Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

What the interns have wrought, 2025

blog.janestreet.com

Post reply on HN
Print view View original

71–80 of 92 posts

Re: What the interns have wrought, 2025

#71

Post by Balinares »

pedroza_alex wrote: post #25
plst wrote: post #7

Earlier quoted context omitted.

Can someone who actually understands the topic explain to me (or link good resources) why/if what they do is useful to anyone? Or are they literally just in the business of making money? (anyone except themselves of course. I'm serious, any hints of irony are unintended)

AFAIU the service HFT forms provide to the market is liquidity. Basically allowing you to sell stocks/options as soon as you feel like it. In turn that leads to more efficient markets since prices converge to their "correct" value faster.

HFT is a different thing from what is being discussed in this thread. With HFT you're talking custom ASICs running within light-nanoseconds range of the target exchange. Ocaml very much isn't in this picture. This is about human-speed trading. Which also provides liquidity and correction of instrument prices towards their fair value, just at a different level.

The societal value of either is debatable all the same, mind you. It's more that wherever you have markets, you have money-making opportunities that can be leveraged, and therefore are.

Re: What the interns have wrought, 2025

#72

Post by Balinares »

infecto wrote: post #16
plst wrote: post #7

Earlier quoted context omitted.

Can someone who actually understands the topic explain to me (or link good resources) why/if what they do is useful to anyone? Or are they literally just in the business of making money? (anyone except themselves of course. I'm serious, any hints of irony are unintended)

Market makers or other similar HFT are providing liquidity in an efficient manner to the markets. The benefit is often debated but for the majority of retail and institutional investors, spreads have never been lower. Instead of a guy at the floor swallowing large margins up, you have bots electronically vacuuming pennies. Of course the whole point for a firm like Jane Street is to make money. To make money means the…

The societal value of liquidity and a narrower bid/ask spread, while non-zero, is not even remotely commensurate with the bucketloads of cash that the top firms bring in.

It's mercenary work, plain and simple. Advanced, interesting, full of juicy maths, highly competitive, rewarding, but mercenary. No one's doing this job for the good of the world, come on.

Give some of your earnings to trans defense NGOs, now that makes a difference and I'll be personally grateful.

Re: What the interns have wrought, 2025

#73

Post by infecto »

strogonoff wrote: post #69
infecto wrote: post #62

Earlier quoted context omitted.

Already said it before but the efficient allocation of capital benefits all of us. From the farmer hedging their crop prices to the local manufacturing company that is drawing a loan to expand their business. What system would you have in its place?

Off the top of my head, for the manufacturing company, why not have people invest in it? If it is doing something good, people will buy shares, and the company will gain the money to expand. For the farmer, if the issue is harvest instability, why not amortise the cost accordingly?

That sounds tidy in theory, but in practice it’s rarely that simple. Investors don’t just fund “good” companies, they fund opportunities with risk/return profiles that make sense relative to alternatives. Plenty of companies doing socially useful things struggle to raise capital because the payoff is long, messy, or uncertain.

Same with farming. “Amortize the cost” only works if margins and credit markets allow it. A small farmer facing price volatility, weather risk, and thin margins doesn’t have the same access to capital markets as a Fortune 500 manufacturer.

The whole point of efficient markets is to reduce those frictions, to better allocate risk and capital so that good projects (whether in manufacturing or farming) don’t just work “in theory.”

For me, no better system exists in the world. It’s not perfect but until there is something that works better I will have the agree with it.

Re: What the interns have wrought, 2025

#74

Post by disqard »

KHRZ wrote: post #2

I was curious what kind of company needed to develop all these different kinds of advanced tech. Turns out it's just so they can be sneaky and trade shares quickly. Kind of sad IMO.

Jane Street is in hot water in India:

https://www.bbc.com/news/articles/c5y0zgrevl1o

Basically, SEBI investigated Jane Street's trading on January 17, 2024, when the firm allegedly made about $86 million in a single day through what regulators term "intra-day index manipulation."

HFT and Ads are two places which (currently) print money, so it vacuums up all the talent, and puts it to use growing their already-huge revenue streams.

Re: What the interns have wrought, 2025

#75

Post by strogonoff »

infecto wrote: post #73
strogonoff wrote: post #69

Earlier quoted context omitted.

Off the top of my head, for the manufacturing company, why not have people invest in it? If it is doing something good, people will buy shares, and the company will gain the money to expand. For the farmer, if the issue is harvest instability, why not amortise the cost accordingly?

That sounds tidy in theory, but in practice it’s rarely that simple. Investors don’t just fund “good” companies, they fund opportunities with risk/return profiles that make sense relative to alternatives. Plenty of companies doing socially useful things struggle to raise capital because the payoff is long, messy, or uncertain. Same with farming. “Amortize the cost” only works if margins and credit markets allow it. A…

> Investors don’t just fund “good” companies

Exactly what I am lamenting in my original comment. Investors chase returns, i.e. money for its sake.

As long as that is conceptually a thing, it is a no-brainer to fund a bad company if you are sure its shares will go up in price during the term of your investment; it is also in your interest (and acceptable within the “money for its own sake” framework) to ensure its shares do go up by helping hype it up; etc.

This all, I believe, is a source of strong and far-reaching negative externalities, which I am far from sure are trumped by its potential benefits.

> A small farmer facing price volatility, weather risk, and thin margins doesn’t have the same access to capital

Why do you need access to capital in order to price in the risks or the cost of relevant insurance? (That’s what I meant by amortising, I might have used a wrong term.)

Why are you “facing thin margins” like it is not an open market where you set your prices and your margin is your choice?

Re: What the interns have wrought, 2025

#76

Post by infecto »

Balinares wrote: post #72
infecto wrote: post #16

Earlier quoted context omitted.

Market makers or other similar HFT are providing liquidity in an efficient manner to the markets. The benefit is often debated but for the majority of retail and institutional investors, spreads have never been lower. Instead of a guy at the floor swallowing large margins up, you have bots electronically vacuuming pennies. Of course the whole point for a firm like Jane Street is to make money. To make money means the…

The societal value of liquidity and a narrower bid/ask spread, while non-zero, is not even remotely commensurate with the bucketloads of cash that the top firms bring in. It's mercenary work, plain and simple. Advanced, interesting, full of juicy maths, highly competitive, rewarding, but mercenary. No one's doing this job for the good of the world, come on. Give some of your earnings to trans defense NGOs, now that m…

I don’t disagree that people don’t get into HFT because they’re trying to save the world. It is mercenary in the sense that the rewards attract talent. But calling the work itself valueless misses the point, liquidity and price discovery are public goods. You notice them most when they don’t exist, and the cost of capital spikes for everyone.

As for the “bucketloads of cash,” that’s just how competitive advantage in markets gets priced. If firms didn’t deliver something real, the money would dry up quickly. Markets are brutally efficient at punishing dead weight.

Philanthropy is great, give to causes you care about. But it’s worth recognizing that the system enabling those donations in the first place is the same one that relies on liquidity, efficient spreads, and functioning markets.

Re: What the interns have wrought, 2025

#77

Post by infecto »

strogonoff wrote: post #75
infecto wrote: post #73

Earlier quoted context omitted.

That sounds tidy in theory, but in practice it’s rarely that simple. Investors don’t just fund “good” companies, they fund opportunities with risk/return profiles that make sense relative to alternatives. Plenty of companies doing socially useful things struggle to raise capital because the payoff is long, messy, or uncertain. Same with farming. “Amortize the cost” only works if margins and credit markets allow it. A…

> Investors don’t just fund “good” companies Exactly what I am lamenting in my original comment. Investors chase returns, i.e. money for its sake. As long as that is conceptually a thing, it is a no-brainer to fund a bad company if you are sure its shares will go up in price during the term of your investment; it is also in your interest (and acceptable within the “money for its own sake” framework) to ensure its sha…

Returns-chasing isn’t some moral failure, it’s the mechanism by which capital gets allocated. If you strip that out, you don’t get fewer “bad” companies, you just get less disciplined pricing of risk and more capital scarcity overall.

On farming: margins aren’t simply “a choice.” Prices are set in global commodity markets, not by a farmer unilaterally. Thin margins are structural, and access to capital or insurance is exactly what helps them survive volatility rather than get wiped out.

I am not sure why you think farmers get to set price on a commodity item. They don’t and because of that will often leverage future contracts or other hedges to bake in prices early.

Re: What the interns have wrought, 2025

#78

Post by wk_end »

Balinares wrote: post #71
pedroza_alex wrote: post #25

Earlier quoted context omitted.

AFAIU the service HFT forms provide to the market is liquidity. Basically allowing you to sell stocks/options as soon as you feel like it. In turn that leads to more efficient markets since prices converge to their "correct" value faster.

HFT is a different thing from what is being discussed in this thread. With HFT you're talking custom ASICs running within light-nanoseconds range of the target exchange. Ocaml very much isn't in this picture. This is about human-speed trading. Which also provides liquidity and correction of instrument prices towards their fair value, just at a different level. The societal value of either is debatable all the same, m…

There’s a few orders of magnitude between “human-speed trading” and the absolute bleeding edge of HFT. A company like Jane Street still does automated trading far faster than any human could.

Sure, Jane Street probably isn’t the fastest in the business, but I wouldn’t be surprised if they’ve got FPGAs or ASICs, dedicated high speed pipes to shave off milliseconds of latency, things like that.

Re: What the interns have wrought, 2025

#79

Post by owenversteeg »

sarchertech wrote: post #29

The thing I’m most impressed by is that I didn’t see a single LLM wrapper project. In fact I didn’t see a single project that directly involved an LLM at all.

Jane Street interns are highly intelligent and hired to make money, not generate hype. I am not surprised.

Re: What the interns have wrought, 2025

#80

Post by siva7 »

MontyCarloHall wrote: post #26
KHRZ wrote: post #2

I was curious what kind of company needed to develop all these different kinds of advanced tech. Turns out it's just so they can be sneaky and trade shares quickly. Kind of sad IMO.

Agreed. Amongst the top 5% smartest quantitatively-minded kids I knew in college, I’d say ~25% of them ended up at quant shops like Jane Street, sometimes straight out of school, sometimes after doing a PhD. I can only imagine what good that brainpower could do for humanity if it weren’t occupied finding cleverer ways to manipulate electronic money.

Unfortunately brainpower isn't the deciding factor if someone will turn out to be the one doing good things for humanity
Post reply on HN