8 ms·
Of the wild theories I've heard to explain the Medallion Fund, my favorite is the "money wormhole." I have no finance qualifications whatsoever -- I just stick
by jjoonathan 7y ago
Of the wild theories I've heard to explain the Medallion Fund, my favorite is the "money wormhole." I have no finance qualifications whatsoever -- I just stick my money in index funds -- but I love a good conspiracy theory, so here goes.
The idea is that you have two theoretically unrelated funds that take complementary positions with uneven odds. One sacrifices performance for the other, effectively transmitting money. The trick, of course, is doing this in a manner so that you have plausible deniability with regulators and, more importantly, so that other market players don't grab the money in flight, because that's what an efficient market would do. Still, consistently creating and hiding complementary pairs of good/bad opportunities is a much easier problem than consistently beating the market.
OK, so you've got a way to transmit money from a "sucker fund" to a "winner's fund" with plausible deniability, what do you do with this ability? In reality you'd probably have many "sucker funds," rotate the one that gets sucked, and limit the extent to which it gets sucked, but still, nobody's going to just buy the sucker fund, not for long, anyway. That's how efficient markets work. Well, one major, well-known, boring inefficiency is the principal agent problem: trusting others to invest your money. They can't just steal it, because that's illegal. They can't just invest in a private fund that they just happen to oversee and which just happens to pay them an enormous salary, because that's illegal. However, if they were to invest in one side of a money wormhole, which is designed to look like a perfectly reasonable investment on its own, in exchange for a cut of the proceeds from the other side of the money wormhole, who's to say the theft even happened? Such an accusation would need to untangle the coordination behavior of the money wormhole to make its case, and those inner workings could be made very convoluted indeed, hidden deep inside complementary pseudorandom behavior.
So, in short, the conspiracy theory says it's a heist -- but one that's well hidden under the veil of statistics and plausible deniability. It provides investment managers looking to monetize the confidence placed in them a way to make that happen without getting caught.
It's probably a dumb theory, but it's enough to entertain someone who knows as little about finance as I do :)
- fyp 7y agoIt's an interesting idea. You can make two funds sum up to a constant by taking opposite and equal positions. But I am not seeing how you can control the direction the money moves. For example if you knew fund A will gain $x and fund B will lose $x, shouldn't you simply not make the trade in fund B?
- papln 7y agoFor the pros, its easy enough to structure a complicated transaction that analytically simplifies to "A loses $X + %risk, B gains + %risk", since you can create instruments with layers of elements like "A buys an derivsative on Foo, and B sells an derivative on Foo", where B controls Foo. The rest of the market is happy to buy from B and sell to A netting a small profit, even if B contols Foo, because they aren't exposeed to the risk of the derivative. when it comes time to settle, A pays a big loss to C, C pays slightly smaller loss to B, netting a motivating profit, and B gets a big gain.
- ThrustVectoring 7y agoYou decide which fund has which position after it books profits and losses.
- SirLJ 7y agoThe "only" problem with your theory is that the "sucker funds" have to lose 100 billion, which won't go unnoticed by the "suckers"
- ceejayoz 7y agoDo they have to lose $100B, or just underperform? People have accepted 2% fees on managed funds despite underperforming an index fund for decades.
- SirLJ 7y ago100B is what Medallion made during those years, the money have to come from somewhere...
- femto113 7y agoIf you layer credit into the wormhole it seems like you don't even need "losses" in the sucker funds. In essence Medallion establishes that it is "low risk", it is then able to borrow money more cheaply than other parties. Through the wormhole it can then indirectly "reloan" that money to higher risk sucker funds at higher rates. The sucker fund can still win, but it pays Medallion back more in interest than Medallion pays to the original source of money. Medallion's capacity is constrained by the amount borrowed by the sucker funds and the margin between what it can borrow vs lend at.
- SirLJ 7y agoI think we already went to deep into the hole on this :-)
- thoughtstheseus 7y ago100% agree. A highly overcollateraled position (not drawn capital so it doesn't show up) can be used to magnify returns. If you're playing in marketable assets you can even have the lender hold the assets themselves while you trade.
- 7y ago
- SamBam 7y agoI'm not sure I understand your suggestion. So there are (simplistically) two funds, Sucker and Winner. They take complementary positions, and, whichever one wins gets transferred to Winner. So far so good. Winner is winning every bet, and Sucker is losing every bet. Now you need people to put money into Sucker, because it has to come out the other end into Winner. How do you convince them to do that? By giving them a cut of Winner's profit? But that's where things don't add up. How can the Sucker investors and the Winner investors both be making a profit from Winner's funds, if the net money going in and out of the combined funds is equal?
- Mvandenbergh 7y agoPeople with illegally earned money put their money into a series of short lived sucker funds which exist to funnel money invisibly into the winner fund where it reappears as legitimate investment returns. The dumb version of this with very simple mirror trades which were easy to track has been seen "in the wild" in Russia as carried out by people at DB. BTW I don't really buy this theory but that's how it would work.
- james-imitative 7y agoThis has been happening in Peru, Uganda, Bolivia and Grenada.
- ThrustVectoring 7y agoIt's simpler than you think. You take a pair of offsetting trades, and later you re-write history to book the losing trade on Sucker and the winning trade on Winner.
- SamBam 7y agoNo, I got that, as I stated. The issue is that half your trades are losing money, so where is that money coming from.
- texasbigdata 7y ago
- bitcurious 7y agoThis complements another wild theory I've heard, which is that the Medallion Fund exists to launder money from corrupt states. Say you have a $200B state pension fund - you invest a portion in the loser fund, earn a portion back as personal wealth on the winner side.
- dankohn1 7y agoThis is deeply insightful.
- james-imitative 7y agoYes. I would endorse this comment.
- dereg 7y agoYour theory is interesting but it's just speculative fodder. Getting capacity in something like Medallion is pretty much impossible. The lowly head of a state pension fund would never have access to such capacity. Doing this sort of thing destroys the reputation of hedge funds and is not a long-term positive expected value decision. (See Bluecrest: https://dealbook.nytimes.com/2014/05/29/fund-within-a-fund-creates-a-conflict/ https://dealbook.nytimes.com/2014/05/29/fund-within-a-fund-c...) While skepticism of Medallion is warranted, we're not doing ourselves favors by explaining away its success with activity that would have much more likely killed the firm vs sustained it.
- ajay-d 7y agoThat makes a lot of sense. And also explains why the the other funds aren't as 'stellar'. I mean all these quants are in the same company, how is one fund perfect, and the others lackluster.
- 1024core 7y agoI believe (not that I know more than anyone else) it's because their own money is invested in the Medallion fund (and not the other funds).
- auntienomen 7y agoThis is wrong. The people at Renaissance have a lot of their money invested in the other two funds. This is one of the main reasons they built the other two funds; they needed a way of managing money that didn't fit in Medallion. Simons, for example, has a ~$21B fortune, and I doubt he gets more than 1/3rd of Medallion's $10B capacity. He's probably got more money in RIEF than in Medallion.
- RookyNumbas 7y agoThis would only explain the performance for the last 10 years. There were no "sucker" funds for the first 20.
- throwaway2048 7y agoHow do you know?
- auntienomen 7y agoThe "sucker funds" would have had to be among the largest in the industry, with lots of external investors to provide the capital. You really think something like that would remain completely secret?
- JackFr 7y agoThat makes perfect sense. Renaissance has always been immune to the claims that it's a Ponzi scheme, because Medallion is closed. But if you used the wormhole techniques to move money from the open to the closed, and voila, unbelievably high returns!
- james-imitative 7y ago> Such an accusation would need to untangle the coordination behavior of the money wormhole to make its case, and those inner workings could be made very convoluted indeed, hidden deep inside complementary pseudorandom behavior. In such a case, it may be (fscking) helpful for the courts of law to go after the people who _didn't clearly break_ the law in the first place. That way, the shady Machiavellians get caught first and the chumps can be mopped up afterwards. Deep as anythink.
- perl4ever 7y ago"so you've got a way to transmit money from a "sucker fund" to a "winner's fund" with plausible deniability" I've been thinking about this actually for quite a while, on a small-time scale. Consider the following scenario: 1. You have a fair amount of money in a Roth IRA, and roughly the same amount of money in a taxable account. 2. The limit on contributions per year is like $6K, and even if you put money in a 401K and then convert whenever possible, that's limited to $19K/yr I think. How can you work around this? 3. Let's assume you're approved for options in your taxable account; and my understanding is that you can buy options in a Roth account - you just can't write them. 4. So, what if you create a synthetic long position by purchasing calls on a broad market fund in the Roth, and selling cash-secured puts on the same fund in the taxable account. And maybe buy some short term corporate debt fund in the Roth to make up for the missing dividends. Not sure if you could also earn some interest on the taxable side. If the stock market goes up in the long run, then you have most of your gains tax-free in the Roth, and minimize your taxable gains. If the stock market goes down for a while, you have most of your losses in the taxable account, and minimal losses in the Roth. 5. I think there are some practical problems, but Matt Levine could probably write about them much more engagingly than me, particularly after someone tries this and gets in trouble for it.
- thoughtstheseus 7y agoThis is called asset location. Structure where your investments are located to maximize returns.
- DrAwdeOccarim 7y agoRight, I didn't think what he was describing was all that risque. What I'm trying to figure out is if there is a way to sell a cash-secured put on a way OTM option from the IRA then buy that put in the taxable account. Then let it expire worthless. Could that transfer money into the IRA and nab a tax loss deduction in the taxable account?
- scottjg 7y agoperhaps it's possible through enough layers of indirection to create plausible deniability... but generally this type of transaction would be considered self-dealing, which is illegal.