@@ -9,7 +9,7 @@ comments: false
99
1010Prediction markets aren't a new concept, but they are a tool being used more
1111widely these days. For the most part, these are being used for pseudo-gambling
12- on events, but there's something more interesting about them for me.
12+ on events, but there's something more interesting in them for me.
1313What if a prediction market could be used as a tool to model risk in the same way
1414What does an actuary do for an insurance company? If we presume that a prediction
1515The market can leverage information asymmetry; could we then use the price
@@ -70,7 +70,7 @@ would not be at least $80 by the end of the day when the user withdraws their
7070funds. In this case, the prediction market on the price would behave similarly to how an options contract works, hedging the risk.
7171
7272However, what if we wanted to hedge some other type of risk, such as the loss of
73- funds due to a protocol hack occurring on Aave, which made it so the stablecoin
73+ funds due to a protocol hack occurring on Aave, which made it, so the stablecoin
7474reward protocol couldn't return funds to the user? We could achieve this by
7575betting "no" on a prediction market that won't be hacked before a set time. If the
7676protocol doesn't get hacked, then the fees won from the prediction market are
@@ -91,14 +91,14 @@ withdrawal because the protocol stayed secure, or their bet is paid out because
9191the protocol was hacked.
9292
9393Additionally, since the prediction market would require evidence of a hack, the
94- The shifting of the yes bet upwards can act as a canary signal to the user to
94+ shifting of the yes bet upwards can act as a canary signal to the user to
9595automatically withdraw the funds. In this way, a hacker could conduct the hack
9696and receive both the funds from the uninsured (since it's an opt-in bet) and a
9797portion of the bet.
9898
9999I'm not totally convinced the incentive structures are properly structured yet,
100100But I think there's merit to the use of prediction markets as an insurance
101- mechanisms . Can anyone come up with a better structure that generates revenue
101+ mechanism . Can anyone come up with a better structure that generates revenue
102102for an arbitrary protocol, protects the user's funds in a catastrophic event like
103103this, and allows the hacker to claim some portion of the funds as a bounty
104104structure (presumably they'd return the uninsured funds to prevent criminal
0 commit comments