Poll: /SG/which-manifold-loan-system-do-you-l

the 50% cap including legacy loans is painful. i'm at like 300% right now

also this breaks long term markets again

@SG for markets which are closed but not resolved for a long time, does interest stop accumulating at the close date or it continues to accumulate until the resolve date?

nice job devs

This version makes Loans usable again; the fees are low enough that it's worth taking Loans in many cases if you see a good deal. And the ability to take out up to 10% of net worth at a time enables that use case.

OTOH, I think it ruins the original purpose of Loans - to make investing in long-term markets worthwhile. The old system did that by making leverage hard to get - you had to wait while it drips in, and then you lose it when the market resolves. So parking large amounts of mana in long term markets (a bit more than 5% net worth in each...) really was optimal. With this new version, you can get the max leverage basically whenever you want. Long-term stuff has no real advantage, and therefore get killed by opportunity cost.

@AhronMaline @SG

OTOH, I think it ruins the original purpose of Loans - to make investing in long-term markets worthwhile. The old system did that by making leverage hard to get - you had to wait while it drips in, and then you lose it when the market resolves. So parking large amounts of mana in long term markets (a bit more than 5% net worth in each...) really was optimal. With this new version, you can get the max leverage basically whenever you want. Long-term stuff has no real advantage, and therefore get killed by opportunity cost.

10x this!!!

@AhronMaline yep, that's why this change was paired with the 5% interest rate

@SG

that's why this change was paired with the 5% interest rate

Interest rate does not help at all - you get the exact same return whether you invest in a sequence of short-term bets or one long-term bet. But with the original loans, one long-term term bet gives you a lot more leverage than a sequence of short-term ones!

@AnonUser Also because, due to compound interest, betting on short markets is more advantageous for interest rates.

@AnonUser Interest rate helps with the problem that there is an opportunity cost for what to do with your dollars (instead of buying mana), not for what to do with your mana (betting in other markets). If there wasn't a peg to the dollar (one way or two way, doesn't matter much), it would be better (IMO) to try to run the mana economy in at zero interest rate. But, given that they care about people buying mana and having the ability to do so (which I support!), I think it makes sense to have an interest rate.

@AnonUser That said, the point you're making otherwise remains valid, and is why I think the free loans complement the interest rate. Both are good.

Interesting! I think this will work out nicely for the way I use the site personally. I can see some users ending up perpetually at their maximum loan amount though - and maybe that's fine.

@SG @mods do you want to count limit orders towards the holding period? I think it makes some sense to reward limit orders with interest, even if it's at a lower interest rate than filled orders.

@Hakari erm, but there's no limit to the amount of orders you have up at any given time. The mechanism of needing to have X balance to put up an order of X is fake. Just the vaguest suggestion on top of a system of unbacked orders.

@VitorBosshard Yeah I didn't mean ALL limit orders, I meant 'limit orders that are filled' should be rewarded interest from the limit order placement date, not the order fill date.

@Hakari ...ok, and this is still interest paid on something I can create millions of without the balance to back it up. It's weird in particular that you want to reward older orders more, when they're functionally identical to newer orders.

I'm not against somehow rewarding limit orders. but that would require an entirely different system of them.

@VitorBosshard you're being a little facetious right now. You can create millions of limit orders but (as I write this) the system is that interest is only paid on positions that were CORRECT.

So:

1) (Millions of?) Limit orders that were not filled wouldn't earn any interest.

2) Limit orders that were triggered but you didn't have the balance to back it up won't earn any interest.

Lastly, older orders are NOT functionally identical to newer orders. Orders should ideally be filled on a FIFO (first-in-first-out) basis. Let me know if this is not the case.

A big part of the motivation for this change is increasing fun and adding more novelty to the platform.

Slow daily drip of trading capital = boring

Leverage on demand = fun

@SG Is this a larger or smaller motivation than that for 'improving the accuracy of long term markets' (or however you want to word it)?

because what you said here is a more believable rationale from your prespective than what you have said about the loan and interest changes up to here imo.

@No_uh my goals with this change are increasing platform accuracy and calibration, encouraging more site usage, and increasing fun. all three are important

@No_uh but the idea that leverage is fun is not something anyone has mentioned yet in this thread

@SG You've obviously got a lot on your plate right now so no rush but whenever it's possible do you or any of the other devs have any results from an overview you did showing that long-term markets were concerningly inaccurate? and if so, can we see those some time? or if anyone else reading this has data showing that manifolds accuracy on longer term markets (say, 1 year out from resolution) is bad, please share it!!

@No_uh things can always be better!

@No_uh an additional motivation is just novelty: i feel like the site has gotten a little stale recently and it's been a while since there's been a big change...

@SG Don't feel that way on my end (re: staleness) but you have a better bird's eye view than I!

re: things can always be better. I don't disagree. Just, if longer-term market calibration/accuracy was somewhat close to short-term, I would've expect more smaller granular changes than if it were a serious problem for taking the site seriously, a more novel approach or drastic change might be warranted.

Thanks for answering

@SG "an additional motivation is just novelty: i feel like the site has gotten a little stale recently and it's been a while since there's been a big change..."

Seems like a bad motivation to do anything in my opinion. Changes are more likely to make things worse than better on average. Most mutations are deleterious. Manifold is/was really, really good. Close to optimal. Hopefully this change will get us even better, even closer to the best the site can be, but the better the site gets, the harder it is to achieve that if your goal is partly "I'm bored; let's change something".

@SG can you add filters for 'listed' and 'ranked' in the market search/sort? I'm probably too new but I have no idea how to identify that.

Tried to sell, pressed "Max," & it won't let me sell

@ChurlishGambit Happening on every market I try. Can't liquidate my positions with interest.

@ChurlishGambit I'm curious, what happens if you sell 1 share first.

@Eliza I don't know—& I'm not in the mood to do more free beta testing for the owners, at the moment, I'm sorry

@ChurlishGambit @SG ?

Will it work?

This seems more confusing than I thought. The interest is actually an increase in shares, but these shares do not appear on the main market screen (which causes me problems when I have limit orders to sell and have to click twice more to find out how many actual shares I have) and does not even appear as the value of the shares. Basically, if I sold the shares, I would have a net gain of 1,500M in my assets, which makes no sense...

@Emanuele1000 would it make more sense if there was a button "collect interest" to press which would collect your interest in shares but not sell anything. You would just increase the number of shares that you hold. Then you could transact as normal with your greater number of shares.

@Emanuele1000 And this goes against my interests, but I'm pretty sure that the interests are calculated incorrectly. The market has been open for two months, and to get 1,500M with 5% per annum, I would have had to have an average of 180,000M invested in that market, and I'm 99.9% sure that's not true...

EDIT: Perhaps it considers all purchased shares, even if they have already been sold?

@Eliza I may be missing something, but wouldn't it be simpler if 34574 shares simply appeared on the main screen as the main data, and consequently also the actual value of the shares I own?

@Emanuele1000 I am also seeing interest rates that are much too high in the same screen. I haven't tried selling yet.

@Emanuele1000 the way they implemented it is you have to do something to cause the shares accumulate because they can't drizzle them in one per second or one per minute or one per day.

@Eliza I'm not a programmer, but if the number appears every time I click on the sell button, even if I don't actually sell, couldn't that number appear directly on the main screen and influence the rest of the numbers?

EDIT: It would also solve the problem of compound interest, because as it currently stands, it is more advantageous to invest in monthly markets than annual ones (or at least to sell and buy back every month, if possible).

@Emanuele1000 Yep. But AFAIK it's "expensive" every time they want to show the new number. If there was a button to press to do the expensive thing, then you would have an up-to-date (within the amount of time you have been on the page) value of your share count at least.

@Eliza Wait, "expensive" how?

@ChurlishGambit in the year of our lord two thousand and twenty six it takes compute time to calculate interest

@EvanDaniel ...so if we view the interest number often, we're costing the site extra money?

@Emanuele1000 your confusion convinced me that the better ux is for interest to just be paid out at market resolution. if we could pay out interest early in mana (rather than shares), that would be good ux, but it's an exploitable mechanism. interest payments at market resolution is simple, easy to understand, and financially sound

@SG that makes sense to me as long as people who were right yet sold early will still receive interest at market resolution proportional to their holding time.

@Hakari yes, they will!

@SG No, I think it's the opposite. Payment in shares is much simpler and more intuitive. It makes it easy to understand that those who make incorrect predictions will not receive any interest, allows users to easily understand how much they have received in interest (it becomes practically impossible to calculate the interest you would otherwise earn, you would have to calculate the integral of your share balance multiplied by the value of the shares, which is not even data you have) and prevents users from receiving payments for markets in which they own nothing. The problem is that the interest in shares was on a hidden screen and did not correspond to the number shown on the main screen.

Furthermore, receiving interest in the future will effectively reduce that 5% per annum, but perhaps this is intentional to prevent compound interest. However, if that is the goal, simply choose markets that actually last less time...

@SG So if I hold for a year, sell, and then have to wait another year for resolution... I get paid a year worth of interest, but it gets paid late? Do I collect interest on that interest for that year?

Maybe another issue is that the further in the future the resolution is scheduled, the less likely I am to ever receive the interest payment (or be around to enjoy it). That's not a problem for the resolution itself because I can sell my stake in the resolution at any time. But I can't sell my stake in the accumulated interest.

@travis Exactly.

This seems like epicycles...

Ooo, a suggestion: it would be great if limit orders could optionally take a loan if you don’t have balance for them when they execute.

@moobunny great idea!

@moobunny Sounds like it needs a warning label. 🔥

@EvanDaniel it could be spicy but it's the time you most urgently need unexpected liquidity, and once the order goes through you can either take a quick profit or shuffle things around to pay it off quickly if you like.

@moobunny I mean if we're trying to convince people to emergency buy mana you're not wrong

The interest behavior is still unclear to me. Reading comments, I think it's actually working like this? Can someone confirm or deny?

- Imagine I make a question with 100m. We can imagine there's 100 shares in this market's liquidity pool (100 yes and 100 no)
- The cost to buy one of those shares in the liquidity pool (or the payout if you sell a share back into the pool) goes up over time, at a continuously compounded rate that ends up being 5% per year.
- Resolutions also take this compounded growth into account.

Is that how this is actually implemented? That makes a lot of sense to me, but... I'm not sure that's what is actually happening. It seems, at first glance, like that implementation should make it so you can buy & sell at any time, and get your interest correctly, right?

An alternative implementation I can imagine, which I think might turn out to be identical, would be for all share balances to grow by some small number every day at midnight or similar. The balance in the liquidity pool, the balance owned by each account, ... Each goes up. I'm not sure if that's actually the same or not, I think it definitely has different rounding behavior, but if you found some way to make it happen continuously, I guess it might be the same net effect?

Or are both of these completely wrong, and it's working some other way?

@DannyqnOht Well I'm also confused about the compounding because it seems like you can get more interest by regularly selling and rebuying your shares (forcing compounding).

@travis I don't think it works that way, they don't want you to compound. But try it out and see.

@DannyqnOht The way I undestood it is different: up until resolution, no interest is paid at all. Then at resolution, if the market resolves YES, then all past or present YES holders get a payment computed as

0.05*\integral (number of shares held) dt

Where t is given in years.

But this is just my reading; I agree it's not at all clear.

@AhronMaline They updated it, the post has changed slightly since the original version, specifically here:

If you sell out of a position early, you'll be able to sell shares from accrued interest as well.

@AhronMaline I think that might ultimately be the better UX

Yeah if you click sell they now let you sell more shares than you "have". Which is odd, surely I have the shares if I can sell them? And it looks like they account for compounding (I bought these shares 56 days ago but I don't know how to calculate what it should be.)

@Eliza huh good point. Okay then I guess it must be the second thing Danny said - all share holdings grow continuously? Including the "fixed product" used by the AMM to compute prices...

Or maybe the Interest shares are "ghost" shares that don't affect the AMM prices, but can still be sold? And also don't earn interest themselves? That would mean you can gain by selling them to buy "real" shares at the same price, as Travis said

@AhronMaline I don't think the liquidity pools are changing, they all still show round numbers. Maybe they just pay that interest out at resolution.

@AhronMaline Well I thought holding 1949 shares for 56 days might have earned me 15 new shares (56/365) * 0.05 * 1949. But they only gave me 11, but in the discussion above they are seeing higher numbers than they expected.

@travis I found a market where I have held 614 shares for about 6 hours and it wanted to give me 41 interest. But I have done a lot of buying and selling in the past on that market so perhaps that is why it is confused.

Leagues and rankings are already kind of meaningless, as small users must work much harder for the same absolute profit. This change makes the problem worse, as you're increasing the amount that new users have to invest to get on a competitive level with the big names.

What's the point of betting with a large vs a small balance? There is very little, other than a fake sense of accomplishment which is priced at tens of thousands of dollars these days. Insanity.

I bought $60 worth of mana back in the day, mostly to support the site, and it doesn't add to my long-term enjoyment in any way (it was a bit of a rush in the moment of course, but that was entirely fake and disappeared as soon as I recalibrated to my new net worth). All I care about is doing well, relative to my starting point.

I like the play money concept, it makes it more interesting and dynamic than metaculus (which feels tedious, while manifold feels like a game). But I think you guys are going about this the entirely wrong way. It's completely pointless to buy any amount of mana if it doesn't get you to a master level balance. So the business model is purely whale-based.

Looking at manifold from a game design perspective, it has all the worst features of gacha games. Achievement laundering. FOMO based sales tactics. Give free accounts an awkward trickle of resources, carefully making sure you can't have too much fun with them. Constantly rub in the pleb's faces how pleb they are.

Maybe mana should be deflationary rather than inflationary? Back when mana could be used for charity, I suggested that accounts should be forced to donate x% net worth every month, which balances user power over time. Those siphoned off winnings could then contribute to a "hall of fame" status you accumulate over years.

Alternatively, there could be "game modes" if you will, where I can start at an even footing with everyone else on the site. I'd be happy to pay $10 to participate in a seasonal league (3-4 months?) where everyone starts with the same balance, and maybe we restrict trading to a theme, or a curated list of markets. The mana from that league would then merge with your account once the season is over, so it'd be like buying boosters and drafting with them before putting them in you collection.

@VitorBosshard They already laughed me out of the room for the last idea, but I still stand by it. You need to use the phrase "colored mana" for maximal effect.

@VitorBosshard That last idea. That's good game design. I like that.

@VitorBosshard access to leverage will disproportionately help smaller accounts compete in leagues

@SG well we certainly can't be helping small accounts use the site

@SG That is true, and you're reducing my access to leverage. Furthermore, 1.5x leverage is very little when there are 4 orders of magnitude differences in net worth between users. It's still thousands and thousands of dollars of buy-in to sit at the big boy table.

But all of this is a red herring; I was talking about the new interest rate, and took that as a jumping off point to articulate some long standing issues I have with manifold's mindset.

Mana is funny money. No internal change to the mana economy will magically make it not funny money. My stance is that you should focus on the fun part, not the money part. Vaguely copying over economic concepts from the real world is pointless.

Of course I have no issue in general with tuning the mana economy to achieve various goals. I just don't see that the current changes will do anything, other than (1) increasing the pressure to spend for people who like to log in every day and bet on stuff and (2) raising the already insane spending standard for whales even further.

I'm here telling you (as a "medium" user) that competing with others is pointless and unfun. Spending $60 on manifold changed nothing about the fun for me. The way you sell mana is completely orthogonal to the way I enjoy the site.

Maybe some whales can buy fun for $10K; I don't understand the mindset. Good for them, whatever. But there isn't a sustainable business model there. You should sell premium features instead. If anything, the changes are way, way too conservative. You should be thinking about 10x more radical changes.

@VitorBosshard "But there isn't a sustainable business model there." that is the bottom line

@VitorBosshard One question is whether we're trying to have the markets be accurate or whether we want them to be fun. Balancing user power (as suggested by you) makes the markets less accurate because the better predictors lose more mana over time. E.g. one could introduce a per-market maximum bet size like PredictIt has, which would make it more fun for new users, but also much less accurate. The best changes to the site would accomplish both objectives simultaneously.

@JonasVollmer fully agreed. I have tremendous respect for the difficulty of the problem, with multiple goals at odds with each other. But in the end, manifold is a business, and (I think?) these changes are aimed at making money, so that was my framing.

I hope we also agree that the current model isn't exactly great for market accuracy either, when people get a say proportional to how much money they're willing to burn for internet points.

At this point, it seems more realistic to get manifold funded as a non-profit, or changing the business model to sell something other than mana.

@VitorBosshard Manifold does not make money

@ChurlishGambit live service games are a rock solid business model. The trick is to sell things that don't cannibalize the core product, to make the poors feel like they're low status but that they're still playing the same game as everyone else. This isn't just so the poors stick around to pump your DAUs. More critical is that if the thing you're selling can be construed as cheating, then it becomes low status, and then the whales aren't gonna buy it.

Fortnite doesn't sell stronger guns or increased movement speed, they sell emotes and character models.

For Manifold, mana is a core game mechanic. People with 1mio mana are playing a fundamentally different game to people just starting out. But a 1mio balance that you bought with dollars is not the same as a 1mio balance that you earned via trading. Thus, only newbies (and people feeling charitable) buy mana. No amount of macro economic fiddling can change this reality. This is a pretty disastrous business model because your most entrenched users are also the least motivated to spend any money.

To a large degree, live service games are in the business of fashion, of manufacturing desire. And the kind of people who would make a play money website may not necessarily do so well in the fashion industry. Unfortunately, that is where they've found themself.

@cthor Manifold used to have one "whale" (in the sense of spending money to buy mana). If someone invented the mechanic to attract just a dozen or so whales who are willing to spend hundreds to thousands per month on the game, they'd actually be in a really good position.

@cthor I mean I get it—the people running this place, don't. They've never managed to break even, much less profit, in the FOUR YEARS they've been running

5% is still really small. Why trade on a yearly market just for +5%, when I can trade on monthly markets which are equivalent to 12x my expected return on investment. This is only helpful for users who have <0.5% increase of their net worth per month.

If you want to be bold, you should make long-term markets high stakes. If a market resolves in a year, for example, both profits and losses should be doubled at the end.

@ItsMe If I am not mistaken, a 100% interest rate would probably make it even more inconvenient to bet on long markets. Because betting on an annual market would earn you 100%, betting on 12 consecutive monthly markets would earn you 160%, taking advantage of compound interest.

@Emanuele1000 it wouldn't apply to short term markets at all. It would be only long-term markets which are vetted by moderators. Each market would be given an appropriate stakes-factor based on how far away it is.

@ItsMe Adding to the workload of unpaid moderators doesn't seem like a good solution

It's a drop in the bucket. There's only a handful of long-term markets which have decent volume, so they could just focus on amplifying those.

@ItsMe someone's got manager brain

@ItsMe Well, of course, you could also decide not to give interest to those who have not held a position for at least 12 months (for example), but you are only shifting the problem. At that point, it is more advantageous to invest in five annual markets (you can double five times, so multiply by 32) than in a market that lasts five years (where you multiply “only” by five). Then, of course, at that point, I imagine that Manifold is not really interested in favouring markets that last 5-10 years over those that last 1 year.

I like this better than the proposal from a few days ago. I personally will avoid the loans almost all of the time, but the interest seems like a reasonable incentive for long-term betting. Together, they seem like they will reduce how likely users are to end up in the negatives.

@UnconditionalProbability How's it an incentive, though? You don't get the interest until later, & now your loans rack up interest all the time too. It's a restriction of the fake money supply, which is a disincentive to all betting

I mean that the interest paid on positions is an incentive for me to invest in long term markets. (If I was using loans to invest long term, the interest fees on the loans would be a disincentive.)

That said, Ziddletwix's argument below about this disincentivising mana purchases is also convincing.

@UnconditionalProbability Considering that those who lose their prediction do not even receive interest, I do not believe that this will prevent people from going into negative mana.

@UnconditionalProbability Why does it incentivize long-term markets? You get the same interest rate on short-term markets. I think it just encourages people to keep a low mana balance. And it encourages ranked vs unranked.

Wouldn't it be simpler to just pay daily 0.00015 * invested portfolio or is that a different incentive somehow?

@SG as you said about having 2x leverage, can we have that instead of the 1.5x rn

@JeromeHPowell Would you buy mana if he gave you 5x leverage for being on the subscription plan 😂

nah, because then it becomes like pay to win ( @Eliza cool market you made btw)

@Eliza Unironically a good idea. Subscription payments help cover risk of account blowup, people with a subscription probably care more about their account, harder to make a new account and resubscribe if you want to abandon it. Subscription plan should probably give you higher leverage available, though that should probably have both a leverage and an absolute cap...

@EvanDaniel Anyone paying a subscription plan to this site should be checked for gambling addiction lol

@Eliza paying for mana "feels like cheating" but paying for leverage doesn't, as much.

I do not think this is a good system. Daily loans were good, interest payments in any form are a mistake.

(I'm about to go out for the night so will write this very quickly & not edit it at all, oops.)

As background, to repeat from the other thread, there is no meaningful difference between balances & investments. Manifold is infinitely flexible, "ranked and actively traded" is an insufficient barrier, and it is still trivially easy to park your balance as a safe investment somewhere and accrue interest. I wouldn't even call it an abuse of the system—normally Manifold encourages finiancial shenanigans! So with minimal additional effort, I will store my excess balance as an investment, it will not induce me to do more meaningful trading (for whales, the limitation there is their own interest, not mana!), but Manifold will now start paying me 100,000 mana per year. I do not think this is a good way for the site to distribute mana.

While underspecified, I will assume that the motivation here is that the old daily loan system meant capital flowed a little too readily, and it would be good for mana to be a meaningful constraint for more users (encouraging them to buy mana). If that is not the goal, please correct me.

First, if the goal is to encourage people to buy mana, then inflation is a big deal. Interest payments are extremely expensive for the site. If I were a new user looking to buy mana, I'd be kinda confused that the site freely hands out $10,000+ in mana to old whales like me for free, and I'd feel like I was wasting my money paying the suckers $1 per 100M rate. Interest payments are extremely inefficient when it comes to motivating users! (They are subtle, no immediate feedback, etc etc).

Next, I recognize that it is hard to be certain what makes users want to buy mana. If you believe that some users will buy mana once they run out and reach the lower loan cap, then fair I guess. But I do not agree! Daily loans are very well-structured for the broader set of non-power users who might login to Manifold, place some bets, run out of mana, and then they'll come back the next day once a loan is available. Most of these users are not at all willing to spend a dime on mana, and that's unfortunate but inevitable, Manifold just isn't super important to them. But they are nonetheless important because Manifold only works with a sufficiently large userbase. Mana sales generally occur when some user wants a big burst of mana for something. These power users tend to be least sensitive to daily loans.

I don't know what numbers were correct for the old system (maybe it was too generous). But I think it was extremely well-designed for Manifold's current constraints—it very freely gave away small amounts of capital so non-power users had incentive to continually log back in, get a loan, and place some bets. But it was stingy in the sense that it never actually paid out mana, keeping mana valuable and encouraging power users who wanted a bunch of mana all at once to buy it. It may not have been working! Fair. I think the new system will be worse.

To be concrete, about what I expect will happen: many non-power users will very quickly grab all the available leverage offered to them, spend it, reach the cap, and then will run out of mana and become less active. I do not believe that these users will buy mana. OTOH, these changes are extremely expensive for the site, in ways that do not effectively encourage activity. And I do not think they do anything to encourage the occasional power user to want to buy mana (except increase the rate of inflation). More broadly, I think structuring things around interest payments is rooted in a mistaken idea that users are ultra rational financial actors that are highly sensitive to subtle differences in EV, rather than just people who want to log into a game each day and place some trades.

@Ziddletwix also, as an addendum, it’s somewhat important for Kalshi to offer interest payments because dollars can always be parked in places that generate interest! Manifold has no such competitors for mana, and it’s free to let the interest rate arise organically. The site should take advantage of that fact, it’s a good thing! The argument for interest rates was based in some theoretical finance brain idea about encouraging betting on long term markets, and I very strongly disagree that in practice it makes any sense, but I won’t rehash that from other threads.

@Ziddletwix Capital flowing freely, encourages betting in long-term markets. Restricting the flow of the fake money, does the opposite. Unclear why the people running this site don't get that lol

@Ziddletwix I agree with this.

I'd like to add that this new system is just massively more complicated. Power users and whales will be able to navigate it just fine. Right now I have a big balance, but once the time comes I don't feel like hunting down single markets where I could request a loan and I don't understand the structure enough to feel confident to take a general loan. It's too complicated and too much work. to get right and not mess up.

Background: Master degree in STEM and top 25 (as in 25 people, not 25%) daily streak Manifolder.

@Primer Yeah I'm not sure why they think making the site more complicated will grow the userbase lol

@Ziddletwix This was a convincing read for me.

Speaking as a creditor, would it make more sense from a "let the market decide" point of view to let peer to peer loans work their magic? It lets users determine each other's risk factors, and the price of loans would price that in. I have reached out to new users in the past who have staggeringly impressive profit charts, and they don't seem to disappoint when I offer them loans. They get lower rates from me with those credentials, and they get to grow their mana nest egg faster. We are all happy with the transaction.

Manifold has the "monopoly on violence" and can seize mana and force a user to pay back a loan against their will and even into negative balance. If users can send a million mana to a sock puppet account, let them draw leverage on that mana, and then send that lent mana plus the principle back to the main account, then they can create a huge negative balance rather quickly and print a lot of mana.

Users defaulting on loans to their creditors is not a bug. It's a feature. Loans involve risk. They are tests of faith. If you delete the risk factor, you create an impetus for shenanigans, and not the good kind.

I invite others to tell me where my views here are incorrect. Hit me with a truth nuke and change my mind.

My comment doesn't address the topic of making long term bets make sense to traders, but I would like to let the comment stand nonetheless.

@SG so it looks like the 50% limit includes the legacy loans? Very painful!!! Before, I would routinely be at about 120-125%. I am at about 110% currently (just because a bunch of longer-term markets I was invested in resolved at the year end), so at this rate I am out of luck fo at least a few months - effectively nothing changed for me since yesterday - yesterday I could not afford loans because the rate was way too steep, and now I can't access them until I pay down 60% of my net worth, and almost all my investments are in markets that would not resolve for at least 10 more months... Does it have to be 50%?

@AnonUser i intend to raise the amount if everything is working correctly...

@AnonUser Yeah 80% of my bets are on markets that close in about a year, or more. I just sank my remaining balance into betting on the good people of Iran, & then I think I'll just sort of...not do anything on here for a while lol

Manifold Markets is now just a forum to me

@SG Explaining that, and more generally communicating things in advance, would have been helpful!

One thing comes to mind, and I don't know if it's something that has already been considered, but since you can request and repay the loan at any time, wouldn't it be equivalent, and simpler, to manage them directly through the balance?

For example, I have 100M in cash. I bet 1000M on a market, and the interface tells me that 100M will be cash and 900M will be obtained through a loan at 0.03% daily interest. After the bet, I will have a balance of -900M. The negative balance becomes the amount of the loan I requested. I complete the daily mission, earn 25M, automatically repay 25M of the loan, and the balance becomes -875M. A market closes, I collect 1000M in cash, automatically repay the loan, and the balance becomes 125M.

The only difference with the current system is that now there is a daily limit that is different from the total limit, otherwise it seems to me that the two systems are equivalent... The "low" daily limit is the only reason why I should not automatically repay a loan as soon as I can and should not request it just before buying in a market, I believe. But I'm probably missing some details.

@Emanuele1000 With daily loans, the per-market structure kinda made sense, since it meant you could accumulate more leverage on positions held for a long time. That makes sense as a rough heuristic for risk management, and fits with the original goal of helping promote trading in long term markets. I'm not sure it was optimal, but it makes sense. I assume the new system does it this way because it was easy to graft on, and the new system is just a pile of rules that seemed good at the time without any overall plan.

@EvanDaniel "without any overall plan" seems to describe this entire project lol

@ChurlishGambit It's not mad science governance unless there's a mad hypothesis being tested and a mad lab book being kept.

@EvanDaniel Yeah, this is just screwing around until the 2024 investor money runs dry

The maximum loan amount is equal to 50% of a users net worth.

Loans on markets are limited to the lesser of 25% of the position value or 5% of their net worth.

I don't see how you can actually achieve the 50% of NW limit if you can't get more than 25% of the value of any one market. Doesn't this per-market limit create an overall limit of 1.33x leverage, rather than 1.5x limit the account limit would appear to allow?

I don't see how you can actually achieve the 50% of NW limit if you can't get more than 25% of the value of any one market. Doesn't this per-market limit create an overall limit of 1.33x leverage, rather than 1.5x limit the account limit would appear to allow?

@EvanDaniel you can exceed 25% per market limit if the value of your investment goes down after you receive the loan.

@EvanDaniel I believe that 25% is the limit if you apply for a loan directly from the individual market interface, while it is 50% if you apply for a loan from the home page. But I don't think there is any reason to differentiate between the two...

Does interest apply to liquidity pools/AMM shares?

@moobunny Yes!

"Available today" should be zero

Are users able to take the other side of margin loans (providing liquidity @ ~5%) or are these platform 'funds'?

@PaulBenjaminPhotographer not right now. also, i'm not sure you'd want to. even with the fee, i'm sure the house will not be breaking even lol

I think I like this. I will need a little more time to create a fully informed opinion.

rn it won't let me take one out, weird....

@JeromeHPowell ok is it just me or is it glitching for you guys too

@JeromeHPowell I can't get one but I may simply be over-leveraged. Most of my fake money is tied up in long-term markets, so I can't really pay off the loans I do have without liquidating them all...but the only upshot of that would be to have fake money to bet on things I find interesting. So I just have no real incentive to continue using this site

@ChurlishGambit hey, same over here i think, though if you do leave.....a donation to Big J would always be appreciated

@JeromeHPowell My loan button is also in the "claimed" state right now. I didn't claim it today.

@Quroe i think we are all overleveraged

@Quroe are you on the latest version of the site? try refreshing

@JeromeHPowell Oh! I could believe that, then.

@SG On Android app. Closed and reopened. Checked for updates on Google Play Store. I think I should be all set for it to work right?

@SG It says I can get like 4500 today, which is misleading bc i have more leverage than is allowed so it should say zero and state that as the reason, but I do like this, we literally developed. a @ManiFed product today to do this on the website

@Quroe does it look like the screenshot i posted above?

@SG No. I see the same old same old on both Android and Desktop.

@Quroe you need to do a hard refresh. don't use android so not sure how to do that

@SG also the math is wrong, 50% of NAV is 1.5x leverage, but this change is great!!!!

@SG I uninstalled and reinstalled the app.

@SG @ian please we need the Loan Amount on the Trades tab in user profile so we can find markets we have loans on

hahah, we @ManiFed were in the middle of a system for margin loans, dammit

you're good @SG

@JeromeHPowell 1.5x leverage cap is pretty small for anyone with diversified positions. You'll still find plenty of takers.

@EvanDaniel yea, still a good feature for us I think!!!!

Not paying interest on my AMM stake is a crime unless you're going to fix the UI/UX to get people to use limit orders significantly more.

@Eliza lmfao

• You will earn interest immediately upon selling positions.

So I can buy a position at 1%, sell it the day before the market closes, and earn 5%, which is basically +500%? I'm missing something here...

@Emanuele1000 no what its saying is if you buy 100 mana in yes at 1% wait a year and then sell (still at 1%) then you would get back 105 mana.

@Kfredric > E.g. if you bet 100 mana on YES at 50% and the market resolves YES one year later, you will receive a payout of 210 mana (200 mana in winnings plus 10 mana in interest). NO bettors will not earn any interest. Note the change from previous behavior.

So shouldn't it be 205 here?

EDIT: Ah, I see, it's a return on the value of the shares AFTER the resolution, how much are they worth at 100%. I don't understand what the incentive is, though. What is the point of this change?

But aside from the objectives, wasn't it easier to simply add the shares? I buy 100 shares. Yes, after a year I will have 105 shares. Yes. It also makes it more intuitive to sell and/or buy back to reduce exposure.

@Emanuele1000 im not 100% sure but what I think they are saying is then your yes shares earn interest. So if you buy 100 mana on yes at 1% to win 10000. The after a year its 105 mana on yes to win 10500.

It's an improvement. I still maintain that getting rid of free daily loans in all cases goes counter to Manifold's interests, for reasons previously discussed that you haven't addressed.

The maximum loan amount is equal to 50% of a users net worth. In other words, the system allows leverage of 2x.

Surely that means 1.5x leverage. I can have 1k net worth, and 1.5k of risk exposure, that's 1.5x?

No interest if you're wrong? Why is that?

@ChurlishGambit That feels like a weird incentive. You would not want to correct a slightly overconfident market.

@ChurlishGambit another way of looking at these changes is that we pay interest on shares rather than invested mana. if the shares that you hold are worthless, the interest you accrue on them will also be worthless

@Eliza Yeah exactly. And it's not really interest, this way. It's a bonus for being considered correct in a year

@SG Yeah but it's like... the bad version of that. Incentivizing holding old positions rather than selling them at profit or cutting losses is a poor choice.

@SG I guess so? But I'm not taking any new loans anyway, I lose no matter what in these new systems. I think you're gonna see activity continue to decline, by taking away a major daily incentive to betting

@EvanDaniel Wait, maybe I'm misunderstanding how the interest earned immediately on selling positions works. This might not be as bad as I thought, maybe it's just wonky.

@Eliza didn't you yourself ask for interest to be paid on shares?

In my opinion, it is much more intuitive if new shares are given (and I believe it also improves market liquidity, but I'm not sure about the mathematical model), so it is clear why you can collect interest in the event of a sale and earn nothing if you get your prediction wrong.

Although I don't understand the purpose of this change, we are not betting money, the markets should not challenge bank deposits...

@Emanuele1000 Paying in shares rather than cash probably also fixes the exploit.

@EvanDaniel you are confused, my understanding is that you can sell at any time and get the current value of the shares, with interest proportional to the length of time you held the position.

This is much better, though I still miss the interest free loans that kind of felt like an incentive to come back every day.

@Velaris They are that, and by combining them with the normal daily streak bonuses and other quests, Manifold can provide a larger such incentive with less mana printing. The combination of the two is a cheaper way to provide daily liquidity to free to play users and small accounts than the daily bonuses alone are, for any fixed daily incentive target. The free daily loans should come back.

This argument is entirely compatible with a moderate cap on loan size (per day, total as % of account, total in absolute terms, a mix, or other).

@Velaris I requested the loan every day, logging in and checking my notifications, even though I had liquidity equal to about 100 times what I was getting from the loan. It was an incentive to come back, in addition to the daily streak.

@EvanDaniel Also I expect that the best cap, from Manifold's perspective, on the free daily loans is higher than what's required just to get a multiplier going on the daily quest bonuses. That need would be served even if you just had a daily cap for the free loans of like M50 or M100 or something. But I suspect that Manifold is well served by bumping up the velocity of money a bit beyond that; it should increase trade volume and open interest. At some point Manifold is taking on more risk than they want, which is why a cap starts to make sense again. Multiple different things push the ideal cap in different directions, I'm not sure what the math would actually say.