Comment by MilStdJunkie | Hacker News Reader

Comment by MilStdJunkie | Hacker News Reader

Parent Full thread

Old college buddy switches around between fintech executive, State Department fixer, and finance journo. I asked him the same question about RT back in 2018, and his[1] reply was memorable:

"Secret sauce is always 'insider trading'"

He expanded on that a bit to also include buying up all the critical fiber nodes between the trading datacenters, and also laying down a new kind of fiber side by side with the old ones that's a teensy bit faster, and most critically, it knows how much faster it is. But the really big breaks - again, according to this buddy, who of course none of you know, so this has zero credibility - always come back to what's basically spies.

I don't think he's correct because of simple economics - spies and insider trades aren't going to be a smooth curve, and your investors will be pissed if one year you give them 5000% and the next 12%. Sure, you can move the money around to make a curve, but will your board let you do that? But yknow what? I don't work in the industry, so benefit of the doubt goes to the buddy who does. I'm not sure I could deal with working in the industry, given that the financial industry's done more to break this country than any other singular force. I would overdose in irony.

[1] He's a very cynical man

They were successful when they were trading at lower frequencies and they hired people who implemented successful strategies that are pretty well known at this point to come work for them. People like Ed Thorpe had similar Sharpe ratios with less leverage so I think their returns are entirely realistic. It also took them about 10 years to get consistently profitable (under different names).

The low-latency networks you're alluding to exist but there is nothing nefarious about them. They aren't insider trading they are just getting market information faster than almost all other participants. I also don't think RenTech does this, it's more of an HFT strategy, and the HFT industry as a whole makes a few billion in profits per year, a lot of money sure, but it's not a huge amount when you consider how many transactions they are involved in.

Also, while the financial industry has caused some major blowups and bad actors often go unpunished it's a crucial part of any developed country and enables almost all big projects to get funded.

I had a high school friend with a relative who was a reasonably-successful hedge fund manager ($500M AUM in the early 2000s) - nothing crazy, no insane quants, just a pretty traditional (for the time) fund, and in moments of candor (usually assisted by THC), the relative basically said the exact same thing:

Most of the alpha in the traditional hedge fund world often just boiled down to getting away with insider trading. It was so rampant that it was just part of the game. The stuff we all heard about in the news was just the particularly egregious stuff the SEC HAD to blow up. But if the SEC prosecuted every instance of anything even remotely insider-y, there would have been no hedge funds left.

(Now to be fair: I'm not really speaking to Renaissance - just to the general industy in the early 2000s. I think other commenters have made compelling points that Renaissance may very well have some actual special sauce.)

Buying up fiber wouldn’t be a way to make money in this space. The space is dominated by microwave transmission links as they’re considerably lower latency than fiber.

Yeah, he was talking on old info, probably because when I was hanging out with him he'd just gotten back from five years in <LOCATION_REDACTED>. Microwave, shortwave, and that weird dual-mode laser AOptix thing have been the rule for a bit, although I don't know if the laser thing ever really worked. Maybe enhancing the other things. Probably by now they have their own satellites.

Microwave has been the standard for about 13 years now. Laser hasn’t panned out as I understand. Satellite wouldn’t help much because the bounce up is longer than just shooting in a near straight line with repeaters for long haul.

They mostly hire Physics Ph.D.s in the same close range of fields. As much as I find the insider trading narrative to be more plausible than any secret sauce, I think their hiring speaks against that.

Eh, I've worked with Math and Physics Ph.D.s throughout my career (and I am one myself). Those qualifications are definitely insufficient evidence on their own that you could beat the market. Not to say that hiring isn't their secret sauce, but you'd have to hire the _right_ Physics Ph.D.s, whatever that means.

I'm not saying that they make money because of Physics Ph.D.s. I'm saying it would be kind of useless to hire those specific people only to end up doing insider trades the whole time.

> As much as I find the insider trading narrative to be more plausible than any secret sauce

I would be interested in hearing why you find it more plausible.

I'll start. I find it less plausible because you actually go to jail for that.

If it's a secret sauce that is based on public information (aka not insider trading), then it's unlikely that no one else but this one company has discovered it. It's just extremely unlikely. Plus, plenty of "godlike investors" were clearly willing to do things that landed them in jail (SBF, Madoff, etc).

Hiring as an edge makes a lot more sense. Quant trading is often not about a single strategy but about the sum of many small ones. And other quants that have been taking similar hiring approaches also had incredible success (eg. Jane Street, Citadel, Jump Trading, Optiver, etc). RenTec just has a reputation of being "the best", which in turn makes the best people more likely to join them, which again in turn makes RenTec better, like a vicious cycle.

I also want to mention that in some years, there's a massive discrepancy between RenTec's internal Medallion fund (+76% in 2020), and its external funds (-19% to -31% in 2020). Not saying there is something to it (they never claimed using the same strategies), but it does seem at least a bit suspicious.

> Hiring as an edge makes a lot more sense. Quant trading is often not about a single strategy but about the sum of many small ones. And other quants that have been taking similar hiring approaches also had incredible success (eg. Jane Street, Citadel, Jump Trading, Optiver, etc). RenTec just has a reputation of being "the best", which in turn makes the best people more likely to join them, which again in turn makes RenTec better, like a vicious cycle.

Weren't you meant to make the case for insider trading? This isn't insider trading.

> I also want to mention that in some years, there's a massive discrepancy between RenTec's internal Medallion fund (+76% in 2020), and its external funds (-19% to -31% in 2020). Not saying there is something to it (they never claimed using the same strategies), but it does seem at least a bit suspicious.

Are you telling me that the insiders keep the good stuff for themselves and the outsiders get fucked? I don't find that suspicious at all.

I'm supporting dna_polymerase's point, that RenTec's edge is much more likely to be hiring smart people than insider trading and more likely to be insider trading than some secret special trading algorithm that no one else managed to come up with in 20 years.

To my knowledge, RenTec aren't market makers, they supposedly run directional bets. It is highly unlikely to do that with their size of funds ($130B) for that long (>20 years) with that success rate (>60% p.a.)

Without signing on to the belief that RennTech is generating returns from insider trading, I will observe that hiring a bunch of Physics/Math PhDs would be a reasonable cover story and therefore wise thing to do, even if your core returns came from insider trading.

In addition, it could be that the PhDs' technical research and work is about how to structure a portfolio strategy such that it can profit from insider trading when the opportunity arises without triggering any SEC alerts.

For example, maintain a portfolio across as wide an asset base as possible, while trading each asset frequently enough that any insider trade inserted into the mix looks random, or at least plausibly deniable, and thus a low-odds case for the SEC to prosecute.

insider trading cannot be scaled up as well and smoothly as what Ren Tec does

I agree that insider trading is an unlikely explanation for the reason you mention.

It could be that an early algorithmic advantage compounded into an incumbency advantage: better research tools; better data; better pay for better scientists.

The 'secret sauce' may have started out as novel discoveries and, over time, become 'just being RenTec'.

This doesn’t seem to me at all cynical, since members of congress engage in insider trading without consequence.

Reply on news.ycombinator.com