Will Manifold change fees from a percentage of trade amount to a percentage of profits by August 31?
11
100Ṁ423
resolved Aug 25
Resolved
N/A

Resolves YES if Manifold changes the fees to be a percentage of net profits instead of a percentage of the amount of each trade, by August 31. Resolves NO otherwise. This will be determined by my judgement and/or how the Manifold team describes any changes to fees.

If both types of fees are in use I'll probably resolve to 50%. If fees are entirely eliminated resolves N/A. The fee percentage can be fixed or variable, either would count.

Currently, on yes/no markets, Manifold charges fees on each trade as 10% x amount_traded x (1 - max_price_per_share_bought).

E.g. buying M$50 YES from 50% to 70%, you pay fees of 10% x M$50 x (1 - 70%) = M$1.5, or buying M$50 NO from 80% to 40% you pay fees of 10% x M$50 x 40% = M$2. (For NO the price is 1 - the probability of YES)

Jul 26, 6:43pm: Will resolves YES if fees on fixed-payout (CFMM) markets change in a way that meets these criteria, even if fees on free-response (DPM) markets do not change.

Manifold
Get
Ṁ1,000
to start trading!
Sort by:

Technically zero is a percentage...

predictedNO

@MartinRandall Haha, I actually anticipated this possibility:

If fees are entirely eliminated resolves N/A.

How does this work if I make several trades on the same market, against a variety of liquidity providers? I assume I'm getting a fee based on my net profit, but how is that distributed?
@MartinRandall Well, today the fees don't go to the liquidity providers, they all go to the author. I don't know whether Manifold is thinking about changing that.
@jack Oh. Well that's clearly wrong. If the fee is for rewarding the creator for making a great market, why would limit orders avoid fees? Fees for using the AMM should go to whoever put up the liquidity for the AMM.
I agree that the current fee structure doesn't seem very good. Paying fees out to the liquidity provider makes sense to me. I think the author rewards of M$10 per unique trader are a better way to reward the author than fees anyway, although it might be even better if you got more rewards for people coming back and making more bets. And fees are always a tradeoff between many different things, limit orders avoiding fees makes sense to incentivize them and allow smaller spreads, and if the person filling the limit order has to pay fees then the author still gets rewarded (although I'm not sure if that's actually how it works).
@jack my experience is that neither side of a limit order pays fees when a buy order matches with a limit order. I don't know how it works when a limit order matches with the AMM.
@MartinRandall Yeah I think matching a limit order = neither side pays fees, which doesn't make sense to me, I think you should pay fees on any portion of your order that gets immediate execution and not pay fees on any order that gets put on the order book and fulfilled later. And when a limit order matches AMM I believe it does charge fees, which makes sense.
@jack I thought that before but now I think that for limit orders the spread is the fee.
Fees are brutally high, should be either waived for exits and/or liquidity provision, or vastly lower That said, no market in the world charges percent of profit for many reasons
@Gigacasting PredictIt does charge fees as a percentage of profit. Capital gains tax is also basically a fee on trading profits. What are the reasons not to / downsides? Limit orders (that are not executed right away) do not charge fees, so there is an incentive for liquidity provision. I think fees as a percentage of profit is much more incentive-compatible. Currently if the market is at 50% and I think 55% is the true probability, I will lose money betting towards 55%. That's pretty lousy - the market probability can only be thought to be accurate to within ~10%, which is quite a wide range.
© Manifold Markets, Inc.•Terms•Privacy