Vitalik predicts the rise of “non-custodial centralized exchanges” in a recent blog post. Is he right? Will traditional CEXs even be necessary in 2023?

2023 is the year of woke crypto. Exchange type isn’t binary anymore; it exists on a spectrum. We don’t just have CEXs and DEXs. We have HEXs (hybrid exchanges), FEXs (fully encrypted exchanges), and T-REXs (a type of dinosaur).

Vitalik, in a recent blog article about safe CEX, talks about “proof of solvency” and how an exchange can demonstrate reserves in excess of customer deposits. He goes on to say that in the long-term, even better than proof of solvency is a theoretical new type of CEX that is cryptographically constrained from ever stealing user funds. Basically, a “non-custodial CEX.”

In recent months with CEX after CEX change burning up in flames, we are seeing more than the dissolution of a few businesses. We are witnessing the death of the centralized exchange model altogether. Trusted CEXs are an exclusive club, and the club ain’t taking new members. Of the remaining giants like Binance, Coinbase, and Kraken, we may still see some disappear but we’ll never see a new CEX acquire enough brand equity to turn a profit and survive.

But is this a bad thing? Historically, retail crypto investors have always needed CEXs for (1) fiat onramping, (2) custody, and (3) timely order execution (e.g., “buy/sell me $100 in bitcoin if the price goes above $20k”). CEXs clearly fill a necessary role in the ecosystem, but for as long as there has been crypto, there has also been an exchange willing to steal it (going all the way back to Mt. Gox).

Only extremely recently did it become possible to actually have decentralized or non-custodial alternatives that are just as simple/convenient as a traditional CEX. So, what makes the future different than the past?

Contrary to the 2017-18 cycle, which saw an ICO bubble that left the industry with little lasting innovation, this latest cycle saw the creation of a host of web2 startups solving web3 problems. Startups likeMoonPay that solve the onramp problem, or Magic Link that solve the self-custody problem. With all of these new technologies, there could be kinks that still need to be ironed out, but for the first time in the history of crypto, it looks like all the right ingredients are there to make a non-custodial CEX.

This is a problem I care deeply about, given that I had family members lose funds both in Mt. Gox and how all these years later in Celsius. Currently, I run the only 100% free “acorns for crypto” app, called Bits (it intentionally makes zero revenue). That said, at the end of January, I’m upgrading it into the world’s first non-custodial CEX able to solve the three key problems mentioned above: competitive fiat onramping (0.2% on stables, all of which goes to a third party), simple self-custody with automatic almost-impossible-to-fuck-up backups, and timely background order execution – all on-chain and gas-free (taking advantage of a recent Ethereum Improvement Protocol).

All in all, the simple user experience of a CEX, the power of a DEX, and the security of a Ledger. My goal leading up to the public launch is to collect user feedback and attract beta testers.

​