# How I explain Across to a Business Person Across solves the problem of transferring assets between blockchains by disintermediating the Broker role and introducing a third party validator. Anyone with a blockchain account can participate and be a broker or be a validator and offer the best price for their services. ## ELIMINATING THE BROKER There exist many blockchains today that do not have natural ways to transfer information or assets to each other so they outsource their inter-blockchain communication channels. This is analogous to the traditional financial system in which different banks from different countries need to be able to wire transfer value to each other. In order to send money to a bank, the recipient bank must speak the same transfer language as you in order to interpret the transfer order correctly. This is why most banks will support multiple transfer networks like ACH and SWIFT, in order to send transfers to other banks on those networks. Transfer networks are legally enforced: ACH transfers are provided for by an American organization and are therefore enforced by the American government, while SWIFT is enforced by a group of international central banks and its transfers are enforced by an international court system. Blockchains could try to tap into existing networks like ACH and SWIFT to transfer funds between blockchains, but imagine how inefficient this would be: - Send BTC from Bitcoin blockchain to an exchange like Coinbase - Convert BTC to USD on Coinbase and withdraw into bank via ACH - Transfer USD to recipient's bank via wire transfer - Recipient deposits USD into Coinbase via ACH and buys ETH with USD - Recipient withdraws ETH from Coinbase to their Ethereum account You could be a broker of blockchain currency this way, but I will demonstrate to you that there is a way to skip some of these steps. The simplest way to transfer information in general is to go through a "broker" who you trust to transfer the information. This broker is like a messenger or a courier. The flesh-space analog to this role is a mailman who you hand sealed letters to and trust them not to read your letter before delivering it to your intended recipient. ![](https://i.imgur.com/5sxs0YV.jpg) In the real world, companies like Wise and Moneygram provide such brokerage services for a fee. Wise and Moneygram are bound by legal contracts enforced in the sender's domicile to not steal the sender's money. Traditional remittance companies like Moneygram and Wise are bound by legal contracts that they enter into with customers, and customers have the ability to bring cases against these companies in the local jurisdiction court. So, Moneygram would need to register for money transmitter licenses in each country in which they operate. On blockchains, there are no such legal contracts. Or rather, it is harder to enforce a blockchain account because the account is pseudonymous. Legal enforcement would therefore take a long time, and a long settlement experience is a risk that no sender would want to assume. For example, let's imagine that the Bitcoin blockchain somehow had a license to transmit money within America. Bitcoin blockchain still wouldn't be able to guarantee that the account that a sender wants to send money to is also an American account. It could be a French, or Korean account. It would be unscalable for Bitcoin to register for money transmittance in every single possible domicile. As an alternative to legal enorcement, you might require the broker to post a collateral bond and economically secure the message. How would you determine the size of the bond? As a first stab, it should be substantially higher than the value of the information you want to send. Economic security via overcollateralization is one way to achieve such security, with the downside that capital efficiency becomes the bottleneck. If I want to send $100 through the broker then I might require the broker to post a $200 bond. No sane broker would risk $200 to gain $100, so I feel safe sending $100 through this broker. *This of course assumes that if the broker steals my $100, I can submit and win a case at a judicial court by claiming that the broker stole my $100 and I am owed their $200 collateral bond.* *The problem of finding a fair and honest court to plead my case to is another problem called the “Oracle problem” that this post will not cover. Let’s just assume that in most cases, finding a fair Oracle is itself a very difficult problem. UMA is one such system without a central actor that can be corrupted.* Its going to be hard to find brokers who can post substantial enough to bonds to cover realistic transfer volume. Moreover, replacing one Broker with a collection of Brokers is going to be very inconvenient for me as a Sender since I would have to interact with multiple Brokers. Is there a way that I could interact with a single party and that single party would be able to scale up its economic security to cover realistic amounts of volume (e.g. $10,000,000+ of volume)? The solution it seems is that we need to eliminate the Broker and replace it with a system that can divide the total collateral cost amongst many Brokers. ## ACROSS IS A MARKETPLACE FOR BROKERS Across is a market for Brokers to compete to fulfill transfer orders. The system allows anyone to be a Broker and transfer messages. In Across, the Broker will send the recipient the sender’s value minus some transfer fees. A little bit later when the Broker’s transfer is confirmed as a real transfer for the exact specified amount (A.K.A. "validated"), Across will pay back the Broker the amount they sent to the recipient plus the transfer fees. The system is capital efficient because the Broker only risks the amount they are supposed to transfer to the user. Senders only need to interact with a single “Broker”. Once their message order is sent to the Across Broker, any individual Broker can claim the order and fulfill it. This individual broker will later be paid back by Across. You might be wondering, who validates the Broker’s order? Across allows anyone else (e.g. another Broker, the sender, the recipient, etc.) to validate the order was fulfilled correctly. The validator will submit a batch of orders to the Across system that they claim should be refunded. How is the validator incentivized to behave honestly? By posting a large enough bond to cover all volume. So, if a Broker wants to be capital efficient, then they would not validate their own order. Instead, a third party validator is incentivized to validate all broker orders. Across therefore is a marketplace for Brokers and Validators to get paid for different roles. - Brokers only need to focus on fulfilling orders quickly, which means holding funds on destination blockchains and quoting the best prices to senders. - Dataworkers only need to validate orders correctly and quickly. This would be a role taken on by an actor with a lot of computing power. Across' differentiates itself with its degrees of freedom offered to Brokers and Validators. ![](https://i.imgur.com/nRSl8Sr.jpg) The validator’s submission enters into a challenge period of two hours in which any other validator can dispute the submission. The disagreeing validator might have tried to validate all of the orders in the submission and produced a disagreeing result, for example one that claims that not every order in the submission was validly fulfilled by a Broker. Disagreements are brought to a judicial system in which a large set of voters can settle the case. Across uses the UMA oracle as a judicial system. This judicial system naturally has a lot of leverage within the Across system because it secures every single transfer. If the judicial system were to be corrupted, then Brokers would be able to steal orders. This is why a decentralized court system that is economically, not legally secured, is the ideal court of choice for Across. Once the submission passes the challenge period, then Across can issue out all refunds to Brokers. The system gets these funds from a pool of assets called the “liquidity provider pool”. Anyone can participate in the liquidity provider pool by lending their assets to the system. The pool is rewarded fees for each successful transfer. Placing all of the LP funds that are used to refund brokers into a a single pool on a single blockchain, Ethereum, is also why Across is the best experience for passive capital lenders. Across can choose how it wants to pay these LP's for their capital, and currently it utilizes a pricing curve that is similar to a lending fee curve that pays lenders more when more of the pool is utilized. ## HOW ACROSS FAILS - The judicial system settling validator submissions is corrupted - The volume of transfers cannot exceed the liquidity provider pool - There are not enough Brokers in the system - There are not enough Validators in the system