Why do crypto OTC desks fail faster than regular markets?

4 min readAug 24, 2020

The genesis of this article was a twitter thread responding to a question from a friend had after trying to do some business. At the time, He was trying to set up a contract for a 10MM reoccurring tranche for a three month period and found that all the names and brokers he had successfully done business with had closed shop or were no longer with a given firm. Coming from a commodities background, he was puzzled how something that traded like cash and commodities could just shut down.

As a two-way trader in this space, I don’t believe that it’s from lack of demand.

From where I sit, it breaks down into six primary business points.

1.) Non-unique services or deals — Many of the desks that have failed or pivoted away thought they could “build it, and they will come.” Meaning that the money would flock to them because they offered a service. However, it is a service that everyone else provided. Nothing made them special. No unique order flow, no great connection, or technology advantages.

Without the customers they expected to show up, you have no counterparties. For most deals, you need two or more sides (capturing the spreads in between), even if you can internalize the majority of the trade. As a customer, often time I would put out quotes or tranches to an eager desk with start and end dates, and they couldn’t fill the order. (i.e., lacking the network) even if I was willing to give them more edge to the trade to do so.

2.) Poor customer service and deal-making ability — In regular markets, if a desk knows you can trade size or have a lean one way, they are on the phone or in a chat with you regularly, shooting the shit, telling you the deals they are working on, all to get you interested in matching.

This doesn’t happen in crypto. OTC desks, mainly exchange based otc desks, try to internalize the entire trade. Quoting round prices that match their risk, rather than trying to match with another party and reduce the carry risk. They want to maximize profitability for themselves, only thinking of a single trade rather than the next 20 trades that could happen. (i.e., this is how most big fx desk work, even if it is by RFQ).

3.) Relying on machinery where human interaction is needed — Over 2019, a few desks have figured out that RFQ or request for quotes was a faster way to do business rather than doing chat for everything. It allows you to submit an order electronically and get quotes back from interested parties. However, the people using these systems are the same 8–10 desks you find everywhere else, basically trading with themselves. These same desks are usually trying to lay off the carry risk of a trade they just made rather than matching themselves (i.e., hot potato).

The rise of RFQ systems has lead to a perceived efficiency that has lead to the fall of the relationship manager. A few of the larger desks have laid off or paid out most of their relationship guys. Based on the most successful desks in crypto today, I think this is a mistake. Customers need hand-holding, deal-making, and quite frankly, a reminder you exist.

As an example, when a customer of Fortress wants to buy 50M of something, he’s not spending his time working order and RFQs to piece it together. He’s talking to his account/relationship manager and saying he wants to buy 50M at so many bps, the desk goes off and fills his order.

4.) AML standards tighten and a surprise to clients — despite crypto twitter’s insistence, all desks capable of doing size comply with AML, money transmitter laws, and other local equivalents. This means any customers they on-board go through a fairly intense KYC process. If you’re a business, all the articles of incorporation must be presented, and any members holding over 10% controlling must also comply with by providing KYC information.

Combine this with proving the source of coin, proving ownership and blockchain analytics before the trade is made (i.e., did it come from the darknet, scam, gambling, etc.). This impacts both the speed of the deal and people’s ability to potentially execute (either for privacy reasons or failing the above process). This makes many people go straight to exchange instead as it’s the easier path. We see evidence of this whenever ICO coins become unlocked

5.) Non-cyclical demand — some of the biggest client in OTC, such as ICO foundations or whales, do not generate additional income beyond their immediate crypto holdings. This means once they’ve sold, the most substantial order flow for a given product tends to trickle down, spreads compress, and there’s not enough meat on the bones for every desk to have a bite.

An example of this was when the EOS foundation took their funding raise in Ethereum and was selling to desks by the truckloads — every day, seven days a week for months. Unrelated desks had to talk to each other to make sure they didn’t step over each other on the market. Once they were done, the OTC supply and demand dropped dramatically for months.

6.) over pricing based on self-perception — this follows point #1 but worth talking about. When you have 20 desks, offering identical services. Customers will eventually cycle through all of them to find the relationship, services, and pricing. The non-competitive ones had to close shop.

The rush to OTC alpha gave way to many parties w/ 0 differentials at the trough. So supply and demand took over, spreads compressed & those who couldn’t fly, will die. The desks that left standing are those with deep networks, exclusive deal flow, or great relationship/bus devs.