Topic: Ricardian Contract and Triple Entry Accounting (Read 5778 times)

He continues on to say this
"We term this triple entry bookkeeping. Although the digitally signed receipt dominates in information terms, in processing terms it falls short. Double entry book keeping fills in the processing gap, and thus the two will work better together than apart. In this sense, our term of triple entry bookkeeping recommends an advance in accounting, rather than a revolution."

The old system of double entry is maintained. So those are two pairs of entries. With the addition of a third pair, the one kept by the issuer. 3 pairs.

Guys,

Thanks for this conversation it’s been helpful as I try to get a solid grasp on this stuff.

I’ve been reading Ian’s papers but I’m left with a few questions.

I think I get the triple signed receipt. It’s signed by Alice, Bob and the server…three parties. Ya?

But the ‘entries’ in triple entry account is giving me some trouble. Ian says that triple entry accounting is…
“pairs of double entries connected by the central list of receipts; three entries for each transaction. Not only is each accounting agent led to keep three entries, the natural roles of a transaction are of three parties, leading to three by three entries.”

At first I thought this meant that the server is using double entry account + receipts…three entries. But why does he say “pairs" of double entries?

Double entry for Alice’s account + double entry for Bob’s account?

And “each accounting agent led to keep three entries”? Alice, Bob and the server are the accounting agents, yes? …Alice is doing double entry accounting for herself and Bob???

I’m missing something here. If someone would tell me what that is, it would be much appreciated!!

Thanks,
Lina

This seems like an ideal way to base a GLBSE style business. Investors would have access to the ledger and have assurance that their dividends are legit and not a Ponzi. Transparency opens up trust with even anonymously based business. Spreadsheet and ledger entries (single or double entry) can be published as a hash on the Bitcoin Network, making them resistant to future editing. Any editing would need to be done by appendage and thus any shenanigans would show up as patterns discoverable on the public Bitcoin Network. If GLBSE were to be reborn someday, this might be a good service to offer.

Here is why:

Double-entry bookkeeping is what made it possible to ferret out wrongdoing, since you could not change something over here, without being forced to change it over there as well, in order to "balance the books." This made it easy to trace back to where the books were originally cooked in the first place. Notice however that you are required to store all of the transactions in order to do this (so you can add them all up and get a balance.)

The point in triple-entry is that the receipt (3-ways with slightly diff versions going to each party and the server) for any transaction IS the transaction itself, and any party cannot dispute a transaction later without producing that server-signed receipt, which also contains his own signature on his original request. Further innovations also make it possible for the last signed receipt to prove ALL transactions (not just the most recent one.)

This means that you no longer have to store all of the transactions individually in order to prove the current balance, as in double-entry. (You certainly may CHOOSE to, for internal accounting reasons or whatever, but you are not FORCED to.) As long as you have the last signed receipt. More on this here: https://github.com/FellowTraveler/Open-Transactions/wiki/Triple-Signed-Receipts

Therefore I think this is an advance over double-entry. The original power of double-entry was the ability to prove balances. Now we can prove balances in a better way with computers than we were able to do in the past, using a single signed receipt, instead of a record of all the transactions, to determine the balance.

Understood loud and clear 

I'm not defending the author, I just think it's a useful tool.

Sounds fine to me. But to call it "Triple Entry Accounting" and to state that it is a challenge to Double Entry Accounting is very misleading. It is simply a stronger way of authentication.

It's not an economics issue, it's more a clever use of cryptography than anything else.

The basic idea with ricardian contracts is that anything that can be traded can be represented with a contract. For example:

Lets say we are using an exchange, and on this exchange I want to trade silver coins.

So I make a contract that says anyone who has it (the contract)can redeem it for silver coins. So to trade silver coins you trade the contract. When you actually want the coins give me the contract and I give you the coins.

Now lets say I want USD for the coins, then I set the ask price at X USD, you give me a contract that says the bearer will be paid X USD, I take it to the issuer and get my money.

These are only some of the basic things that can be represented in a contract, share, bonds, futures are all really just contracts, and you can make them as detailed as you like. And its the contracts that are exchanged on a market.

*Note you must understand basic public key cryptography for the next part to make sense

Now with ricardian contracts it's just an electronic form of a contract, a text document or textfile. And the contract is signed(using their cryptographic key pair) by the issuer(whose public key is available to verify), it's signed by the market exchange it's traded on and finally it is checksumed, so that it can't be changed without it being noticed.

And the really useful thing is that the checksum is the contracts id on the exchange, instead of using a ticker symbol like USD or GLD, instead it's something like 24ea1163ea6c9f5dae77de8c49ee7c03

And it's up to the buyer to properly understand these contracts and to know whether the issuer is trustworthy so they dont get burned or cheated.

My understanding of Triple entry accounting is again it's a cool use of cryptography.
The basic idea is that the receipt becomes the transaction. The buyer pays the seller, the seller makes a receipt showing payment,it is signed by the seller, the buyer signs the receipt showing a trade, then the exchange signs the receipt and the receipts id again is a checksum of the signed receipt(meaning that it cannot be changed by any party). This process is now proof that the transaction happened, it is verifiable and impossible to forge. Or something like that.


No, the argument is that the receipt IS the transaction, or at least it is the full record of it, and when talking about an exchange the transaction is the swapping of contracts(electronically). To actually get the goods you must redeem those contracts (I think this is the clearing process of stock trade).

I minored in Economics and am halfway to my masters in accounting.

The author has correctly identified that double entry accounting makes errors of a certain type more noticeable, but seems to fixate on this function. One of its main goals is to correctly match revenues to expenses in the same accounting period, which is arguably more useful than any other function of the system.

The "triple entry" he has identified is actually a double entry split across two liability sub accounts. Here's the entry:

Debit: Alice's Liability Account...... 100
Credit:       Bob's Liability Account........ 100

If he is making the argument that the receipt is the asset, then fine. But the asset doesn't actually go anywhere; there is no transaction needed for it. It has simply changed title from Alice to Bob but remains on the issuer's (bank's?) books as an asset. The transaction (receipt) still plays the same role as a traditional paper receipt; it is back-up for the transaction. I suppose you can create asset sub accounts entitled "Assets held for Alice" and "Assets held for Bob;" but this would be redundant. Then you can do another entry with this transaction:

Debit: Assets held for Bob...... 100
Credit:       Assets held by Alice...... 100

But that's still a double entry (the debit and the credit) and not a triple entry. I don't think the author know's what an "entry" is and has mistaken it to mean the number of places a transaction is recorded.

Just wondering if our esteemed economics colleagues have heard of these theories or are willing to investigate.
I have looked into in briefly but many of the hard-core terminology was above my geek level.