Coinbase’s quest to stay profitable throughout cycles

Hey!

When Coinbase went public, Brian Armstrong, the company’s co-founder and CEO floated the idea that in 5 to 10 years, more than half of the company’s revenue would come from non-trading fees.

Armstrong and his team didn’t retreat from this ambition during the crypto winter of 2022. In February 2022, they informed investors that the company expects to lose $0.5 billion in Adjusted EBITDA, but will continue investing in building new products.

The bet was controversial in the moment (everyone else rushed to reach profitability), but absolutely meaningful in the long term. Armstrong and his team wanted to make Coinbase less dependent on volatile trading fees and maintain profitability throughout crypto cycles.

In today’s essay, I take a look at how this bet is playing out so far (spoiler aller: better than expected).

Enjoy!

Jevgenijs

p.s. if you have feedback, just reply to this email or find me on Twitter/X

On the day of Coinbase’s IPO, Brian Armstrong, co-founder and CEO of the company, gave an interview to CNBC stating that he didn’t see margin compression (meaning, fees going down) in the short or the medium term. This should give the company sufficient time to diversify its revenue, with non-trading fees contributing 50% or more to its revenue in “5 to 10 years.”

In later interviews, he built on this message, saying that the goal of increasing the share of non-trading fees, which are less volatile than trading fees, is to maintain profitability throughout crypto cycles, which are common for this novel industry. Coinbase went public on April 14, 2021, reporting its highest-ever GAAP Net Income in the first quarter of being a public company.

The down cycle, or “crypto winter”, came in less than a year, with Coinbase’s net revenue collapsing from roughly $2 billion in Q2 2021 to $0.8 billion in Q2 2022. This was a year of major crypto bankruptcies, culminating in the collapse of one of the largest Coinbase’s competitors, FTX.

It wasn’t a bad year only for the crypto industry. The Federal Reserve finally started raising rates, so the Nasdaq Composite finished the year down 33% (the S&P 500 was down 19%), some Fintech companies lost more than 90% of their value, and everyone rushed to deliver profitability.

What stuck in my head was that Brian Armstrong and his team said the company would lose up to $500 million in Adjusted EBITDA, but would keep investing in diversifying its business. Times and times again, he would speak about crypto cycles and the need to grow the non-trading component of the revenue to go through cycles like that.

…and that was the right call.

The crypto industry survived, and trading volumes on Coinbase started coming back in late 2023. While the retail trading volumes are still nowhere near 2021 volumes, Coinbase reported $312 billion in trading volume in Q1 2024, which is comparable to Q1 2021. Coinbase also returned to profitability.

The fees for retail customers did not come down. Yes, the crypto winter took away one of Coinbase’s competitors, FTX, and the world’s largest exchange, Binance, lost market share in the U.S. due to litigation. However, 2024 also brought spot Bitcoin ETFs, giving retail investors cheap access to Bitcoin via their brokerage accounts.

In the meantime, Coinbase’s non-trading fees (reported as “Subscription and Services Revenue”) increased from $56 million in Q1 2021 to $511 million in Q1 2024…

…which represented 32% of the company’s net revenue in Q1 2024. Subscription and services revenue represented as much as 53% of the company’s net revenue in Q3 2023. That’s the level that, on the day of the IPO, Brian Armstrong envisioned to come in “5-10 years.”

In its latest Shareholder letter, Coinbase management guided for $525 - $600 million in Subscription and services revenue in Q2 2024, which would represent 57-79% YoY growth. This guidance implies a 37%-43% contribution to total net revenue, as per current analysts’ estimates.

So what are those non-trading fees, or “Subscription and services revenue”? The components of this revenue stream are 1) stablecoin revenue, 2) blockchain rewards (or staking), 3) interest on customer funds, 4) custodial fees, and 5) various other streams of revenue such as “Coinbase One” subscription revenue.

Over the last 12 reported months (Q2 2023 - Q1 2024), roughly 44% of the subscription and services revenue came from stablecoin revenue, 26% came from blockchain rewards (staking), and 13% came from the interest earned on customer funds (and institutional lending).

Stablecoin revenue is the interest that Coinbase earns on the USDC reserves. USDC, a U.S. dollar-pegged stablecoin issued by Circle, is the second largest stablecoin with a market cap of around $32.5 billion as of this writing.

Circle exchanges fiat dollars for USDC, and then invests the proceeds into the government and corporate bonds earning a yield until USDC is not exchanged back into fiat. As of this writing, $29 billion out of $32.5 billion were held in The Circle Reserve Fund, a money market fund managed by Blackrock.

Coinbase has a special arrangement with Circle where it retains all interest on the USDC reserves for the stablecoins held on the Coinbase platform (Circle is paid a fee). In addition, Coinbase and Circle split interest income 50/50 for the stablecoins not held on the Coinbase platform.

Blockchain rewards, or staking, is the revenue that Coinbase earns from providing staking service to its customers (and from staking its corporate crypto assets). Thus, Coinbase customers can stake their assets (such as Ethereum) and earn staking rewards from the respective blockchain.

Staking directly with a blockchain might require a certain minimum amount of the respective crypto asset (e.g. 32 ETH in the case of Ethereum) and technical know-how. Coinbase removes the minimum barrier (by pooling resources of multiple users to meet the blockchain’s minimum) and makes the process as simple as making a few clicks.

Coinbase charges its users for providing the services by taking a cut from the rewards received from the respective blockchain. Thus, you can see “Blockchain rewards” in the revenue (staking rewards received from blockchains), and “Blockchain reward fees” in the operating expenses (staking rewards passed through to customers). Coinbase also stakes its corporate crypto assets earnings staking rewards.

I will skip the interest income, as earning interest on customer funds is quite common in the financial industry. The only novelty in the case of Coinbase is that this revenue line includes Prime Financing or the fees that the company earns by lending money or crypto assets to its prime brokerage clients.

Coinbase launched custody, which in the case of crypto means storing client assets in cold storage, back in 2018. However, the service got its spotlight with the launch of Bitcoin spot ETFs. Thus, Coinbase is the custodian for 8 out of 11 U.S. spot Bitcoin ETFs that launched earlier in the year. The company is also expected to be a custodian for 5 out of 8 upcoming spot Ethereum ETFs.

There are other subscription and services revenue sources (such as the revenue from the “Coinbase One” subscription), but I will stop here. Nevertheless, there is one more non-trading revenue stream worth mentioning. That’s revenue from Base, a Layer 2 blockchain, and other payment-related fees (reported under “Other transaction revenue”).

In August 2023, Coinbase launched a Layer 2 blockchain, Base. Layer 2 chains compensate for Ethereum limitations, making transactions faster and cheaper. Coinbase is the only “sequencer” for Base (at least for now), which means it earns revenue for “processing” all transactions on Base (sequencers “pull” transactions together and submit them to the Layer 1 chain, which in this case is, Ethereum).

As you can see from the table above “Other transaction revenue” more than doubled in Q1 2024, which is primarily attributed to the growth in Base transactions and, consequently, revenue. Base is (if I am not mistaken) the youngest Layer 2 chain on Ethereum, but it is quickly gaining traction. This is an example of Coinbase using its massive userbase to distribute a new product.

All right, back to where I started. Has Coinbase diversified its revenue enough to maintain profitability in the next crypto winter? In addition, to launching new products and services, Coinbase used crypto winter to optimize its cost structure (including executing layoffs).

Thus, as you can see on the chart below, operating costs (excluding transaction expenses) have been stable at around $660-670 million per quarter over the last year. What’s notable is that the trading volume on Coinbase more than doubled from Q1 2023 ($145 billion) to Q1 2024 $312 billion, but the total operating expenses remained pretty much flat.

If you recall from the above, Coinbase reported Subscription and services revenue of $511 million in Q1 2024, as well as guided for $525 - $600 million in Subscription and services revenue in Q2 2024. As illustrated on the chart below, Subscription and services revenue is still insufficient to cover operating expenses but is quickly getting there.

However, together with transaction revenue, which even at the peak of the crypto winter remained at roughly $300 million per quarter, Coinbase’s revenue should cover the operating costs. So this back-of-the-envelope calculation suggests, that Coinbase is close to having achieved revenue diversification that would allow it to remain profitable throughout the cycles.

Of course, the next crypto winter will not only hurt Conbase’s trading fees. It will be a headwind for the Subscription and services revenue too. Thus, staking rewards are earned in crypto assets, not in U.S. dollars; so, a fall in, let’s say, the price of Ethereum, will result in a decline in staking revenue even if staking volumes remain the same.

The value of custodial assets, and thus, custodial fees, will be lower, Base activity will, most likely, decline in line with the overall market, some customers will cancel their “Coinbase One” subscriptions…and so on. Circle will also not earn 5%+ interest on the USDC reserves forever. But…remember, in 2021 Brain Armstrong gave a timeline of “5 to 10 years”, and we’re just 3 years in. Let the team cook, as they say.

I hope you enjoyed this write-up and became at least a bit more “Based”!

Cover image source: Coinbase

Disclaimer: Information contained in this newsletter is intended for educational and informational purposes only and should not be considered financial advice. You should do your own research or seek professional advice before making any investment decisions.