Consider a river moving through a town and suddenly the
source of the current disappears. The first user of the water, closest to the
source receives the water flow and those downstream do not. This is not a case of first come first served;
this is a case of planned first use of resources.
Money in a gross real time payment system acts in the same
manner with a receiving financial institution put in place of the first user of
the resource. The difference is there are no natural laws governing payment
systems; payment systems are entirely a human creation, and function absolutely
according to controls we place on them. Payment systems can run uphill to butcher the
metaphor precisely.
Once in a bank’s domain, payment instructions, law, and
policies govern the distribution of funds. It is possible to do the same thing
with small values and with communities having trouble selling goods and
services because of the appearance of poverty or the distance from potential buyers.
If buyer knows a push of funds to a seller always results in transfer of
whatever purchased, then trust between buyer and seller need not exist; just
trust in the payment system.
However, a payment instruction, cannot give details for
distribution of payment. If a payee requires reimbursement for shipping, the
payer pays in advance (typically). The question is can we engineer the payment
system to hold the funds partially and let the flow to payees go in stages, and
if we do it, what happens if the transaction becomes interrupted. Is this a
desirable use of the control functions of a payment hub at all? Diagram 15
shows the different paths.
Certainly the concept illustrated in diagram is possible and
the refusal prompts an instruction from the receiving financial institution to
replace the money (less the transportation costs to the payer). However, the
service is unwieldy, and funds stuck in limbo, essentially without ownership is
a logical nightmare (what happens if the receiving financial institution goes
bankrupt or the goods become destroyed in transit for example). The concept stands in stark opposition to a
fundamental principle of payment system architecture; specifically - speed
good, delay bad.
That is not to say that a community or a single seller may
purchase a type of escrow agent, which receives the payment and then acts on
the instruction of the payer. Effectively it is a new type of payment hub, one specifically engineered to delay fund distribution according to payer instructions; it creates
trust where none exists otherwise. The
question is the price for such a service greater than that of the aggregate
fees related to current card payment techniques? The answer may well rest in
the liability of funds for failed transactions.
Next Blog: The
gyre of behavior and evasion detection filters.
No comments:
Post a Comment