Hacker News new | past | comments | ask | show | jobs | submit login
RenTech Created the Ultimate, Tax-Free IRA Account for Employees (bloomberg.com)
85 points by throwawaymath 43 days ago | hide | past | web | favorite | 123 comments

The backstory, for those unfamiliar with Renaissance: they have more than one fund available. Their highest performing fund, Medallion, is the one which made them famous in the 90s. In the early 2000s it was closed to outside investment (likely due to capacity constraint for the strategies it uses). It's now exclusively an investment vehicle for employees and their families, somewhat like a large proprietary trading firm.

The really interesting thing here is that Renaissance received clearance to roll employee IRAs into the Medallion fund. That leverages significant tax advantages on a fund which already provides significant returns for its employees.

Do new employees get to participate in Medallion?

Yes, but they're restricted in how much they can contribute.

> The firm initially terminated its 401(k) plan for employees in 2010, a step that permitted them to roll the savings into traditional IRAs. Then, employees took advantage of a rule change that year allowing affluent Americans to convert their traditional IRAs into Roths. The following year, Renaissance applied for clearance from the U.S. Labor Department for employees to invest the accounts in Medallion, which the agency granted and made effective in January 2012.

This is entirely convoluted. A Roth 401K would have achieved the same thing, 10X higher annual contribution limits, and allows cheap liquidity from borrowing 50% of it at 2 points above fed funds rate. The conversion opportunity was nice but has very little to do with what has happened since 2010 for anyone that invested later.

Many Reg D offerings also allow for 25-30% of the fund to be held by tax-deferred accounts. US Department of Labor wasn't necessary here.

> In turn, the IRA money -- held by about 250 employees -- grew to more than 4 percent of Medallion’s gross assets from about 1 percent five years earlier.

So what's the US Department of Labor for again? I'm missing something..... this must come down to the fund's structure. In any case it is nice they pulled it off.

My approach to letting 401K/IRA/Tax-Deferred accounts make a ton of money is to give them a separate share class. That share class gets some outside attention and liquidity preferences sometimes. It was inspired by Bain Capital's approach, when that was in the news during Romney's presidential run and it became 'controversial' regarding how they circumvented tax liability. I didn't confirm thats what Bain Capital actually did but thats what I came up with and its passed.

> The conversion opportunity was nice but has very little to do with what has happened

You can’t invest in a Roth account if you earn more than a fairly low level. You can, however, convert IRA assets into Roth assets at any income level provided you pay the tax on conversion.

Disclaimer: I am not a CPA. This is not tax advice.

There's no income limit on Roth 401ks as far as I know, where did you hear that there is one?

>You can, however, convert IRA assets into Roth assets at any income level provided you pay the tax on conversion.

You seem to be saying you can convert traditional IRA money into a Roth IRA. That's true, but how do you get traditional IRA money? There's an income limit when you make traditional IRA contributions if your employer has a 401k:


That's a tax deduction limit not an IRA contribution limit. You make an after-tax IRA contribution (up to the IRA contribution limit of $6K/year in 2019) not subject to any income limit and then convert to Roth IRA.

If your firm's 401k plan allows it, you can contribute up to 32k post tax (it's not deductible) and then immediately roll over to a Roth 401k (google for "mega-backdoor Roth"). People don't realize that the IRA and 401k sections of the tax code are completely different, and they both have Roth sections, and the limits in one don't apply to the other.

Yeah, but that's not really "traditional". And if done fast enough, the "tax on conversion" will be essentially nothing.

The Medallion fund has been restricted mainly to RenTech employees since 2005 as the firm took steps to keep its size around $10 billion. The fund has historically averaged annualized returns approaching 80 percent before fees, but such gains can slump when it gets too big. Even employees face annual investment limits, and Medallion also typically distributes its profits every six months instead of reinvesting the gains.

How is that even possible? even warren buffett only averaged around 15-20 a year

Return on capital isn't a super meaningful metric for capacity constrained trades. If a fund earns 80% returns, but has no means to compound the resulting profits through the same mechanism, whoever receives them naturally puts them into something with worse returns, so their wealth still grows slowly over time.

I think Medallion is somewhere between HFT and stat arb, probably a mix of multiple strategies along those time frames. The faster you trade independent opportunities, the more you recycle capital, and prime brokers extend tons of leverage. When I worked in HFT, our profits were bound by other factors way before cash, and my desk's ROC was far higher than this when doing well (even when doing poorly, ROC was quite high. It was the expenses of finding those returns that killed us).

If all the money is employees', Medallion is basically just a prop firm. The employees paid out of the fund are essentially partners/owners, and the rest earn discretionary payouts of management/performance fees.

Even within the fund structure, most profit is return on labor, not capital. I'm sure if you compare margins paid to partners in professional services firms like law or consulting vs. typical publicly traded companies, they're also far higher, but Cravath, McKinsey, etc. won't let you buy in as a passive investor; you have to work for it.

ETA: If you're wondering how it's possible to earn 80% returns, or even more: there are myriad tiny inefficiencies you can trade on given the right research and infrastructure. I'm sure 80% is simply the point at which Medallion makes the optimal $ per year relative to risk. They could probably throttle back, make less $ on smaller capital, but far higher percentage return, if they wanted.

Can you give an example of a strategy that dam it's course and reached a point of no longer being viable?

You might notice that the prices of trading A for B, then B for C, then C for A results in more A than you began with if done instantaneously.

But you can’t do that for a trillion dollars, because the “price” of an asset is just the best offer to buy or sell, and those are capped at the quantity of the best offer. Once you exhaust those, the arbitrage might not exist anymore because the remaining best offers will be worse. Say the market has capacity to actively exploit arbitrage for $10,000. You will earn the same amount of dollars if you have $1,000,000 or $10,000 if you exploit it to the max, but your return will be 100x better if your base is $10,000.

i've been thinking about your comment for about 12 hours now (over night after reading it last night). i don't quite understand the force of it.

>If a fund earns 80% returns, but has no means to compound the resulting profits through the same mechanism, whoever receives them naturally puts them into something with worse returns, so their wealth still grows slowly over time.

i don't see why this matters at that AUM. treat it as a fixed annuity for the fund members and it's still fantastically successful (if you know of a savings account that i can safely withdraw 80% of 10b from every year for 20 years please let me know).

>Even within the fund structure, most profit is return on labor, not capital.

i think this is just a matter of "levels of abstraction". if i invest in a mutual fund that has portfolio managers and analysts are my returns thereby ROC or returns on labor? obviously according to GAAP they're returns on capital but inside the mutual fund there are people laboring away. if i in-house that mutual fund along with all of that research infrastructure why does it suddenly become return on labor? it's the same economic activity.

Most of what we know about Rentech's strategies comes from Senate hearings regarding the sketchy things they do. The common theme is that they find strategies with fairly low returns (~3%), and very low risk. They then use massive (and illegal) amounts of leverage, as high as 16x, to turn those low risk, low return strategies into outstanding returns. James Simons and Robert Mercer are among the largest donors ever to both political parties, so it isn't much of a surprise that these hearings haven't really gone anywhere.

Of course it isn't just that simple. Leveraging a strategy should increase it's downside risk along with it's returns. But somehow Rentech does it while never actually experiencing that downside risk. That's their mathematical genius. And the fact that they employ so many brilliant mathematical minds leads me to believe that part may be real.

The latest Senate hearing link: https://www.hsgac.senate.gov/imo/media/doc/REPORT-Abuse%20of...

> They then use massive (and illegal) amounts of leverage

Why do you think their use of leverage is illegal?

Not me, The IRA thinks it's illegal. There are limits on how much leverage you are legally allowed to use, and Renaissance has made a name for itself by trying to bypass those laws in every tricky way possible.

> There are limits on how much leverage you are legally allowed to use

No, there aren’t. Reg T limits initial leverage for retail investors in equity securities to 2:1 and FINRA rules maintainence leverage to 4:1 [1]. There are no similar broad-market rules for institutions. (Retail investors, likewise, can leverage much more for FX and—more commonly—real estate.)

[1] https://www.sec.gov/reportspubs/investor-publications/invest...

They weren't in violation of Reg T. They were in violation of Reg U and Reg X. You are right that there is no law saying you can't use more than x leverage, but there are absolutely regulations governing loans in margin accounts and using securities as collateral for leverage. And The government believes Rentech was in violation of them.

> there are absolutely regulations governing loans in margin accounts and using securities as collateral for leverage

No, there are not for institutional investors. Rentech fell into an interesting dispute (tigger with Deutsche Bank) with the IRS regarding long-term capital gains. If I want to lever my institution 10,000:1, and can find a lender who will lend against it, there is no law prohibiting me from doing so.

Disclaimer: I am not a lawyer. This is not legal advice.

But there are laws preventing a broker from lending you that money. The hearing cites the SEC net capital rules which Rentech and DB worked together to bypass. There are pages dedicated to how RenTech DB and Barclays knowingly and purposefully circumvented rules with some changing of terminology. Starts on page 79 [0].

I definitely should have been more clear with my original statements. Maybe something more along the lines of: at rentech's scale, using a margin account at a prime broker, you cannot leverage your firm 16x.

[0] https://www.hsgac.senate.gov/imo/media/doc/REPORT-Abuse%20of...

I understand where you’re coming from. Let me, too, be more clear.

The Senate report is crap. Yes, Reg T and FINRA rules limit the loans B-Ds can provide clients. But leverage, for Reg T’s purposes, is constrained to lending. I can buy a 3x leveraged ETF [1] as a retail trader without violating Reg T. (With options, I could easily increase that leverage without borrowing.) All of this is not only permitted, but common.

Reg T does not exist to protect investors. It exists to keep broker-dealers from going bust from dud margin loans. (And thereby prompting a systemic crisis.)

The leverage RenTech took is in the non-lending and non-systemic (legal) category. The exposure that most closely puts RenTech in the lending bucket is the exposure Deutsche Bank carried on its balance sheet for tax purposes. (This tax avoidance was the core of the scandal.)

Long story short, unless you’re a politician, it doesn’t make sense to talk about RenTech’s illegal leverage. Lots of other market participants, by regular practice, are levered far more.

[1] https://etfdb.com/etf/TQQQ/

If a senate report causes issues for you, I don't think it's exactly crap. But I understand your point :) It also keeps mentioning Reg T which I agree is not relevant here. I was wrong to say that RenTech used illegal amounts of leverage, thanks for explaining that. I'm still not convinced it's wrong to say that RenTech worked with BDs to bypass rules that would have stopped them from using the leverage they did in the manner they did.

Was the leverage in the non-lending category though? Isn't that like saying their gains were long-term? That's the whole point of this, no? They used some different terminology to reclassify loans and taxes, in order to use more leverage and pay less tax than they normally would have.

I guess another way to put it: would there have been a way for RenTech to hold the same portfolio, using the same leverage, with the same payout characteristics, that no government agency would have issues with? Maybe the answer is yes, but I doubt it.

> would there have been a way for RenTech to hold the same portfolio, using the same leverage, with the same payout characteristics, that no government agency would have issues with?

Yes, quite easily. In fact, highly-leveraged portfolios like the one RenTech held are an essential feature of market making, which was historically done using banks’ balance sheets. RenTech’s shenanigans were around tax. Everything else is commentary.

(On the Senate report, the whole thing isn’t crap. But that section is crap as in it’s written for political purposes and has limited bearing with respect to the law.)

And most peoples mortgages are > 2:1 which is a loy higher than almost all funds/investment companies gearing.

the subject of this Senate hearing was widely written about wrt rentech back in 2014.

This is more about what rentech does after they've made all their money trading... To avoid capital gains taxes. Sure, it's just another way to make more money, but it's not their core strategy.

How does the leverage work considering it comes out of the returns.

If I invest $1B and borrow another $9B, my base returns need to exceed the borrowing costs of the $9B just to break even.

It was never about returns exceeding borrowing costs. Rentech is probably the greatest money manager of all time. They can make returns that exceed the borrow costs.

It was always about the amount of borrow available, and the tax paid on the gains. The (illegal) strategies they used allowed for much larger borrow amounts, as well as only paying long term capital gains, when in reality there were millions of trades.

The simple setup: Rentech pays the bank $1B for a call option. The call option is on a $10B pot of money. The bank then hires Rentech to invest the pot of money for them. After a year or two, Rentech then exercises it's call option, which gives it everything in the pot except the bank's $9B plus a fee. The banks loved this fee. And since this was a year long call option, all gains are taxed as long term. They avoided $6B in taxes doing this.

For a modern hedge fund, leverage doesn’t work that way... e.g. if I buy $1000 of AAPL (Apple) and (short) sell $1000 of MSFT (Microsoft), what is my “capital” and what is my “leverage”? My net capital use is $0 - I got $1000 selling MSFT and spent it again buying AAPL... in the end, it comes to a combination of margin required by the broker(s) - i.e. what they estimate is the black swan scenario where AAPL goes down, MSFT goes up, you start losing money on both trades and they can’t close the position fast enough - ultimately it depends on volatility but can be netted across many trades, and the overall volatility / risk of your strategy (i.e. how much you’re potentially willing & able to lose in a year).

When you short you need to post capital as part of the repo trade. You also need to borrow the shares which typically costs more than the interest you're earning on the posted capital.

So no, your net capital use is not $0 in your example

That depends on your borrowing costs. I doubt they'll be paying margin rates lesser mortals pay. ultimately they are screwing their lender because the lender is taking all the risk here without (seemingly) adequate compensation.

Their lenders likely understand the risks pretty well. Their lenders are going to be large banks that have entire departments dedicated to modeling risks like these, and making sure that the risks don't exceed certain limits on these accounts. Nobody's getting screwed, except arguably the government out of some taxes here.

>Their lenders likely understand the risks pretty well. Their lenders are going to be large banks that have entire departments dedicated to modeling risks like these, and making sure that the risks don't exceed certain limits on these accounts.

Three words: Mortgage Backed Securities. Large banks with entire risk modelling departments have made multi-billion-dollar errors in recent memory.

Three letters: TSA

I have a complete lack of faith that any institution whose day-to-day experience completely differs from their catastrophic risk scenario is ever truly prepared.

The personal and organizational strain of maintaining constant vigilance against an invisible enemy is simply too high.

People get lazy, complacent, and think they're more prepared than they are. That's just human nature.

This is absolutely true. Due mainly to wishful thinking as well as some bizarrely bad statistical assumptions, the risk of something like 2008 was simply not forseen. There's absolutely no reason to think that other unforeseen crises couldn't arise in the future.

You're right. But there's very good reasons to think that they won't come from levered quant strategies like Renaissance's. They're extremely diversified and tend to be market neutral.

Yes, you can name one time that that happened. But do you understand the types of strategies RenTech is using? They're diversified across thousands of equities on the long and short side. They're market neutral. Taking on amounts of leverage that seem crazy to you is actually very reasonable in this setting, and it's very easy to show that this is reasonably safe.

Nobody outside of RenTech knows what trading strategies RenTech are using, because they're extremely secretive. There is a strong suspicion that they are generating very high returns by passing on concealed risks to other, less sophisticated players - in the case of RenTech, literally anyone else in the market counts as less sophisticated.

That's really not true. Everyone knows what they're doing. They're doing statistical arbitrage. There's many firms that do it, Two Sigma, Citadel, etc.. The only thing secretive about it is the exact details of each strategy. Renaissance, for instance, pioneered techniques like using satellite imagery to track sales. One of their early strategies noticed that the market tended to go up on days when the weather was nice in NYC. Things like this are what they're doing, but on a grand scale and with a lot of very very smart people working on them.

While it is possible that they're making their profits by shifting risk onto unsophisticated players, it's certainly not necessary for them to do that to make money. They have the smartest people in the world working for them.

Tbond returns about 3%. Not sure how they are going to get money less than that.

It was 1.9% during the time of some of these trades. But yes, you're right, I should have said 3% over the risk free rate

It is definitely possible, though it becomes much more difficult at scale. Asymmetries in financial markets are pervasive, and it is technically quite difficult to methodically explore that space analytically. Going deep mathematically into modeling these things is beyond the ken of most funds. For the most part, they are biased to search for patterns in parts of the phase space where it is relatively easy analytically. Motivated individuals that want to search the technically more difficult parts of the phase space still find plenty of gold nuggets because so few people look there or the scale of the opportunity is too small for other funds. RenTech is just really good at systematically mining parts of the phase space that are mathematically out of reach of other funds.

While 80% is extreme, 15-20% is not and I know many people that can do that consistently. Anecdotally, one strategy I developed many years ago that was designed to be easily applied to retirement accounts (so that close friends and family could, and do, use it) has returned an average of 20% annually through boom and bust, outperforming the S&P500 every single year. It didn't require being super-clever, just observant and analytical.

I think we tend to overstate the difficulty of getting a good return to scare people away from pursuing naive and/or stupid strategies in a futile effort to find easy returns.

Particularly when these analytical advantages may compound over time rather than dissipate. For instance, suppose a firm was exceptionally good at pairing world-class mathematicians with a team of mathematically-exceptional programmers who could reconceptualize the math wizard's insights as financial time series models and steer them towards areas of applicability. Now, when the next world-class mathematician comes to the office on a recruiting trip, he recognizes a higher expected value & better colleagues, joins, and the engine gets more repetitions with which to improve itself.

Hey, it's me, your close friend :P

Berkshire Hathaway has two orders of magnitude more assets under management than the Medallion Fund.

Even smaller funds only average at best 10-15%. 80% is just nuts and to do so with no down years, minimal volatility. I think it's more than just market timing

It's the law of large numbers. Virtu's infamous net trading profit histogram showed the firm had one losing day in over 3 years: https://www.zerohedge.com/news/2014-03-10/holy-grail-trading... (1)

Again I've seen the same just running a single HFT desk within a larger firm. The only time we ever lost money was from rare technology errors. Trading equities, even if one position spikes 5-10% bad on news, you will still make money, because it's just one little position out of the thousands of tickers you trade. Even guys making far fewer bets in asset classes like FX only ever lost on extreme dislocations like the Euro/Swiss unpeg.

If you make a large number of bets, even with just a tiny statistical edge, you will be consistently profitable. RenTech probably isn't profitable every day, but I bet over a year they make at least as many bets as someone like Virtu makes in a day, so it's not surprising that they never have a down year, provided they have the edge.

1: Now does this mean Virtu the business made a profit above cost every day? Probably not. But it does show that consistent trading profits are achievable.

I'm not sure RenTech and Virtu are comparable though (although I'll admit I don't work in the industry, so correct me if I'm wrong), but based on the article from Matt Levine on his Bloomberg column "Why Do High Frequency Traders Never Lose Money?" [1]:

> Imagine how suspicious it would be if, for instance, your local supermarket made money on every carton of milk that it sold. That just seems too good to be true, doesn't it? How can they know the price of milk before you do? Shouldn't they be losing money on half of their milk, and making it on the other half, so that things balance out? Doesn't the fact that they always make money suggest that they're ripping you off? [...] That is, Virtu (like Goldman) is selling a product, and that product is liquidity, and it charges for that product. High-frequency trading firms are in the business of acting as middlemen, providing a valuable service by letting buyers and sellers trade as soon as they want to, rather than waiting for fundamental sellers/buyers to come in on the other side of the market.

As I understand it, RenTech is taking an investment position which is why the returns are remarkable whereas virtu is (as Levine puts it), selling liquidity.

[1] https://www.bloomberg.com/opinion/articles/2014-03-20/why-do...

RenTech doesn't publish much so I'm merely conjecturing based on their volumes and other information that's in legal filings. There are a lot of ways to provide liquidity and fair pricing over diverse time horizons. It doesn't really matter how they do it. If your bets are independent and you have a statistically significant edge, you are basically guaranteed to make money with proper bankroll management.

From what Virtu's published, they make two-way markets, and once filled they scalp a tick, arbitrage in another product, or cross if the market becomes weak. Maybe RenTech does something like buy underpriced oil producers whose prices haven't moved up after bellwether stocks in the industry like XOM and CVX have, then sell once the spread between them converges. I'm sure both firms have loads of different tactics. The key thing is that they're mildly better than chance.

Virtu makes money because they take advantage of a structural setup in the market: they trade retail flow from the likes of Schwab and Robinhood.

This is vastly different than making money in the markets just by connecting to an execution venue and receiving market data.

In the former you have a clear edge over the market because you receive dumb uninformed flow. In the latter example you are left to fend for yourself and challenge the EMH purely through your math skills and insights.

Most smaller funds only average that much, yes. But there are outliers.

Without knowing more, my default assumption would be fraud with returns driven by ponzi scheme.

I mean, it's kind of ridiculous to just make blanket statements like that. This is a hedge fund that has a long history of good performance. The whole point of a ponzi scheme is that you have to keep getting more money from outside investors to pay the original investors. Rentec does the exact opposite. They refuse outside money and keep their employees from investing too much in it.

With that out of the way, obviously no one except those who work for them knows what they actually do. What we do know for sure if that they are highly levered, use an entirely systematic/algorithmic investment process, and do a shit load of trading. So even if they don't earn a lot on each trade, they do so many that are just barely profitable and highly levered that it ends up being a lot of money.

Ok sure, but 80 (eighty) percent returns? That's... inconceivable.

It’s all relative. Renaissance is famous because they do it on (low) double digit billions. On the other hand, 80% at a proprietary trading firm - where you might be working with double digit millions - isn’t spectacular. I can tell you that several small outfits in Chicago and the Bay Area return well over 200% consistently. But they are severely capacity constrained and basically disburse profits to the partners and traders each year. Their capital is only in the eight figure range.

Can you elaborate on what drives the capacity constraint for someone not in the industry?

I don't think its inconceivable, but comparing it to a hedge fund makes it look pretty crazy. First note that they look more like a prop fund, in which typically returns are labor-constrained rather than cash-constrained. Annual returns there aren't really comparable to investing, as they're more performing services (like arbitrage) for the market than putting in money. Second, the fact that being able to invest is a perk of employment means that part of their pay structure is via Medallion returns. Income that other firms would have to subtract from returns because it is spent on payroll still counts for Rentech even though said returns are going to employees.

Why? You don't need to do a ton of constant trades. While my portfolio is fairly small, I make probably half a dozen year and sometimes less or more but in total it averages to 500-1000 percent a year

A good starting point..if wall Street is consistently shitting on a stock, it's probably something to check out and do some research on.

their gross return is probably just a couple percent but then they leverage the crap out of it, which they can easily do since the strategy is designed to have very low volatility.

But leverage magnifies volatility to the extent that it increase gains.

Correct. There's no free lunch when it comes to volatility.

I disagree. But more importantly, it doesnt matter.

We cant invest and dont have enough information to assess whether it is a ponzi scheme or not.

Its possible that this one fund has generated excessive outperformance with low volatility, but we should start by assuming it didnt, then prove otherwise

They reportedly don't reinvest returns, which means that the 80%/year is paid out. A ponzi scheme can't return more money than is put in; Medallion does so every 15 months.

There was a discussion of this company the other day and the suggestion that unlike the other outperforming hedge funds that turned out to just be insider trading, this one is real and they are just that smart. I am skeptical; they do not have a monopoly on brains. What they do have is some shadiness in their history:


I am sure there is more like this waiting to be uncovered.

I know a few people who work at RenTech. I would be very surprised if the operation turned out to be fraudulent. It certainly could be, but I very much doubt it.

If you have your ear to the ground you can find many proprietary trading firms which have similar or even superior returns. It's much harder to find a fund deploying similar strategies at the same capacity.

As a corollary, every single new hire at the firm is expensive in more than the traditional ways. They only have so much room in Medallion, and so they tend to be both extremely secretive and extremely protective when hiring. They've lessened up on that somewhat recently, but it's still there.

What are the people that you know like? Are they all multimillionaires now? Are they genius level smart by conventional standards. Did they apply for a job there or where they “plucked” from else where? How do they make such a shitload of money there?

I can't speak for throwawaymath, but the people I knew who went there generally were academics who had already done great things or from whom great things were expected. I don't have much use for the term genius, unless we're talking about Grothendieck or Witten. But yes, we're talking about people who stood out among the crowds of extraordinarily smart people in top tier academia, and who'd actually managed to _do something_ with their intelligence.

I think it's fair to say they were generally recruited via social networks. They knew what RenTec was, they knew someone who worked there (maybe an old grad school colleague or a former student), and when they reached out, RenTec already knew who they were. I don't know of anyone who went there who didn't already have a reputation in academia.

So contrary to astazangasta's original claims, they do seem to me to have something of a monopoly on brains. There are other smart people at other firms, but I don't know of any other place that has such a concentration. (Aside from Jeff Dean's office, I suppose. ;)

I'm not one of those people that have a moral objection to making money, or holds to crazy conspiracy theories about hedge funds, or anything like that. But it does seem kind of sad when you lay it out like that.

Eh, these aren't people who were forced out of academia because of scarce resources. They left because they wanted to work on the kinds of problems RenTec provides, in the kind of setting RenTec provides. When they left, they freed up funding for someone who wanted to stay in academia.

We do have rather interesting incentives in our society to induce so many talented people to do essentially zero-sum game trading instead of science or math.

Commercial work will always pay better than academia. Roles like trader/quant where your value is measurable & portable will always pay better than being a drone in a big machine like Google. If anything, front office finance roles aren't underpaid: people with similar skill sets are underpaid elsewhere.

I don't work in trading anymore, so I'm not talking my book here. Trading is zero-sum at a transactional level, but has knock on benefits beyond profit and loss:

-Making it easy for companies to raise capital through IPOs or offerings (without a robust secondary market for securities, people will be less likely to invest)

-In commodities: Letting businesses bear the risks they want and insure against the ones that aren't their core business

-Liquid markets let real people trade in and out of investments without friction at fair prices

-Providing accurate price signals to other businesses and the broader economy

So I don't think I was saving the whales, but I don't think it was wasteful, either.

Also, as a mildly clever OCD math guy who's semi-good at writing fast C++ code, I don't think I would have been curing cancer anyway.

ETA: At least in the case of HFT, if you accept that markets need intermediaries of some sort, it seems more efficient to have a few dozen tech/math guys do the same job thousands of guys in mesh vests were doing years ago, and cheaper.

I knew, or at least knew of, some of the vest guys as fathers of people I grew up with. As in any group it’s a mixed bag, but I feel pretty confident that none of them were missing their true calling in pushing forward the boundaries of our understanding of hidden markov models or getting speech recognition to work. We’ve used those kind of resources to free up people to, I don’t know sell whole life insurance policies or something. I’m sure there’s a sense it which it is efficient, like I said I’m not a centrally planned economy guy, but there does seem to be something off about it.

I mean there are computational cancer models that need to be written and optimized.

I know some folks there too, and I sincerely doubt they're fraudsters.

But character references aside, the claim that Medallion is skimming the other funds just doesn't add up: Medallion is at capacity and has been for at least a decade. It's been producing at least $3B / year in profits during that time, and that's redistributed to RenTec employees as cashflow, either as bonus (from their infamous 5/45) or as profits returned to the Medallion investors. So, we're talking about minimum $30B spun out in cash over the past decade. That's basically the total AUM of their other funds.

There was a conspiracy theory making the rounds for a bit based on Simons connections to Columbia. He rode a moped down there in the 50s, and I think won a tile or carpet factory in a card game. I don't think anyone but the haters give it any credence, but it shows you what kind of person Simons is besides "smart."

I think it's a fair guess that they are just better at playing the game than most. The people I've known from there are all very good, and alums have included people like Lenny Baum (aka hidden Markov Baum-Welch) and Elwyn Berlekamp. They also were almost certainly the first or second (Thorp knew) fund to truly figure out optimal bet sizing via Berlekamp's association with Kelley. Mercer's work in speech recognition at IBM is also evocative; speech recognition has been around for a long time, but it's always been really difficult and inherently a time series problem, like markets. Lots of Mercer's old team got pulled into Rentech looking at different kinds of time series.

Tax optimization is just another part of the game. I know a bunch of guys who renounced their US citizenship and moved to Bermuda to run their fund; 36% compounds pretty quickly.

I don't think the card game story is true, alas. Recollection is that Simons convinced his father to invest and they nearly lost their shirts when the tile factory took longer than expected to become profitable.

The story about Simons hiring essentially the whole IBM Speech Recognition group is true though. I have a neighbor who worked there in the early 90s. He's got a paper from that period with 4 other co-workers, and he's the only one of them who didn't end up at RenTec.

Article does not seem very well written. It presents margin trading as some sort of special privilege. It also makes some very vague suggestions about "large partnerships". And then it name drops RenTech, even though they really aren't that special in this connection.

> the suggestion that unlike the other outperforming hedge funds that turned out to just be insider trading, this one is real and they are just that smart.

They probably aren't just smart, they also have some process that isn't just about being clever. They optimize their execution, they make sure their research platform is top quality, they make sure they have the cheapest funding, they make sure people want to stay and work there, they keep an eye on what other firms are doing, and so on.

The thing about the investment business is there's a lot of focus on being smart, to the detriment of everything else that you need.

Here are a few insights on the rentech process from Nick Patterson, one of their senior statisticians there for a decade:

[...I joined a hedged fund, Renaissance Technologies, I'll make a comment about that. It's funny that I think the most important thing to do on data analysis is to do the simple things right. So, here's a kind of non-secret about what we did at renaissance: in my opinion, our most important statistical tool was simple regression with one target and one independent variable. It's the simplest statistical model you can imagine. Any reasonably smart high school student could do it. Now we have some of the smartest people around, working in our hedge fund, we have string theorists we recruited from Harvard, and they're doing simple regression. Is this stupid and pointless? Should we be hiring stupider people and paying them less? And the answer is no. And the reason is nobody tells you what the variables you should be regressing [are]. What's the target. Should you do a nonlinear transform before you regress? What's the source? Should you clean your data? Do you notice when your results are obviously rubbish? And so on. And the smarter you are the less likely you are to make a stupid mistake. And that's why I think you often need smart people who appear to be doing something technically very easy, but actually usually not so easy.]

[[at] my hedge fund, which was not a very big company, we had 7 Phd's just cleaning data and organizing the databases]


quotes from ~30:06 & ~38:03

Nice find.

I met the dudes at one of their few spinoffs, Merfin/Edgestream. I didn't work for them, but it was abundantly obvious they pretty much used really simple tools with excellent risk management and execution.

I've seen systematic portfolio managers who switched firms and then took a year or more to begin trading -- when they expected a maximum transition time of a couple months -- because they took for granted the tools and data infrastructure at the prior firm. Even tasks as mundane as handling ticker changes can have an unexpectedly large time cost when you're the one who has to handle them.

That may be why there doesn't seem to be a large rentech diaspora, as opposed to Julian Robertson's tiger cubs, for example. One can't take the process with them and it's really hard to reinvent.

Yep, you sound like you probably know this, but a few people left RenTech for another firm back in the day. They never got those returns off the ground at the new place.

Edgestream was founded by people out of RenTech? Really? I didn't know that.

Main one: https://littlesis.org/person/215249-Sandor_Straus#

Berlekamp kicked around the idea of starting a new fund as well. I knew a guy who interviewed with him.

And here I've been assuming that you used to work for them...

No, I'm familiar with Edgestream by name and reputation only. I've never really looked into the origins there.

For what it's worth, I'm based out of Irvine, across the country :)

Ah, forgive me. Your comments fit the profile of someone I know socially who'd worked at Edgestream before leaving to do research in communications.

Irvine leads to other presumptions, of course. That golden sunshine..

That golden sunshine indeed.

> They probably aren't just smart, they also have some process that isn't just about being clever. They optimize their execution, they make sure their research platform is top quality,


> The thing about the investment business is there's a lot of focus on being smart, to the detriment of everything else that you need.

Yes, precisely. RenTech's research platform is key. All hedge funds recognize that data is integral to their success. But most funds drown in the amount of data they try to process. RenTech does not. A large number of their research team works directly on innovating data processing, not just strategy design and development.

maybe it's market making. they act as a middleman between buyers and sellers of options and other financial instruments

I suspect the reason for IRA vs 401k is entirely regulatory. For any values I try in a spreadsheet, paying the taxes up front results in larger post-tax amounts later (as long as the initial income tax rate is greater than capital gains rate).

Can anyone find another reason?

Edit: yes, I'm entirely ignoring predictions of tax rates rising in the future since that's not interesting mathmatically

By saying "IRA vs 401k" I think you're misspeaking. There's a Roth IRA, and a traditional IRA. There's a Roth 401k and a traditional 401k. I think you're trying to make a Roth vs traditional distinction, not an IRA vs 401k distinction.

You're right that the article confuses various things. But you're wrong in your analysis. Capital gains rate has nothing to do with this. Whether it's Roth or traditional, and whether it's 401k or IRA, you never pay capital gains rate, you always pay income tax rate.

You're also wrong in saying paying up front is better in terms of tax rate. In terms of tax rates, Roth vs traditional both have the same rate (assuming you have the same tax rate now as in retirement) (and IRA vs 401k also both have the same rate, regardless of tax rates now vs retirement).

Roth does have several advantages though. It has a higher effective legal contribution limit. It also allows a higher density of value per dollar, which is useful in a fund like the Medallion fund that is limited to $10B.

I'm assuming you're talking about Roth versus non-Roth. Mathematically, you're right, they are equivalent if your tax brackets don't change. It's pretty obvious once you get the correct formulas.

The are reasons to prefer Roth: higher contribution limits -- paying the tax liability up front lets you save more for later. You can also pull out contributions to the IRA tax + penalty free, and pull out earnings for limited purposes (medical, downpayment on a first home). And you're not required to withdraw which can save you down the line.

Contributory IRA and regular 401K are basically the same thing: deferred taxes, required distributions.

Roth IRA is different: no required distributions, no tax. (or a least they promise no taxes).

You might want a Roth IRA because of ...

#1 Higher tax rates in future.

#2 You're not going to spend all your money, Roth IRA can provide tax free distributions for your heirs over their lifetimes. An inherited contributory IRA or regular 401K is taxed at distrubtion and distribution is accelerated (up to 5 years).

The upfront deduction and credits help people invest. When I was a kid, it was about 25% of my investment.

Ignoring the tax rates is dangerous with the abuses of these programs, it’s going to happen.

Interestingly, this private, closed fund constantly has its returns leaked, PR articles about it, etc. You are welcome to invest in their open, and very mediocre, funds though.

> You are welcome to invest in their open, and very mediocre, funds though.

Are you really though? Show me how an individual can invest in any Renaissance fund (e.g. RIEF) They are only open to institutional investors as far as I know - i.e. those looking to invest say, 250m or more.

Notably, Madoff hedge fund was also very undercover and hard to get into and had was quite picky who's money they accept.

Isn't this Robert Mercer's company?

James Simons founded Renaissance in the early 80s (ish). He brought on Robert Mercer from IBM later. Then Mercer took the helm when Simons retired around 2010, if I recall correctly. Mercer himself stepped down last year (likely due to political blowback from the rest of the firm).

So the short answer to what you're asking: yes.

I think the short answer is: kinda. Mercer was co-CEO, with Peter Brown. He was demoted when his political activities impacted recruitment.

My understanding is that Simons is still holds the majority of RenTec shares, with Mercer around 3rd place. RenTec made Mercer rich, but it is isn't his.

I agree with you. I'm responding to what I think is the spirit of the question. These days if someone is asking about RenTech in the context of Mercer, they're doing so because they're mostly heard about the firm as an auxiliary topic to Mercer's political activity.

And last I was aware, yes, Simons still has the majority of shares. If I recall correctly that's publicly disclosed though.

I thought there was an income limit for contributing to a Roth IRA? I'm guessing all of their employees with access to the Medallion fund would be above that limit. I must be missing something.

'The firm initially terminated its 401(k) plan for employees in 2010, a step that permitted them to roll the savings into traditional IRAs. Then, employees took advantage of a rule change that year allowing affluent Americans to convert their traditional IRAs into Roths. '

Looks like they rolled all their savings in to traditional IRAs and then converted them in to Roth IRAs. This procedure, known as a 'Backdoor Roth' conversion, effectively allows unlimited Roth IRA contributions beyond the income limits. If you are scratching your head and asking, why this is legal - so are many other people, but it appears to be accepted as within regulation by tax experts.

Backdoor and Mega-backdoor Roth conversions do not allow unlimited contributions. They allow an unlimited conversion of (limited) contributions.

As for whether it ought to be legal, my view is that multiplication (of which both income taxation and compound investing returns are forms) is commutative and taxing an amount now (the conversion) and then having it grow for 20 years is no different than having it grow for 20 years and taxing it then (traditional IRA), assuming growth and tax rates remain the same.

A Roth conversion is in essence a bet on your marginal tax rate at withdrawal being higher than your marginal tax rate today.

Your comparison assumes that the alternative to backdoor Roth is a traditional IRA contribution. The (ab)use of backdoor Roth contributions is commonly done by high-earning employees who have access to a 401(k)/Roth 401(k) plan. In this situation none of their IRA contributions are actually tax deductible at all!

So the 'backdoor' Roth contribution allows such people to invest post-tax dollars w/ Roth benefits when the IRS income limits seemingly make it clear that they shouldn't be able to contribute in this fashion. Why does the IRS have income limits on Roth IRA contributions but then allow them to be so easily bypassed in this manner?

> assuming growth and tax rates remain the same.

Big assumption.

> Why does the IRS have income limits on Roth IRA contributions but then allow them to be so easily bypassed in this manner?

I think TFA answers this. The limits were put in originally, and then as part of Bush-era tax cuts, as a way to raise current revenue, the conversion was allowed knowing that billions in taxes would be paid today in order to convert. They predicted $9 billion and got something like $30 billion instead.

Taxing the money now, when a large part of it could be passed via inheritance tax free later, is not necessarily bad policy.

I have been suspicious of Roth IRAs because I keep seeing sales taxes increase in the US and a push for something like a VAT. Which might negate most Roth savings (but not in this case).

I agree 100%.

I like cash over magical promises that can wiped away with the stroke of a pen.

Interestingly enough, this is also supported - sometimes explicitly, sometimes implicitly - by several top tech companies. It can be a bit tricky to pull off if it's not systematically endorsed by the organization, but it's often very doable.

There is for direct contributions. However the strategy used here is sometimes called the backdoor Roth IRA because there are no income limits to rolling a traditional IRA into a Roth.

There is also mega backdoor after tax 401k to Roth IRA

Correct, no limits. But taxed.

It's taxed just the same as if you were below the income limit.

This article highlights what I hate about financial reporting. Take this quote "If you are expecting to earn a rate of return of 20 percent or higher, it turns out to be a really good tax shelter." Or this one "The difference: ordinary people don’t have access to Medallion, one of the most successful hedge funds of all time."

It's basically insinuating that the techniques Renaissance used are somehow unfair because Medallion has such high returns, as if those returns were anointed by God instead of the work of the people at the fund. In addition, it insinuates that those high returns will go on forever - well, there have been more than few money managers who have seen their years of overperformance decimated by 1 or 2 bad years.

It's fine to argue that a Roth IRA may not be good tax policy for the country, but the only thing "special" about Renaissance is they appear to be better at investing than most of their peers.

No, they also have created legal advantages for their employees that aren't available to most people. E.g. I cannot do any margin trades in my IRA and am subject to settlement rules if I try to do day trading. Additionally, if I founded my own startup I would not be able to invest my Roth IRA money in it and let it grow tax free.

Via a self-directed Roth IRA, you probably could invest in your own startup.

Looked into that back when I was founding a startup - lawyer said it was a no-go due to self-dealing rules.

Your lawyer was wrong, then. If you look around, you'll find coverage of people successfully doing their startup shares in their Roth IRA, and I know one billionaire personally who did that.

It’s a facts and circumstances thing. Yes, it’s possible to do. It’s not possible in some circumstances. (Not possible for S-Corp, not possible if the company will be buying assets [inc IP] from the IRA holder, probably not possible if the company isn’t profitable enough to pay a salary from operations [you can’t work for free for your IRA holding, it’s legally risky to use your IRA to funnel you current salary], etc)

GP’s lawyer could very well have been right in advising that GP couldn’t do it under that specific set of facts.

>probably not possible if the company isn’t profitable enough to pay a salary from operations

Yeah, that was definitely the case.

This is precisely what Mitt Romney did with his Bain Capital shares.

What about self-directed IRAs?

> It's basically insinuating that the techniques Renaissance used are somehow unfair because Medallion has such high returns, as if those returns were anointed by God instead of the work of the people at the fund

The general assumption I've heard from friends in the industry -- which may be justified or may not be, I am certainly not qualified to say -- is that they're juicing the Medallion fund with returns from their other funds, they're just using clever trading mechanisms to effectively "launder" what would otherwise be highly illegal transactions.

The advertising mileage they get out of Medallion alone would probably make the activity worthwhile, but I'm sure the tax-free income from the Roth IRAs doesn't hurt either.

The chronology doesn't really line up for this story. The other funds didn't come online until Medallion had had 15+ years of stellar performance.

Applications are open for YC Summer 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact