A very well known expert wrote to me privately the following critique of the pyramid model that I developed:
"I don't really agree with the pyramid scheme analogy (though no analogies are perfect). In a pyramid scheme, the risk gets greater in every round and is passed further and further from the originator. Here, the risk is still with the bank (though limited liability passes it back to the depositor). Also, the money exists (deposits always equals loans, ignoring investments and securitisations) it is just not necessarily liquid. The Austrians argue that you do not need regulation at all. If you simply applied normal procedure everything would be okay - ie if the bank cannot meet all demands then it is deemed to have folded. As such, they argue, the market will bring about higher reserve ratios (as happened in the past before lenders of last resort)."
I responded:
1. It is a financial pyramid for real - it is not an analogy
As far as I am concerned a pyramid scheme is not really an analogy. If you lend with loan to deposit ratio above 100% IT IS a pyramid scheme (by definition of a pyramid scheme). By very nature of exponential growth lending with loan to deposit ratio above 100% leads to collapse regardless of any other risks. (Here I applied analysis based on Cobham Thesis from computational complexity.) That is why I separated the risk directly stemming from pyramid growth from any other risks in finance.
The risk with pyramid growth can also be looked at in two ways:
- the risk of collapse of the system with a lending with loan to deposit ratio above 100% is practically 100% (it is deterministic)
- the risk when, at which cycle, of a pyramid growth it is going to collapse is far more complex: it depends on others risks, market confidence, uncertainty, etc. but, considering the point above, with every round the risk of collapse gets greater as the collapse becomes closer (this answers your first issue with my approach).
I would suggest the following analogy: if you build a bridge on a major road that can bear maximum 5T load (and do not put any warning sign), it is only a matter of time when a car weighing more than 5T will crash a bridge. But it is really difficult to say when: a traffic pattern on this particular road would have to be studied etc. And this can be quite complex: so whilst, realistically, a collapse of the bridge is deterministic when it happens is not.
This is precisely why pyramid schemes are illegal in law and lending with loan to deposit ratio, as it is building a financial pyramid, is illegal.
2. Risk growth and propagation in a pyramid
With respect to the argument of your objection to my model (as it is not really an analogy). You wrote: "In a pyramid scheme, the risk gets greater in every round and is passed further and further from the originator. Here, the risk is still with the bank (though limited liability passes it back to the depositor)." As already mentioned above, in my model, risk also gets greater in every round. You consider bank as an originator. In my model I consider individual bankers who made decisions then collected their smaller or larger bonuses for their "performance" as originators (plus regulators and some politicians). As I hope you clearly see in my model, risk, if any at all, is also is passed further and further away from originators (i.e. individual bankers). A bank as such is its shareholders, customers and ultimately taxpayers in case of government rescue. They all are a pyramid scheme customers duped to this rather nasty business. Bankers, regulators and some politicians are pyramid purveyors and I do not see any risk associated with them at all. Moreover, unlike Albanian gangsters who could not force already cheated customers to cough up more bucks, our pyramid purveyors do it by forcing taxpayers to subsidise failing pyramids. So what is the risk of pyramid originators? None. Indeed it looks like a perfect pyramid. It is the same as if Albanian gangsters created a pyramid with customers money only. (It is not an emotive language: this is just reality.)
In fact when you consider pyramid customers: the early customers that participated (e.g. took a loan 5 years ago, bought a property and a year ago sold it and cashed it out) are winners. This also exactly what happened in Albania. These early gains, or gains for some time on the paper, also work as a way of luring customers into a pyramid. If you early enough and do not exit too late, cash it out, you are a winner. In the City and in Albania. So a pyramid has winners: but this is extremely disproportionate to overall loss. (This actually also answers your concern: "the risk gets greater in every round and is passed further and further from the originator". In fact it contributes to validate the pyramid model of this crisis showing its practically recursive character, i.e. every pyramid node is an originating node of a (sub)pyramid.)
The way of passing the risk is still the subject of my studies. But by passing risk around you may stabilise pyramids but as they keep on growing, you will still face inevitable end: a collapse.
3. Money v cash (M0)
I am not clear what you call money. I generally refer to cash (M0): i.e. hard cash and my approach to loan to deposit ratio is based on cash reserve growth/depletion analysis. In that context: with loan to deposit ratio 100% exactly, value of loans equals value of deposit, but banks generate no cash reserves in the process. Technically it is a borderline pyramid; i.e. its growth is linear (a sum over a number of deposit/loan cycles) not exponential that makes the system so lethal. Having written that I would still not recommend it.
The
"Exercise/example - how does it work? shows how
"not necessarily liquid" money becomes toxic waste (and that this is inevitable) when loan to deposit ratio is above 100%.
4. Regulations
I do not feel at all about regulations: I do not consider myself as financial markets expert so I am quite agnostic on that. My model only infers that if you lend with loan to deposit ratio above 100% for long enough the system will always collapse. (And generally that "long enough" will not be that long.) The financial risk management techniques, spreading and minimising the risk, may ONLY delay this deterministic fact, but if lending with loan to deposit ratio above 100% continues the pyramid gets bigger (i.e. ratio of banks balance sheets to cash on the market) and when a collapse comes it is simply more spectacular (such as the one we observe now as compared to Albanian pyramid collapse). That's why it is important not to over-regulate as, in extreme, it may be even lethally counter-productive. (My mathematician's instinct tells me that the rules, whether tight or "liberal", should be simple, consistent, i.e. free of contradictions and complete, e.g. no loopholes. Human instinct tells me that whenever money is not tightly controlled it gets stolen - look at the MP's expenses now, but tight control hampers human initiative so it is economically counter-productive. The answer is in the balance, which I do not know.)
Having considered your comments, which I greatly appreciate as they "stress-test" my work, I do not see how they undermine my modelling of the financial crisis. I think your concerns are dealt with within the model.