Great read. What do you think of the market and competition in general? Every superapp wants their own Prediction Market. Do you think the market will stay fragmented or consolidate and Polymarket/Kalshi will supply all the apps with odds in exchange for fees?
Also, I'm not very sure if market making on PMs makes sense... Is it a worthy endeavour for institutions or not necessary for this new platform?
Very appreciated, I think binary markets will stay fragmented unless we somehow get a single liquidity layer that everyone can build on top of. But these binary markets are also limited by nature and how much they can scale, because risk is rarely just binary. So going forward, I see them staying fragmented but nice for people who just want to place basic sports or political bets and don't really care about slippage.
MM on PMs doesn’t really make sense at the moment; it’s very difficult due to total loss risk and having to quote wider spreads. On the long tail of events (especially niche ones), you almost don’t want to get filled as an MM.
In theory, you could make a case for market making on conditional events if the traded outcome is price-based, since it would essentially appear as a regular order book (but that’s still far down the line). Kalshi, for example, uses an internal market maker (which isn't even profitable) to provide depth and subsidize liquidity, which isn’t sustainable. Polymarket gets decent flow (especially around sports, elections and other niche events), but that setup isn’t ideal either for broader use cases.
So imo for prediction markets to scale and be used for more than just sports and elections, they’ll probably have to move more towards scalar or gradient markets — “by how much” instead of “if.” We’ve talked to some really interesting teams building around that.
Strong take on prediction markets as data infra rather than just betting. The Polymarket-ICE deal really does signal a shift from fee-based revenue to data monetization, which makes alot more sense at scale. One thing I'd push back on slightly is the oracle problem for non-binary events like "rainfall hedging" becuz verification gets messy quick and dispute resolution can eat into efficency gains. Still tho, the Arrow-Debreu resurgence angle is something I've been mulling over too.
Very much agree. The oracle resolution problem (especially for non-binary or continuous events) is something we've actually been digging into quite a bit. We've spoken with some really interesting founders working on this. Curious if you’ve come across any projects that are tackling this from a new angle?
Great read. What do you think of the market and competition in general? Every superapp wants their own Prediction Market. Do you think the market will stay fragmented or consolidate and Polymarket/Kalshi will supply all the apps with odds in exchange for fees?
Also, I'm not very sure if market making on PMs makes sense... Is it a worthy endeavour for institutions or not necessary for this new platform?
Very appreciated, I think binary markets will stay fragmented unless we somehow get a single liquidity layer that everyone can build on top of. But these binary markets are also limited by nature and how much they can scale, because risk is rarely just binary. So going forward, I see them staying fragmented but nice for people who just want to place basic sports or political bets and don't really care about slippage.
MM on PMs doesn’t really make sense at the moment; it’s very difficult due to total loss risk and having to quote wider spreads. On the long tail of events (especially niche ones), you almost don’t want to get filled as an MM.
In theory, you could make a case for market making on conditional events if the traded outcome is price-based, since it would essentially appear as a regular order book (but that’s still far down the line). Kalshi, for example, uses an internal market maker (which isn't even profitable) to provide depth and subsidize liquidity, which isn’t sustainable. Polymarket gets decent flow (especially around sports, elections and other niche events), but that setup isn’t ideal either for broader use cases.
So imo for prediction markets to scale and be used for more than just sports and elections, they’ll probably have to move more towards scalar or gradient markets — “by how much” instead of “if.” We’ve talked to some really interesting teams building around that.
Strong take on prediction markets as data infra rather than just betting. The Polymarket-ICE deal really does signal a shift from fee-based revenue to data monetization, which makes alot more sense at scale. One thing I'd push back on slightly is the oracle problem for non-binary events like "rainfall hedging" becuz verification gets messy quick and dispute resolution can eat into efficency gains. Still tho, the Arrow-Debreu resurgence angle is something I've been mulling over too.
Very much agree. The oracle resolution problem (especially for non-binary or continuous events) is something we've actually been digging into quite a bit. We've spoken with some really interesting founders working on this. Curious if you’ve come across any projects that are tackling this from a new angle?