I canât support this proposal as it stands because I donât believe Symphony Finance has an abundant user constituency, nor do I think any of the grant proposals have a real shot at increasing the economic sustainability of the protocol or of growth layerwide. Iâm saying this early in this cycle and laying things out without too much structure because the Symphony team have been unbelievably accommodating in answering my many questions, and I want them to be able to respond to this and prove me wrong before I present my thoughts to the Shadow Committee.
The fundamental economic model doesnât work for how the product is most often used.
The claim here is that setting a limit order earns users yield and thus presents an advantageous value proposition to users relative to competing products like 1inch. The problem, though, is that the vast majority of limit orders arenât held for long enough for users to even break even on the limit order fees, much less earn enough yield for it to be worth their while. The protocol earns money on these unit economics, but because users evidently havenât seen a lot of value, there also arenât enough users to lift Symphony out of cash-burning mode.
The vast majority of limit orders are placed for under 12 hours.
There is negligible yield earned on these orders in that amount of time, and itâs uneconomic for Symphony users relative to 1inch limit orders on Optimism.
Assuming a 5k order at 5% APY, users need to leave their orders up for over a day â over 2x the modal amount of time â to break even with 1inch. So something like 80% of people whoâd placed orders with Symphony were worse off than had they placed with 1inch.
You would need to place a 10k limit order at a bull-market stables rate of 14% APY for Symphony to be worth the while of people in the middle of that under 12 hour bracket.
Both of these conditions seem unlikely for the vast majority of cases today.
Setting aside a comparison with 1inch, it takes almost 4 days to break even on a 5k limit order charge at 5% APY:
And how much would they make from yield on that 10k order? $1.30. A rounding error for somebody trying to make money on, say, a directional ETH bet.
Bottom line, for their yield-bearing limit model to beat 1inchâs or other examples on the market since at least early 2021, Symphony must either 1) dramatically increase yields, which seems like a nonstarter as a low-risk yield instrument in todayâs market, or 2) court longer-term limit orders, which appear in short supply/interest. Otherwise, as a user I care much more about placing the order and making money on whatever the order is intended for do than earning yield while the orderâs waiting to fill.
And itâs not obvious to me that thereâs a real constituency of people who need to have some amount of money up for long amounts of time. My thinking is that in most cases of long-term limit buys, people are more likely to farm how they want to and follow the desired asset in question, buying when the time comes â unless weâre talking about setting stink bids, which seem more of a phenomenon with CLOBs, where the amount of liquidity provided by market making is more variable and fat fingers more common than what you see in AMMs. Happy to be wrong here â maybe MEV/frontrunning leaves open some opportunities.
One note, though: I think stop loss orders do make sense, and perhaps thereâs something there (could imagine people leaving orders up on their USDT sells for months, for which there could actually be a pretty cool integration with other protocols), but thereâs not much evidence of uptake to date. Would love to see some data from Symphony suggesting otherwise.
Itâs late and Iâm tired, so maybe my math here is wrong. I kind of hope it is.
There hasnât been real evidence of PMF
Hereâs their daily volume on Optimism Yolo since launch:
and daily amount of limit orders placed
Entire days have gone by with no orders placed.
Even on Polygon, Symphonyâs home eco, there were few days above 30 DAUs on Yolo v2 one year after the protocolâs launch.
I havenât done a ton of research into 1inch limit orders, i.e., whether they enjoy a ton of use, but anecdotally I havenât seen much use, and I donât anticipate much relative to, say, order book dApps like dxdy.
But fundamentally, itâs not like these limit orders are expensive in low-gas ecos. This tells me that fees arenât really a bottleneck; I just see a general lack of interest in this niche instrument. The concept is cool, but does the market care? Iâm not sure it has.
The user grants as structured arenât going to drive growth.
The grant currently offers users rebates on their gas/limit order fees. Thatâs 0.05% money back on a limit order thatâs been placed. Assuming an average limit order of $100 (consistent with the graph below), thatâs 5c per order. And assuming 2-3 trades per person, thatâs still not very much.
Itâs not clear to me where the gas rebates really come in for users unless theyâre looking to court programmatic traders, in which case there should be some discussion of that route.
Iâm trying to game out how much of an incentive really would get people to place limit orders on their profile on a continuous basis after the fact, and even then I look to the example of Polygon incentives and see a large dropoff in use â which makes me wonder how providing incentives in this way on Optimism would lead to any different of an outcome.
Regarding marketing distribution, Iâll just put it quickly. The high-LTV users are likely to be the long-term limit order players, programmatic traders, ponzistable hedgers, niche players interested in a niche instrument. Iâm pretty bullish on TikTok as a marketing channel, but Iâm not sure this product is the kind of thing that resonates with the kind of audience I imagine coming in high volumes from TikTok.
Iâm actually sympathetic in principle to the idea that subsidizing integrations of high-potential products are the right kind of investment in that they catalyze broader growth: but to my view, the potential for PMF has to come first, and thereâs still not evidence of this.
Overall
Iâm going to stop short of suggesting outright that Symphony Finance consider pivoting their model given their hitting something of a wall in runway, as the team seem highly skilled, and their UI really stands out in defi, but thereâs a user gap here that Iâm not sure is obviously overcome by a governance grant or bootstrapped by artificial activity. There are some potentially interesting avenues that take real advantage of defi composability (e.g., i could imagine an end âtraderâ being not an individual but an org using this for one side of a broader strategy), but if theyâre working those leads or use cases, this information needs to be highlighted more clearly so people like me can pick up on it.
Itâs tough for me to have to say this, as itâs been a pleasure speaking with @kakashi and seeing the great care with which the informationâs been provided, but I think there just needs to be a clear path Iâm not seeing.
