Payment systems

Kipochi the startup I helped found 3 years ago is in the news again. While the whole story ended up pretty painful for me I did learn a lot from the experience. It's been a while now so I think it's time to tell the story from the inside.

First some background

I started Kipochi in Feb 2013 together with a consortium of Danish guys with experience in telcom in Africa. I brought technology from a previous startup Economi.co that I shut down and they were in charge of money and sales.

Our original goal was to partner with telcos in Africa and Asia to create an alternative to local operator owned mobile money operators (MMO) like MPesa.

This was to be based on Bitcoin and allow interoperability with any vendor. Our goal from a tech point of view offer a consumer wallet based on USSD and mobile web. This would be branded by local operators.

We would also provide a solution for operators to use their existing agent network to buy and sell bitcoin.

Updated 4/21/2016: Updating this as this is also now affecting the Gift Card industry. Added crypto currencies in general. Moved hosted crypto currency wallets like Coinbase and UpHold up to a 3 since the industry has changed a lot since 2012 (No more BitInstant or Coinbase). Removed Liberty Reserve for obvious reasons and modernized it a bit.

About 15 years ago after a bunch of E-Gold exchangers lost money from fraudulent purchasers performing chargebacks Start e-gold app developer JP May published his May Scale of Money Hardness (only on wayback machine now).

The idea is that the easier it is for a payment to be reversed, the softer it is. I’m trying to update it for 2012 (now 2016) so the next generation of financial startups don’t get burnt more than they already have been and learn the lessons learnt by the e-gold community.

Original May Scale

Hardness	Item
1	Street cash, US dollars
(Hard)
2	Street cash, euro currencies, japan
3	e-gold
4	Street cash, other regions
5	Interbank transfers of various sorts (wires/ACH etc), bank checks
6	personal checks
7	Consumer-level electronic account transfers (eg bPay)
8	Business-account-level retail transfer systems
(Soft)
9	Paypal and similar ‘new money’ entities, beenz
10	Credit cards
(Ridiculously soft)

Updated May Scale

Hardness	Item
1	Street cash, BitCoin and other intrinsic crypto currencies like Ethereum and DASH, Gold/Silver Coins
(Hard)
2	Western Union and other money transmitters
3	Account based electronic currencies fire walled away from banking system (e.g. UpHold, CoinBase, BitGold)
4	International wires
5	bank checks
6	ACH, personal checks
7	Consumer-level electronic account transfers (eg Dwolla, AlerPay)
8	Business-account-level retail transfer systems, credit cards (brick and mortar)
(Soft)
9	Credit cards (via internet or phone)
10	PayPal
(Ridiculously soft)

updated 3/8/2012 Based on comments I moved things around a bit. Ordinary bank wires are more less reversible than ACH’s. The scale shouldn’t cover default risk only reversal risk so I’ve moved all street cash to 1. and added gold/silver coins to it. I’ve added electronic currencies like Liberty Reserve that are fire walled away from the banking system to 3. Any account based system that interfaces directly with the banking system will be in 7, so AlertPay belongs with Dwolla as their banking connections can also put undue pressure on them. I still consider Personal checks risky. At least until it clears so I’m leaving that at 6. Also see The BitCoin Wiki’s Payment Method page for alternative rating

Why do I place Dwolla above PayPal? Mainly because they exclusively use the ACH system and not the credit card system, which makes it harder for people to perform chargebacks.

Why is PayPal below Credit Cards then? While it is a blended system of ACH and CC, their risk management procedures makes it even more risky to merchants than using a straight credit card processor. Just by virtue of being successful or having one or two chargebacks they will freeze your accounts.

What can we learn from the May Scale?

Purchasers benefit by having softer money as it reduces risk for them and merchants benefit by having harder money.

If you are selling game credits, subscriptions to a web service without any significant cost to you the benefits and ease of accepting soft money is fine.

It is slightly trickier for merchants sending physical goods to users. The merchant does have a risk, but the shipping infrastructure does provide some insurance and documentation that partly alleviates it.

If you are selling financial instruments, real estate, cars and other high value items you should not accept anything higher than 5 on the May scale. As a matter of fact due to anti money laundering laws since the original May scale was written you probably shouldn’t accept anything less that 5 either.

If you sell BitCoin you are in a little bit of a different situation similar to e-gold exchanges of yore, due to the growing hostility to it from financial institutions and governments. Depending on how high your spread or transaction fee is you could accept the risk of accepting bank transfers, in particular if you limit the size and frequency of transactions with customers until you feel you can trust them. But I would probably stick to 3 or below.

One risk potential risk with Number 3 is that you might accept a Western Union transfer from a party who is on the terrorist watch list, in which case you could get associated with them. This is a particular risk if you are not registered with the FSA (in the US) or similar elsewhere and did not perform some level of Know Your Customer.

Why soft money?

Most of the soft money is based on book entry systems with central authorities. One of the benefits of this is that in case of fraud it is possible to contact this central authority and reverse a transaction.

BitCoin is also a book entry system, but there is no central authority so there is no way of reversing a charge.

So if you as a consumer buy some coffee with BitCoin and the merchant doesn’t send it you there is no real recourse unless you know who the merchant is and are able to take the merchant to small claims court.

Because of this there is a trend within the BitCoin community of having centralized BitCoin denominated book entry systems that temporarily keep the funds in a reversible place. 4/21/2016: Do these even exist anymore? There are also BitCoin escrow agents. These are potentially good solutions, of course you also need to trust the operators of these or they may just run away with your BitCoin. There were probably more fraudulent than trusty escrow services in the e-gold days. I would expect the same with BitCoin.

There is also a risk to the merchant in BitCoin. See the recent Linode Bitcoin fiasco. If they were using softer money they could have called the central authority (PayPal etc), freeze their account and have dodgy transactions reversed.

Anyway none of this is easy. If you are trying to do something with BitCoin or other kinds of alternative financial services you need to think about it for your business.

The debate continues. Please read the first part of my response OpenTransact the Payment Standard where everything is out of scope first.

Manu wrote a new response which I will respond to in a separate blog post. First let me finish responding to the original

Generally this post will again reflect the differences in approaches. OpenTransact is a single layer simple pragmatic standard for performing payments nothing else. PaySwarm is a fully featured idealistic multi layered approach where you must buy into a whole different way running your business.

A Facebook friend suggested that OpenTransact vs PaySwarm is like Libertarianism vs Socialism. I don’t quite buy that in practice as I know that PaySwarm is not about forcing anyone to do anything.

However the basic PaySwarm philosophy of wanting to design a whole world view is very similar to central planning or large standards bodies like ANSI, IEEE etc. OpenTransact follows the market based approach that the internet was based on of small standards that do one thing well.

The W3C Web Payments Community Group was launched in August 2011 because Digital Bazaar had created their PaySwarm spec. I have been working with several others since 2009 on bootstrapping an open grassroots created standard OpenTransact with pretty much the same purpose as PaySwarm. I immediately joined the W3C group to see if we could work together towards our obvious common goal.

Different philosophies

So while we both shared this common goal, the fundamental philosophies of PaySwarm and OpenTransact could not be more different.

PaySwarm attempts to solve every single problem up front and thus creates a standard that is very smart in many ways but also very complex. It’s background is I understand in a P2P media market place called Bitmunk where licenses, distribution contacts and other media DRM issues are considered important. Manu Sporny of Digital Bazaar has also been a chair of the RDFa working group so PaySwarm comes with a lot of linked data luggage as well.

OpenTransact comes from the philosophy that we don’t solve a problem until the problem exists and several people have real experiences solving it. I have also been very active in the OAuth community and believe there are many things both good and bad that can be learnt from that process and how developers took to it. OpenTransact also follows the tradition of early web standards of having most parameters as optional.

Currency is one of the most popular buzzwords right now and there are lots of different definitions. I’ll try to unify them in some way and talk about some of the issues involved. Over the next couple of posts I’ll try to analyze what currency is.

If you ask most people on the street, currency is what they have in their bank account and in their wallet.

Gamers will tell you that currency can also mean numbers on a screen earned and spent within a game.

Silicon valley hipsters will also try to say that Virtual Currency is the latest monetization strategy out there, often without realizing what it really means.

Community activists also like to remind us of all the successful community currencies that have sprouted up in the past few years.

So my first attempt at a definition is:

A currency is a fungible asset that can be transferred from one person to the other.

Now under that definition we may also need to include stocks, options other securities. As they are generally transferrable and fungible. Most people wouldn’t consider them currency, but they fit the definition perfectly. I’m personally quite happy to consider them currencies.

Virtual Currencies?

What makes a currency virtual or not? It’s not wether it has any real value as World of Warcraft Gold clearly has value. I’d say it depends on the backing of it. So an attempt at a definition:

A virtual currency is a currency backed by the promise of it’s issuer.

Closed loop currencies

A closed loop currency is a currency only meant to be spent with the issuer. Good examples are Starbucks Cards, but most game currencies are also closed loop currencies as you can only use them within the games.

What about Whuffie, Page Rank and other reputation currencies?

That is a good question. These are currencies that are objectively awarded and taken away based on your standing/actions in a community. Pagerank is often also described as a currency. Most of these break my definition above as they aren’t generally speaking transferable.

Whether they are really fungible is also a good question. A Google PageRank of 8 would have to be worth 8 times a PageRank of 1. But clearly that is not the case.

In my next post I explore a few different flavors of what many of us think is just one currency, the all mighty US Dollar.