I've been thinking about how Predyx and other lightning based prediction markets might finance their operations, without undermining their core function of eliciting information from people.
The standard approach, of offering less-than-fair odds, guarantees long-run profitability (as long as you have enough customers), but it also creates a friction for participants that reduces the information value of their transactions. So, what are some less frictiony options for generating revenue?
Low hanging fruit
- Close markets in real-time: Rather than prespecifying a closing time for some markets, like sports, it's better to close the market at the moment the outcome is realized. This both prevents post hoc transactions and enables late stage transactions. This should be easily automatable (I say as someone with no idea how to do that), with the right resolution criteria.
- Round off shares: Shares and sats are discrete, so just make sure any necessary rounding is always in the house's favor.
- Set initial probabilities well: Use whatever external information is available to open markets as near to the "right" value as possible.
- Arbitrage: whenever markets are related to each other, make sure to resolve any illogical odds automatically
The point of these four is to avoid giving away free sats. None of them reduce productive use of the market. Keeping markets open up until the outcome is realized will probably greatly increase the number of transactions, since that's usually when the most information is coming in.
Third party support
- Ads are the most obvious form of third party revenue
- Sponsorships are the more interesting one: Allow sponsors to boost a market's visibility. This is similar to advertising, but it also capitalizes on the possibility of a market being of particular interest to someone.
- Charge for market creation: users should be able to create new markets (this will also enhance trade quantity and site traffic), but it should be costly to create a market. If prediction markets really provide higher quality information, then it's reasonable to charge for it.
- Arbitrage: Monitor external odds and whenever a gain can be locked in, place the bets (buy the shares) that guarantee a gain.
Bitcoin stuff
- Routing fees: The volume of sats moving into, out of, and being held in these markets will require a fairly large lightning node. Following some helpful tips to optimize fee revenue will generate some sats for logistical stuff that had to be done anyway.
- Treasury strategy: Take out loans against the revenue generated from all of the above and buy bitcoin: NGU -> repay with a fraction of the bitcoin, NGD -> repay with site revenue.
Bitcoin and Lightning Competitive Advantages
These aren't revenue ideas. They're just a couple of advantages lightning and bitcoin provide over fiat that should allow charging lower spreads than a traditional prediction market or sportsbook.
Traditional betting or prediction platforms are earning depreciating fiat, while a bitcoin based platform earns appreciating bitcoin. Traditional spreads must therefor be larger, in order to pull in the same real return. This also means the users' odds are worse on fiat platforms (again in real terms), even if the listed odds are the same, because their winnings will have depreciated by the time they receive them. Technically, this opens an opportunity to charge even higher spreads, but as mentioned in the intro, that would be bad for the information purposes of the market.
Lightning has much lower transactions costs than fiat transactions. So, even with tighter spreads, a lightning platform can net a better (nominal) return per transaction.
@mega_dreamer, I imagine most of those ideas were already on y'all's radar, and obviously you're already doing some, but I wanted to get them out of my head and onto digital paper. Hopefully, some of this will provide some useful food for thought.
