low.

Does this change how I should pick suppliers?

It changes the weighting. In a good market you can afford to be a little generous on terms because the counterparty can absorb a bad month. In a bad market, the counterparty that is one bad month away from failure is the one that will quietly fail to ship, and that is a loss with no fraudster involved at all.

Bad Times Make Better Liars

The worse the economy gets, the more dishonest counterparties there are, and the more honest ones go under. Both of those happen to the same buyer, in the same months, and they are not the same problem.

Why a bad year changes the arithmetic

Whatever side of the table I am on, the thing I actually want out of a transaction has never changed. On the buying side: the money is safe and the goods genuinely arrive, and what arrives is what was expected. On the selling side: the product goes out and the whole of the payment comes back. That is the whole ambition, and it is not a large one, which makes it remarkable how often it is not met.

What a bad year does is change what is available to everyone at once. A company that was already running thin has fewer options than it had, and the pressure does not arrive as a decision to become dishonest. It arrives as the options getting narrower. When one customer is late, and then another is late, the cheapest source of cash available to some companies is a consignment that is never going to be paid for. Nobody had to change their character for that to become possible. The arithmetic changed, and arithmetic is not a moral category.

It is worth adding that the pressure is not reserved for one side of my table. A company that is short of cash will still take an order, will still send a quotation that looks correct, and will still answer every question I put to it. What a counterparty under strain looks like is not erratic. It is competent. That is what makes it hard to read, because a competent counterparty with a weakening position behaves in almost exactly the same way as a competent counterparty with a strong one, right up until it does not.

That is why I find the sentence "that supplier would never do such a thing" unhelpful in a bad year. It is a statement about character, and character is not the variable under pressure. What is under pressure is the position of the company - its order book, its receivables, how many months of breathing room it had in hand. A company's position does not know or care which customer is writing to it.

What a bad year does to the honest companies too

Here is the half of it that is easy to leave out. A downturn also produces more companies that will go bankrupt, stop paying, or simply stop answering. They will leave goods that nobody collected and money that never arrives, and in most of those cases there is nobody to be angry at. The company meant what it said when it said it. It ran out of a year that had not been counted for.

That group is harder to deal with than the cheats, and I do not enjoy admitting it. A fraudster breaks a pattern you can see - the account changes, the story changes, the patience runs out in a recognisable way. The honest company that fails does not break a pattern. It goes quiet, and then it is gone, and the loss is the same size.

What makes the honest failure hard is that there is rarely any warning inside the relationship itself. Nobody is hiding anything, so there is nothing to notice. What gives way is not their character but their timing, and timing is precisely the thing a distant counterparty cannot see arriving.

More counterparties who will take the goods and not pay - because the option has opened up and it is cheaper than the alternative.
More counterparties who were never dishonest and will fail anyway - because their own customers are doing the same arithmetic from the other side.

Neither of those groups announces itself, and on a given day I have no way to tell them apart from the outside. That is worth saying plainly, because the temptation in a bad market is to believe the second group has been eliminated. It has not been eliminated. It has simply stopped being the interesting case.

So what am I actually looking for

The answer I have settled on is not a better read on intentions, because intentions are not what changed. It is to look for a counterparty whose position is strong enough that honesty is their cheaper option. Not good people - a position where taking the goods and not paying is worse for them than shipping them and being paid.

That is a structural question rather than a character question, and it is a more useful one, because a company's position shows itself in ways a person's goodwill does not. It shows in volume that predates my order, in a name that appears in more than one place rather than only on my own screen, in an operating history long enough that somebody other than me has a record of it. None of that guarantees anything on a particular day, and I am not claiming it does. It changes what a particular day is likely to look like.

What I cannot do is see the accounts. I should be honest that the method has a hard edge: in a bad year an honest company that fails still fails, and no reading of its character was going to stop that. What does change is where I put my weight and how much I have riding on a first transaction with somebody. That is not clever. It is simply the part that is actually mine to decide.

The same logic runs the other way, and it is why a counterparty I have traded with for years is not automatically a safe one in a weak market. History is evidence about the past. What I am trying to understand is the next few months in that company's position, and history is only a rough guide to it.

The thing I take out of a bad year is not that the world is full of liars - though it is fuller than it was. It is that the same pressure is working on both sides of the table at once. A counterparty who is one bad month away from failing can take your order away from you without any dishonesty at all, and that is the version of this problem that has no villain in it.