This question will resolve positively on the 1st of January, 2030
➕
Plus
154
Ṁ560k
2030
97%
chance
This market resolves positively on the 1st of January, 2030. This question is meant to find out whether Manifold users are incentivized to correctly predict on longer-term markets, and, to some extent, what the implied discount rate is. Note that with Manifold's lending functionality, you can bet the first M$20 for free.
Get
Ṁ1,000
and
S1.00
Sort by:
predicts YES

Im basically looking to get out of my position because of Manifold's onerous loan system. Limit order set for 99% on NO.

predicts YES

@MarcusAbramovitch It seems like the last time you bought YES here was in November. Don't you get your money loaned back to you after 25 days? How does cashing out now increase your balance?

predicts YES

@BoltonBailey Manifold has been doing some weird stuff for the amount you can get back in loans based on some leverage ratio. And also, it's a decaying 4%/day, not a constant 4%/day. You never get all the mana back. It's more like 90% back in 2 months or something

predicts YES

@MarcusAbramovitch Ah ok, good point

@JonathanRay 2% for 7 years? I will take your bet.

These questions only measure what the most optimistic trader thinks the discount rate is

Everyone else who knows the discount rate is higher obviously has much better opportunities than betting NO on questions that'll resolve YES in year(s). Because Manifold's discount rate is obviously much higher than these questions imply, we instead borrow tens of thousands at exorbitantly higher interest rates from whales like @jack and easily profit more than enough to cover that interest. Those loan interest rates are closer to actual discount rates, and are usually at least 60% APY from my experience

The interest rate should be at least 5%, the current risk-free rate in the real world economy. Instead, it is below 1% here. Insanity! Sell your YES shares and buy something that will generate a higher yield, such as betting against Bitcoin rising.

predicts YES

@BowTrix That doesn't work for several reasons:

  • Manifold loans greatly decrease the discount rate on Manifold.

  • Mana and $ cannot be freely exchanged. There is no way to do the equivalent of buying T-bills with mana.

I don't get it how are the first M$20 for free?

predicts YES

@YohanProYT That part of the description hasn't been updated since Manifold changed how loans worked a few months ago, it's no longer accurate.

@IsaacKing Oh okay so there were some kind of "freebets" in the past?

predicts YES

@YohanProYT Yeah, it was the same idea as today's loans, but they were limited to M$20, and you got them all immediately rather than 2% per day.

I'm confused, isn't there (strictly dominating) arbitrage with https://manifold.markets/Nu%C3%B1oSempere/this-question-will-resolve-positive-114eccf1cb27 ? Why is this one higher?
@JoyVoid the reality is the discount rate is so steep that most people aren't bothering investing any significant capital into these.
@MattP Probably another reason to remove fees
Oh, I think I misunderstood the intention here. You're not looking for the true discount rate of M$, you're looking for the residual effects that it has on market probabilities due to there only being a finite number of people who take out loans.
Does this actually measure the discount rate though? There's no downside to buying M$20 of yes, and enough people doing so will drive the price to 100% regardless of what the real discount rate is. Indeed, the loans were implemented in order to prevent discounting from affecting market probabilities; i.e. to prevent markets from displaying the exact type of behavior that you're trying to get this market to display.