How Augur v2 Can Succeed
Big Challenges and Big Opportunities

If Augur v1 was like trying to stream HD videos in 1995, Augur v2 is something like trying to build Uber in 1999. The infrastructure and building blocks (0x, Uniswap, DAI) are a bit further along, but it’s an even more ambitious undertaking and still…early.

That said, it may soon be too late.

If Augur doesn’t start getting real traction in the months ahead, its long-term success and relevance are anything but guaranteed, given shifts in the competitive landscape and regulation. 

So the goal here is to take sober but optimistic look at the state of Augur, its current challenges, and what to do about them. The focus will be what can be done at the ecosystem level by independent community members, devs, and entrepreneurs building on Augur in v2 to drive the project forward. Augur is a radically open and decentralized project, so it requires contributions from many people to succeed.

Why Bullish

Before diving into the challenges that Augur faces, let’s touch on why the project has so much potential and may be one of the most underestimated projects in crypto.

In short, I believe that Augur can realize the promise of prediction markets and aid in DeFi’s ambitions to open up finance, while delivering a user experience that blows centralized alternatives out of the water. It is also one of very view projects in DeFi right now that has potential appeal and product-market fit outside of existing Ethereum users.

I see Augur as the most sound prediction market protocol and oracle in existence. It is decentralized all the way through, which makes it resilient to censorship, corruption, and single points of failure. Decentralization may be a nice-to-have for some projects, but it’s a matter of survival for prediction markets (at this point in time), which have too often been regulated out of existence or utility. Coupled with being open-source and open-equity, this resilience gives developers and entrepreneurs the ability, trust, and incentives to build on Augur. 

I consider being decentralized, open-source, open-equity (and early) as the base ingredients for a prediction protocol to become liquid, which is ultimately the only thing that matters. Liquidity is what delivers utility for users, as well as robust predictions.

Augur v2 introduces faster market settlement, cheaper market making and pooled liquidity on 0x Mesh, stablecoin denominated trading with DAI, a smoother signup and fiat onramp flow, risk mitigations for Invalid markets, and onchain affiliate incentives for user acquisition.

Perhaps more significant than anything at the product level is the market. Augur v2 arrives in the middle of the most-predicted event in the world, the U.S. Presidential election cycle, and amidst an avalanche of new interest in DeFi. 

Augur has potential product-market fits in 1) high-stakes political bettors who are underserved by centralized alternatives due to low limits, high fees, and limited access and 2) in DeFi enthusiasts who want to speculate on crypto-related outcomes like future TVL or the release date of Ethereum 2.0.

The Road Ahead

Now that we’ve seen Augur’s potential, let’s look at the challenges that lay ahead and what to do about them.

Open Order Books and Ethereum Are Still Early

0x Mesh allows cheaper market making and shared liquidity pools across different UIs. But it doesn’t resolve the frictions of decentralized trading and even introduces some novel challenges. 

Costs for filling orders on Augur’s Reference UI will be prohibitive for small to medium-size traders. Posting orders is free but manually canceling or modifying orders incurs fees, impeding market making at scale. The disparity between order creation and filling costs may open up attack vectors involving spam orders. Finally, retrieving the state of open order books on Mesh is slow, which limits the UIs ability to sort markets and display their current state in an efficient manner. 

I think that the future of prediction markets is open order books. Market makers need order books to manage risk and attain edge. Even if liquidity provision is ruled by bots, these bots will operate on order books. But it will take time for open order books to scale. Until then, Automated Market Makers (AMMs) may be the best way to bootstrap liquidity and minimize friction for both liquidity providers and traders. If you look at the broader DeFi world, AMMs dominate DEX liquidity.

Courtesy of Matthew Finestone

AMMs like Balancer can deliver lower-fee, lower-friction access to Augur markets. And these can work in symbiosis with order books. For example, most large orders may be routed through 0x to minimize slippage, while most small to medium orders may be routed through AMMs to minimize fees.

Augur’s devs have already tackled the phenomenal engineering challenge of building a market around global order books in the early 2020s. So Augur now has the potential with some modest additional efforts to be the first prediction market platform with both order books and AMMs, working in symbiosis. These efforts are already underway, with AugurDAO (a new Aragon Dao in the making) and Guesser working to make this a reality. 

Fragmented Liquidity

I recently wrote about the risk of fragmented liquidity and clone markets on Augur v2. I won’t repeat much here, but in short, liquidity may fragment among markets, thwarting network effects for Augur growth and stunting the potential of global liquidity pools. I look at some potential solutions in the article linked above. The Reference UI may address this in some ways, but it will be important to build alternative, curated UIs to address this as well as other limitations. 

Invalid Risk

Augur v1’s biggest pain point was likely Invalid markets. Thankfully, the sort of Invalid scam markets that plagued v1 are less likely in v2. However, other types of Invalid risk remain. 

Markets sometimes cite centralized resolution sources, which may go offline, be manipulated, or attacked. While Invalid risk may be insured against, given a willing counterparty, the risk of manipulated valid outcomes is harder to hedge. The Augur community will need to continue to advance and refine market standards and terms that are resistant to single sources of failure or manipulation. 

While known invalid risk like obvious errors or vulnerabilities in market terms can mostly be priced in and signaled via Tradable Invalid, unknowns such as event delays and cancellations remain a risk. In fact, for uninsured traders, losses are greater in the case of Invalid resolution than they were in v1.

So offering Invalid insurance at scale may be a useful (and profitable) service. Additionally, for some markets, full, automatic Invalid insurance can be an effective solution. Micah Zoltu has proposed a solution for fully insuring traders and has written an AMM that achieves just this. This sort of solution may work well in cases like sporting events where delays or cancellations are relatively common, where volatility increases at a set time allowing liquidity to be pulled, and where most users are placing short-term buy-and-hold bets, so the lack of transferability and early profit taking (in fully insured shares) are more tolerable.

Settlement Delays

Some markets in v2 may still face slow and uncertain resolution. So early settlement will be a useful service for traders and a potentially profitable one for providers. There are already efforts underway to offer this service.

Settlement providers will buy winning shares from traders at just under expected payout and collect the difference. From the trader’s perspective, you pay a small fee to collect winnings right away and avoid the risk of an unexpected outcome.

Given risks around market settlement delay and Invalid resolution, early settlers will need to manage risk with caution. By preemptively staking added REP on correct market outcomes, they may expedite resolution. Between that and limiting instant settlement to select, vetted markets, it may be a profitable service.

DAI

In Augur v2, traders use DAI rather than ETH. In the long run, stablecoin denomination is a big win for Augur as it reduces currency risk and expands the universe of potential users. This may not be true, however, in the shorter term.

Augur v2 may start off in a sort of wilderness where on one hand it’s not yet accessible for most non “crypto-natives” (users who would prefer DAI) due to remaining frictions like fiat onramp fees, usability and trust barriers around decentralized apps, while at the same time, some (or many) crypto-natives would prefer to trade in ETH rather than DAI.

Particularly if ETH is in a bull market, many users may prefer ETH exposure. Users may maintain some ETH exposure by borrowing DAI with ETH collateral, but this adds a friction point and doesn’t let a user maintain as much exposure due to a collateralization ratio.

The single largest bettor on Augur v1 (the midterms whale), told me that he wasn’t planning on using v2 since he wanted full ETH upside and didn’t trust DAI. While he may be an outlier, I doubt he’s the only one.

Enterprising devs may build solutions to let traders denominate in ETH and maintain full ETH exposure. For example, a platform that let users make a trade in ETH with most of the ETH being swapped for DAI-denominated shares (via Balancer) and the remaining portion buying a levered long position in ETH (using an Augur scalar or another protocol like dYdX under the hood) that would offset the DAI exposure.

Liquidity

I see liquidity and its associated network effects as the most important, if no only, long-term moat for a prediction market protocol. Liquidity is, in part, an emergent factor that arises as a result of many other things (some mentioned here). But I also think that it’s essential to jumpstart liquidity in the early days of a PM protocol. This is especially true in a high-friction environment such as Ethereum and where the team developing the protocol cannot add liquidity itself for legal reasons. 

Guesser and AugurDAO are working on architecture and incentives for liquidity provision in Augur-Balancer pools. Additionally, integration with Hummingbot and other market making tooling may facilitate liquidity provision on 0x a bit further down the road. 

User Acquisition

Even if Augur v2 hits the point where it has liquid markets and a sufficiently frictionless UX, nobody will use it if they don’t know that it exists. The reality is that only a tiny fraction of the universe of potential Augur users is even aware of the protocol’s existence. So user acquisition and marketing efforts will be critical to introduce the world to Augur. 

Some channels I think are especially promising are crypto trader YouTubers, DeFi podcasts, political betting venues, and Twitter…but there are many others. Marketers may do targeted air drops or “free rolls,” small amounts of Augur shares or Dai earmarked for Augur usage, given to users who have a high chance of conversion. I think it will only make sense to ramp up user acquisition efforts once there is an Augur UI with reliable conversion — otherwise, these efforts may backfire if newcomers are turned off by a high-friction UX.

Moving Forward

Augur’s success hinges on the ingenuity and hard work of its developers, community, and surrounding ecosystem in the months ahead. The challenges are real but so is the potential.

I don’t think there’s ever been a better time to start building on Augur or to get involved with the project in other ways. If you’re interested in getting more involved in the Augur community check out the AugurDAO Discord

Thanks for reading. To stay ahead with fresh insights on the future of prediction markets, join The Augur Edge.

Thanks to Ryan Berckmans for reviewing a draft.