Financial Cryptography

Money costs: a dollar, a penny, a system, an experience

John Kyle found some US coin and note costs. For the dollar:

Rep. Jack Metcalf then began questioning Allison about these issues in more detail. Allison explained that when the United States Mint produces a dollar coin, it spends 8 cents on production costs and issues the coin into circulation at face value (100 cents), depositing the coin in the Federal Reserve for 100 cents. The 92 cents difference is seignorage, essentially profit. In the case of a dollar bill, on the other hand, the cost of producing the bill is 4 cents, and the Federal Reserve issues the bill into circulation at face value, investing the 96 cents difference in U.S. Government bonds. The interest the Federal Reserve receives goes to Federal Reserve expenses (about $2 billion), retained earnings (a few hundred million), dividends to member banks (another few hundred million); the rest goes back to the Treasury. Metcalf noted, "It seems like an arcane system that could have been invented only by somebody who was mentally deranged."

And for the penny:

The Mint estimates it will cost 1.23 cents per penny and 5.73 cents per nickel this fiscal year, which ends Sept. 30. The cost of producing a penny has risen 27% in the last year, while nickel manufacturing costs have risen19%. ... But consumers should not hoard coins or melt down the change in their kids' piggy banks, says Michael Helmar, an economist and metals analyst at Moody's Economy.com. He says the process of melting the coins, separating out the metals, then selling would be costly and time-consuming.

"If they were made out of gold, sure," he says. But "there are just too many other costs."

OK! It's nice to know it is ok to hoard and smelt the gold coins. For those interested in the cost of payment systems, here's a list of the things that the business plan needs to consider. All standard fodder for the FCer. Over at Oyster, they seemed not to have paid attention to the basic recipe:

Back in 2005, Transport for London (TfL) announced it had shortlisted seven potential suppliers to transform Oyster from a ticketing system into a means of paying for goods such as coffee and newspapers. Trials were scheduled to start before the end of the year but didn't materialise. And, at the end of April, TfL announced that none of the shortlisted suppliers had been able to meet their criteria and the rollout had been put on hold for the time being. ... So what went wrong? It seems that the technology behind the scheme was not at fault. Dave Birch, director of consultancy Consult Hyperion and organiser of the Digital Money Forum, told silicon.com: "What's become clear is that it's more complicated to sort out commercial arrangements than to sort out the technical arrangements [with e-money]."

It appears that issues with the payment processing side of the project - division of revenues and payment processing costs, for example - were the main reason the e-money scheme was hobbled before it left the starting gates. The question of who would pay for the cost of deploying the necessary infrastructure was a sticking point. For example, without financial support from the banks, retailers were unlikely to agree to cover the equipment costs themselves.

That would be numbers 1, 2, 4, 5, 6, in Win Derman's list. I often characterise the question this way: it costs order of a million to build a software payment system, and order of 100 million to build a hardware token payment system. The difference is the tokens, and who you get to pay for them...

Which leaves us wondering why these lessons aren't learnt? Win Derman says "In looking back over the last 30 years, there’s no question the industry has witnessed (and still is witnessing) tremendous change. But despite the disruption, the same fundamentals still apply. There are 13 questions I’ve learned to ask about any new payment technology to evaluate its potential for success in this space."

Maybe experience is priceless?

---

In looking back over the last 30 years, there’s no question the industry has witnessed (and still is witnessing) tremendous change. But despite the disruption, the same fundamentals still apply. There are 13 questions I’ve learned to ask about any new payment technology to evaluate its potential for success in this space.

1. Are you solving a problem that needs changes at the point of transaction? Then build a 5-7 year implementation schedule because of terminal replacement cycle.

2. Are you solving a problem that has limited geographic applicability (for example, the 1980s chip card in France because of the poor telephone network versus the magnetic stripe in the zero floor limit environment of the United States)? Either you will be tied up in politics or limited in scope to a subset of the card world.

3. Timing is everything. We worked for 30 years to grow the Visa credit card business into a trillion dollar business. The Visa debit card grew to that level in less than10 years because the network infrastructure already existed, the name Visa was already well known, and the political battle between the credit card and debit card groups within the banks was over.

4. If you are selling a proprietary product or service and hope to extract a royalty on every transaction, you will be trying to push a huge rock up a large hill in an environment where even competitors need to share technology because banks and merchants want standardized services - not proprietary implementations.

5. If your product or service needs approval from more than one industry group (such as airlines and banks), plan on a very long and complicated negotiation to achieve consensus.

6. If you expect to change consumer behavior to implement your product or service, remember issuers will be very reluctant to risk making significant changes that could damage a $4-5 trillion business.

7. If your product or service is aimed at fraud reduction, remember that eliminating 100% of fraud - which is highly unlikely - only changes the issuers' bottom line by a fraction of 1%. That doesn't leave much room if implementation or operational costs are high.

8. Technology by itself without a surrounding set of business rules, prices, and procedures that make economic sense will not fly.

9. Any product or service that does not show significant results within 12 to 18 months will take a long time to get anyone’s attention.

10. After you develop a timeline to get your product or service sold, double it.

11. Who is your champion in the target organization? (We tried to have the credit card departments where Visa had its contacts introduce the Visa debit card to the banks. We made very limited headway for almost 30 years until credit card people became senior managers in the banks. Then implementation went very quickly.)

12. If your plan is to introduce your product or service in competition with the existing card products, remember that you either must spend a great deal of money to get a piece of the business (like Discover did) or you will be vulnerable to the big guys overwhelming you during your buildup phase as they easily could have done with PayPal - if they had focused on that market niche earlier.

13. If your target mark is the bankcard business, the good news is that it is a $4 trillion dollar business and growing. The bad new is that it is a $4 trillion dollar business and growing. There is a potential for a big payoff but it is hard to turn the business in a new direction - kind of like trying to turn the Exxon Valdes in Prudeau Bay.

Posted by iang at May 12, 2006 06:03 PM | TrackBack