Thesis
ERC-20 spot trading is increasingly moving to decentralized venues despite UX hurdles endemic to their Layer 1 implementations. Futures trading will follow the same path as Ethereum scaling solutions are released. The protocols that enable these futures will stand to profit massively from this migration. In advance of this migration, I’ve added PERP to the Pineapple Street Allocators portfolio. (Though I didn’t add enough, and was a classic dick for a tick not buying my third tranche in the mid 14.00s).
Rationale
Product
While much has been made about crypto’s terrible UX, there’s no denying that self-custody of your assets provides for a much smoother trading experience than its centralized counterpart. I’ve personally got no fewer than 4 centralized exchange accounts. Moving money and crypto between them is operationally annoying, and causes complexities from an accounting perspective that I intentionally ignore except for the two terrible weeks every year when I do my taxes. Being able to keep all of my assets in an account that is instantly compatible with and connectable to an infinite number of protocols with which I can swap, borrow, lend, pool, stake, et cetera is an incredible upgrade that I think we often take for granted.
This reality — that for the vast majority of users, the only advantage to a centralized exchange is the fiat onramp — is becoming apparent. We’ve seen an explosion in spot volumes in decentralized venues.
Beyond the ease of access, there are a couple of other factors that I think could be driving this growth. UX is certainly still an issue; the asynchronicity of blockchain transactions is not a straightforward problem to solve. However, I think if you take a look at user interfaces today and compare them to those that first started popping up in 2019, you’ll agree that they’re greatly improved. Additionally, there’s an interesting combination of products that assume a lot less technical expertise in their users (I’m thinking e.g. PoolTogether here, not e.g. Curve, which I think is intentionally obscure but that’s another story) and users who in general have a much better understanding of what’s happening on the back end. This is completely anecdotal and I’m not sure how to measure it, but the level of discourse in the vast array of Telegram channels and Discord servers where I lurk has risen by a number of degrees over the last couple years. As more people start being more comfortable sending an Ethereum (or any other blockchain) transaction, the benefits of self-custody become more apparent to them.
This will only improve with as rollups and sidechains and various layer 2s gain traction; the ERC-20 approve-transact flow will go from:
- find a trade you want to make
- remember that you have to approve a dapp to spend your tokens
- send the approve transaction, pay gas
- switch browser tabs while you wait for transaction to mine
- get distracted doing something else
- remember the trade and come back at some later date
to:
- find a trade you want to make
- remember you have to approve dapp
- approve dapp, send transaction, receive confirmation instantly
- execute trade
Finally from a product perspective, I’d be remiss not to mention the incredible growth of the dollar-denominated stablecoin market. Though the below thread wanders into heady dollar hegemony waters, it is illustrative of the fact that the dollar is the de facto unit of account in every crypto trading venue where it matters.
The ability to hold dollars in your Ethereum wallet, and to understand how your trades across various protocols impact that dollar balance, makes the trading experience WAY easier to grok.
Regulation and Knowing Your Customer
We’re also seeing regulatory headwinds in some form or another reflected in the operations of every major crypto exchange, most recently with Binance announcing mandatory KYC worldwide.
I’m not making the argument that we won’t see some sort of regulation in the decentralized space. But I do think that sort of enforcement will take time to figure out, and will result in a game of cat and mouse that, for a while, will mean that there’ll be non-kyc options for those traders who look for them. Meanwhile, Compound and Aave are providing institutional-grade onramps into DeFi that are embracing KYC regulations. DyDx, whose token is launching this month, has blacklisted any user that they suspect of being from the usual list of bad actors: Syria, North Korea, the United States, et cetera.
This is probably a topic for another post, but suffice it to say that I believe the option to transact without submitting personally identifying information to a centralized authority will continue to be a draw for a certain subset of users, and as those users are pushed out of centralized exchanges, they’ll move to decentralized venues.
Market
The above isn’t exactly rigorous analysis; the decentralized product is better than the centralized product, users are waking up to that fact and using the decentralized product more. But the product we’re talking about is spot crypto trades. Which is a big market, but is anywhere from 5–10 orders of magnitude smaller than the crypto futures market, which is still executed almost exclusively off chain.
At this point we know three things which I’ll list below, along with some notes where applicable:
- Crypto spot trading has been migrating to decentralized exchanges
a) not all spot trading, but a meaningful amount!
b) DEX tokens are valuable; they make up a large portion of DeFi’s market cap
- Crypto futures trading has largely not migrated to decentralized exchanges
a) margin management is not straightforward and is hard to solve in a decentralized way
b) futures protocol tokens make up a small portion of DeFi’s market cap
- Centralized crypto futures volumes are much greater than centralized crypto spot volumes
So this sets the stage for a big move up in decentralized futures protocol trading volume, and likewise a big move up the corresponding futures protocols governance tokens. My investment scope is currently limited by the Enzyme investable universe, which at the moment means I’m limited to the Perpetual Protocol as an investment option in this space. Luckily, PERP is awesome.
Protocol Dynamics
Perpetual has seen great growth this year. They’re deployed on xDai, and employ an architecture based on their own virtual automated market maker (vAMM). I won’t really go into the details of v1 because v2 (aka Curie) is being released imminently, and it represents a huge change in the way the protocol works.
Curie is a set of protocol that accepts user deposits as collateral and manages that user’s margin and leverage to mint a corresponding amount of virtual tokens for the user to allocate as they see fit into market-specific Uniswap v3 liquidity pools.
That’s a lot to digest. In practice, it means that you can deposit USDC to the Perp Clearing House and mint either vUSDC or vOtherToken in an amount equal to yourDeposit * yourDesiredLeverage. You can then take those virtual tokens and trade them in the vUSDC/vOtherToken liquidity pool to take a directional view on that token pair. Your position functions essentially as an NDF; when you close out, it's always deliverable back to USDC.
The best set of docs I could find has a good example of how this works. In it use Alice deposits 100 USDC to the Clearing House and wants 2x leverage. She gets 200 vUSDC, which she then uses to buy 1.96 vETH in the vUSDC/vETH Uniswap pool (not exactly 2 because of the slippage her trade incurs). The Clearing House does all the accounting. At this point, Alice has 100 USDC of collateral and a 1.96 vETH long position with a 200 vUSDC cost basis. She closes out when vETH has increased in value by selling her 1.96 vETH for 220 vUSDC. Her PnL is equal to the vUSDC she receives minus her cost basis (220–200), and her collateral account increases by 20 USDC.
There are a couple of big upgrades here. The first is that every positive USDC of PnL in Alice’s position is accounted for by a negative USDC in the offsetting Market Maker’s position. This means that with good margin controls, systemic risk to the protocol in the event of a large move in token price is less than the current setup. Currently, because every participant trades against a monolithic x*y=k liquidity pool, there is potential, if every participant is the same way, for a liquidation cascade that causes losses greater than the insurance fund (the dynamics of which I'm not going to go into here). Individuals are certainly still prone to be blown up, but the risk is largely siloed within the participants themselves rather than spread across the protocol as a whole.
The second big upgrade is potentially more important to the adoption and increased usage of the Perpetual Protocol. In the centralized world, we’ve seen FTX and Deribit win market share (in futures and options respectively) because of optimal portfolio margining systems. Perp (taking inspiration from FTX, according to the documentation) will implement a similar portfolio margining system, meaning that trader can deposit USDC, enter a long position in wETH, and use the PnL from that position as collateral for a position in another token pair. (This is according to Messari here, though I’m chasing it down in the Perp discord too).
Finally, and this is the part I know the least about, Curie (and UniV3) will be running on Arbitrum for increased throughput and lower fees. It’s hard to suss out exactly from the docs how this will work, but the promise of low fees in today’s world of NFT mints is promising. I’ll be keeping an eye on this and will update with a subsequent post when I know more about it.
Token Economics
TBD! There’s currently a very generous staking program but it seems like the particulars are about to change, given two recent governance posts by Delphi Digital. It doesn’t seem like the net rewards will decrease, but I do think the way they’re distributed will change. I’ve stayed quiet on the first post and the second was just released today so I’ll read and update in a subsequent post.
Risks
The bit I mentioned above about risk being siloed within individual positions comes at a cost. You need someone to provide liquidity in the first place. Given the Perpetual Labs cap table (Alameda, 3AC, CMS, etc), I would think that there’s a plan to bootstrap liquidity. There’s likely a good bit of structural arbitrage opportunity for the firms that are poised to undertake the operational steps required. It’s not quite standing in line at a Japanese bank to wire Yen to dollar accounts, but the various on-chain / off-chain transfers and what I assume is an asset bridge onto and off of Arbitrum aren’t exactly straightforward. Given the friction, It seems reasonable that prices on Perp will react more slowly to market news than prices on the centralized exchanges, providing a classic buy low sell high opportunity.
This friction is another risk in and of itself. No one really knows how the Layer 2 scaling wars are going to play out as it relates to liquidity. Will there be enough volume on Arbitrum to make Perp a meaningful decentralized futures exchange? This sort of chicken and egg scenario is hard to predict, but I would guess (and I’m buying PERP to back it up) that Perp itself can be a draw to bring users to Arbitrum. It’s currently handling the vast majority of decentralized futures volume on Ethereum; I believe that it will continue to do so.
And then there is the ever present regulatory overhang. Will the prohibition of US users be enough to stave off the SEC and the CFTC? What happens if the EU follows suit? How long can the cat and mouse game continue until there’s a truly decentralized front end hosted on IPFS to which Gary Gensler cannot serve a cease and desist? My personal view, having bet my career that DeFi will succeed as a product and an asset class, is long enough.
Final Thoughts
Things I didn’t mention, but are important and may be topics for future tweets or essays:
- the funding rate mechanism and oracle providers
- liquidation dynamics
- UI
- the potential for anyone to spin up a Clearing House for any arbitrary market
Things I’ll be watching:
- the tokenomics governance proposals
- liquidity on Curie
