There is an argument against interest that does not require you to believe anything at all. No theology, no ethics, no view about usury. Only arithmetic.
Debt at compound interest grows exponentially. The real economy, made of human labour and physical resources, grows linearly, and not even reliably.
Run two such curves against each other for long enough and there is exactly one outcome. The exponential one wins. Not might. Must.
Which means a financial system built on guaranteed compounding returns is not risky in the ordinary sense of possibly going wrong. It is a structure that arrives at a fixed destination, and the only open question is the date.
The standard reply is that growth compounds too, so the curves can stay matched.
Over short runs that is true, and it is how the system has kept moving. Over long ones it fails, because compound interest has no ceiling and production does. A rate of return does not answer to any of that. It is a number in a contract.
Global debt has passed three hundred and fifteen trillion dollars, about three hundred and thirty per cent of world output. American federal debt alone is past thirty-six trillion, with interest now exceeding a trillion a year, more than the defence budget. The state borrows to service what it has already borrowed.
2008 is usually filed as a crisis that was handled. It was not handled. It was postponed, and the postponement is the interesting part.
Central banks created something over twenty-five trillion dollars from nothing and put it into the financial system.
So the response to a crisis produced by excessive financialisation was a very large quantity of additional financialisation. The disease was administered as the treatment, and then the recovery was announced by measuring the asset prices.
The Islamic objection to riba is usually presented as a moral rule, and a slightly awkward one, a prohibition to be worked around with clever structuring. Read it structurally and it is not a rule about fairness. It is a refusal to let a claim on wealth grow by itself, detached from any real activity and any real risk. Remove the guarantee, insist that a return must be attached to something that actually happened, and the exponential curve loses its engine.
Whether that is workable at civilisational scale is a serious question and I am not going to pretend the answer is obvious. Muslim-majority economies have adopted the extractive model wholesale, which tells you the framework is not currently operative anywhere, and a framework nobody runs is a hypothesis rather than an alternative.
But the arithmetic does not care whether anyone has built the alternative. Exponential against linear resolves one way. The only decision available is whether that resolution is chosen or arrives on its own schedule, and systems that wait are not usually given the choice.