BitGo wants to build a 'virtual order book' to radically transform crypto's market structure

Mike Belshe runs arguably the sleepiest business in crypto: a custodian that stores digital assets on behalf of its clients. But the former Google and Microsoft engineer and current CEO of Goldman Sachs-backed BitGo has big ambitions to add some sex appeal to his firm in 2019, adding trading elements that could shake up the ecosystem's market structure. In an interview over cappuccinos in midtown Manhattan, Belshe outlined those ambitions to The Block. 

"The Genesis deal is actually the tip of the iceberg," he said, eyes-wide, referring to a partnership announced earlier this year with Genesis Trading, a New York-based market maker, which for the first time allowed BitGo's custody clients to execute a trade through Genesis without having to move their coins out of the firm's cold storage systems. 

That's just step one, according to Belshe. "As a custodian, our job is to keep really good tight security and that means keeping it cold storage," he added.

"And while our competitors are off trying to figure out how to move it off of cold storage quickly so you can get on to the exchange, we believe that model is actually broken."

As such, Belshe's vision is to build a broader platform through which BitGo's clients can trade across multiple exchanges through a virtual order book maintained by BitGo, allowing them to keep their funds in cold storage and tap into the market's broader liquidity. The solution could solve a major problem facing investors in the spine-tingling volatile market. Namely, it reduces the chances of a clients' funds being jeopardized by an exchange hack and eliminates many of the difficulties associated with trading across multiple exchanges, Belshe said

Following is our conversation about those plans, mildly edited for clarity and length.

Chaparro: Explain how this model you are envisioning differs from the current market structure.

Belshe: The model that they have with exchanges is "Gee, well, if I want to get the best price, I need to be hooked up to three or four exchanges anyway." And then if I want to get my capital out, I have to send it not to just one risky exchange, but to three or four different risky exchanges, right? Well, BitGo as a trust company now can take both fiat and crypto, and we can bring that here, and we can bring the exchanges and clients into BitGo and actually swap the asset inside BitGo itself, without ever taking any capital risks ourselves. 

So, we can see exactly how much the capital exchange has with BitGo, we can see exactly how much the client has, and now we can serve as a global virtual order book across all the exchanges that are partnered with BitGo. So, we call this global liquidity, which is really where we're trying to get to. So, it solves at least three big problems. One, it's way capital efficient compared to anything we have on the market. Ask anyone who's trading across these exchanges, and they complain. They have to put their money in all these foreign places. And they don't have full access to any of them. Second thing is you get to keep it in cold storage, literally all the time. And the third thing, is if it actually takes off, it's a bit of a network vector, you'll get the best prices anywhere. It'll have an order book that's bigger than any of the exchanges, because it will be the aggregate of all of them. So, it's a way to introduce market structure into a system that has none.

Chaparro: So, how do you convince Coinbase or Gemini, which both compete with you on custody, to sign up for this?

Belshe: So, those guys may be the last to come, but Bitcoin is already connected to over a hundred exchanges globally. And I think that if you just put a few pipes in place, you can start facilitating, here's my South Korean exchange, here's my European exchange, here's a U.S.-based exchange, all under a single liquidity pool. That means you'll be able to instantly drop the arbitrage across them to zero, because you'll have perfect efficiency. And that's going to result in increased flow to those exchanges, so they'll get a little bit bigger, so no another three sets are going to want to come in, and another three sets, and as it gets bigger, more funds are going to want to come in, because you've got larger access to liquidity, at a better price, that's safer.

Chaparro: But wouldn't this platform introduce a new, much larger, attack vector for nefarious actors?

Belshe: So, this is not a perfect market structure. However, it's markedly improved over where we are today.

Today, the problem is the exchanges keep failing. That is the single point of failure. Let's break it into two. Let's have experts that do custody, security, storage. Let's have experts that do trading, but those two don't have to be the same party, and they're a check and balance against each other. And exchanges can use multiple custodians and the custodians can verify that the funds are there. So, let's get all that to work.

So, if it works, and all these people keep coming in, then eventually you end up with these very large custodians. Which you point out, becomes a new risk. The custodians now know where the funds are. Well, the good news is this fabulous technology we have has an answer for this, and you've heard of this, it's called atomic swaps. So, ultimately what you've done, you've digitized the dollars, you digitize the assets, you separate the trading from the custodian, and now you can use atomic swap to do those trades, and the last thing you need is the protocol around how you do atomic swaps that the custodians agree upon. So that you can have your custodian, she can have her custodian, and yet you can atomic swap between one another.

I think this provides a really solid foundation. And we may end up with having a bunch of competitor custodians who are doing the same model and they also are hooking up to multiple exchanges. But that's not a bad thing! That's a good thing.

And then we just need to figure out how to do that atomic swap between these custodians.

Chaparro: What does that look like for the individual clients — in terms of user interface?

Belshe: So actually you've got the order book, and then you've got bids and asks and volumes that go with it, and then you've got an aggregated order. Yeah, but we won't do the UI format. So, we don't want to be the place where you come to trade. That's not our thing. There's a good half-dozen OMS softwares that exist today. And we would just interface with that. They're going to do a better job. They give people choice. And again, I really do think we should avoid having the custodian have any perceived connectivity to being in the exchange. In this model, the custodian keeps everything tight and secure, you can have your cold storage of segregated, you can have your cold storage that's in omnibus, you can have it accessible to you on any of the exchanges that you want, and then the exchange takes care of the order matching.