The Augur Master Plan
5 min read
Augur’s purpose is to democratize and decentralize finance. We’ll do this by enabling anyone, anywhere, at anytime in the world to create and speculate at a low cost for the first time. If Bitcoin gave us decentralized currency and Ethereum brought decentralized computation, Augur will enable a decentralized financial system. We have cryptocurrency, now we need something to use it with. As a side bonus, prediction markets give us better forecasts for the future, more direct hedging and speculation mechanisms, and finally fulfill the vision set out by Hayek[¹] and Arrow & Debreu[²] long ago.
The first version of Augur will likely be somewhat slow and slightly expensive (think pennies and many seconds per trade), but it’ll certainly be a beautiful glimpse of what’s to come. Any new technology has higher costs in the beginning and lower costs once it’s more optimized later on. The long term plan is to overtake all trading though liquidity and network effects, although in the beginning these limitations mean the first markets and activity on Augur will be surrounding new markets or markets that are currently expensive or limited in certain ways. Think Chinese and Russian investors wanting to speculate on US stocks, or vice versa, people wanting to speculate on sporting events more cheaply, or people who just want to create a new market for something that doesn’t exist at the moment due to multi-million dollar startup costs for creating new financial products. With augur we can remove that million and drop the cost down to “multi dollar.”
Since it’s run on Ethereum, it cuts out the middlemen and brings costs down to the economic minimum to operate things securely. On Betfair you’re paying 10%+ fees, on Augur it’ll likely be 1% or less. For the first time people will be able to trade on a censorship resistant, global trading platform without having to trust counter-parties. Liquidity will be truly global because Ethereum doesn’t care if you are from China or the US or Russia, you’re just a pseudonymous address.
Futures contracts will resolve without trusted third parties, but instead with a security model similar to Bitcoin or Ethereum themselves (vulnerable to 51% attacks, but always forkable in the event of one). In Augur’s case a fork is much more clear cut because it’s over reality, people won’t want to trade or make markets on a platform where it says Obama lost the 2012 US Presidential election. People will even be able to use leverage (using scalar markets on Augur you can effectively get leverage). For example, a market on the price of Apple with a range of 150 to 250 is more levered than one from 0 to 300. Leverage is when for every dollar you bet your position is equivalent to some lever ratio, so 3–1 leverage means for every dollar your position goes up or down you actually lose 3 or gain 3 dollars. This is in contrast to margin, which is when you are borrowing money or shares to buy or sell. Margin with margin calls is functionally almost the same as leverage in practice.
Once 0x is out later this year or next faster trades and dark pools will be possible. 0x enables partially off chain trading so people can create, modify, and cancel orders quickly without having to post a transaction to Ethereum each time. This’ll also decrease costs for placing and taking orders in general. Dark pools being a desirable thing may come as a surprise, but big whales aren’t going to want to be moving the market by placing super large orders on the book, they’ll prefer dark pools just as they do in traditional financial markets. Proof of stake will help with faster trades as well since block times will decrease by about a factor of 3.
While the initial version of Augur will have markets denominated in ETH (which although that’ll work fine for short term trading / markets), for longer time horizon events the volatility will be too high. You don’t want to go long “Apple” on Augur, be right about the company, and end up losing money due to Ether volatility. The solution to this is to allow markets to be denominated in stablecoins, or cryptocurrencies that are stable with respect to the dollar or other fiat money. Makerdao is working on stablecoins, and we do actually also get part of this to some degree out of the box. On Augur you can make a market speculating on the price of ETH in USD denominated in Ether as a way to get a rough stablecoin where one side is levered long ETH and one is short ETH but long USD. Since the collateral/currency it’s denominated in is in ETH the short side is effectively the stablecoin and the long side is effectively levered ether. If Maker is out by then and live it’ll likely be a superior option though.
We’ll also need to add the capability for automatic rollovers via an additional contract; I wrote a high level interface for a contract to do this the other day. The idea behind automatic rollovers is when you’re speculating on a prediction market, say an ETH-USD one, they have expiration dates like futures markets, so you can’t just buy one side and hold indefinitely. An automatic rollover contract would buy perhaps the “What will ETH be at the end of June?” contract. When June is over it’d sell the complete sets for June and take the remaining funds and buy complete sets for the July contract. People would actually be buying shares in this pseudo-ETF that always owned the current month ETH price market and automatically rolled over by selling the last month’s positions each month. This would enable users to just buy one asset and hold instead of having to exit each month and buy back in manually. Compared to regular futures contract rollovers, with this type people would just be paying the complete set fees, but not losing any money due to market movements from exiting and entering a position. Based on some napkin cloth math, they’d likely be a better deal, and certainly no worse than existing futures ETF tracking error (some of the top oil ETFs over the past year had a 75% cost compared to spot price returns, crazy)!
The main remaining piece of the puzzle is to enable more transactions per second, which means using sharding and/or raiden. This is probably the hardest and furthest off. My estimates for the various pieces that’ll help with it:
— Proof of stake (about 1 year and perhaps a 3x improvement)
— Wasm (2 years and a 10x improvement due to much more efficient and faster contract execution times)
— Sharding (3–5 years is my estimate and a 100x improvement)
— Raiden (1.5–3 years and a 1000x improvement)
Once we have it though, we’ll finally have a system that could take over all of the world’s markets, and that’s pretty damn powerful.
1. Release a v1 (“Buffett”) according to the release schedule, it’ll be cool, pretty, and a bit expensive
2. Integrate 0x and release a v1.5 that allows partially off chain trading so people can create, modify, and cancel orders quickly (“0/Zero”)
3. Integrate Maker once out (or if not out, explain how to use / create contract for difference stablecoins) and release a v2 (“Soros”)
4. Add automatic rollover contracts
5. Release a v3 with super high transactions per second using sharding and/or raiden (“Simons”)
What is needed to make a killer prediction market platform
If Augur is to succeed it needs to be 10x better than than any existing system, and this is what is needed:
1. Fees lower than current options available (e.g., \< betfair)
2. Censorship resistant and trustless global trading platform
3. Ability to create your own markets
4. Markets that resolve without a trusted third party
5. Stablecoins (cryptocurrencies that are stable with respect to the dollar)
6. Faster trades
7. Higher transactions per second
10. Automatic rollovers
11. Dark pools
(written by: Joey Krug)