I find this stuff fascinating, and this article is way above average for online posts about proprietary/algorithmic/quantitative/low-latency trading (very leaky Venn diagram there). I have a few nitpicks but overall it's informative and it's an interesting format: viewing an industry through the lens of a particular firm, especially one as fascinating as Jane. Anything that develops literacy in modern finance amongst the lay public is a good thing in my book.<p>If this stuff floats your boat I'd also recommend any of Carl Cook's talks, e.g. <a href="https://www.youtube.com/watch?v=NH1Tta7purM" rel="nofollow">https://www.youtube.com/watch?v=NH1Tta7purM</a>. Optiver is AFAIK in a somewhat different business than Jane, but they're also players (or were last I had any inside baseball).<p>Too many people got their worldview on this industry from "Flash Boys", and I say this as a Lewis fan, is criminally stupid at best and in bad faith at worst (if you want a well-researched, accessible alternative: <a href="https://www.amazon.com/Trading-Speed-Light-Algorithms-Transforming/dp/0691211388" rel="nofollow">https://www.amazon.com/Trading-Speed-Light-Algorithms-Transf...</a> is about a zillion times better).<p>It's a pretty short list of places I'd ever go through some grueling and semi-arbitrary gauntlet to work for, but Jane is on it for sure.<p>I hope the author(s) do Medallion next.
>the winners get a job from which people routinely retire rich in their 30s, and the losers... don't<p>Honestly, I find this ridiculous. Firstly, Yes, working at Jane Street is a well paying job and you'll do well out of it. No. People aren't routinely retiring in their 30s. I don't understand where this absurd idea comes from. Look at all the rich people in the world, look at how old they are, and ask, are they retired? No! People who are driven and smart don't suddenly earn their first $5m go off and buy an annuity. They're more likely to go off and found their own trading shop at 30 than they are to retire.<p>Secondly, you know what happens to people who don't get hired after their internship at Jane Street? They go to HRT, to G Research, to Jump, to Citadel, to Optiver, to IMC, to XTX, if they're really unsuccessful they'll go to Google, Microsoft, Amazon, Meta. These are not people desperate for a job.
Generally good article that delves into how a market maker functions. Couple points:<p>Re EA: quoting my boss, EA isn't really a dominant thing in the market maker space, it's pretty much only espoused by a couple high profile individuals (mainly, SBF).<p>Re strategy: point about how there's little strategy involved in being a market maker is off-base. Everything is ultimately strategy: do I continue to pour resources into a strategy that is losing money in the hopes of eventually seeing pnl? What markets should I focus on? What is the best use of time for each employee that maximizes pnl/head? etc etc<p>One thing that I thought article got right is that most work involved in market making is about avoiding trades, not making them. Capturing the bid ask spread is conceptually easy. The hard part is avoiding trading with toxic counterparties.<p>Part about put options is especially apt. Market making during a crash / recession (like right now) is especially difficult because all the non-toxic counterparties have stopped trading as much (people like to trade a lot more during a bull market than a bear one). By setting up a structure such that you profit during a crash (either by buying puts, leaning net short, or through some other method), you introduce an uncorrelated return stream that can really help.
> it's hard to argue with success: Jane Street earned $6.3bn in the first half of 2020, up more than 10x from the year before ($, FT).<p>It’s actually quite easy to argue with that. It’s 1 (or 2 at best) data points. 6.3bn is meaningless without knowing the capital put to work to achieve that. And finally if your 10x YoY it’s just as likely you had a bad year before as a good one this year.
Regarding the last point in working at Jane Street versus research on fusion/cancer:<p>You could maximise more good by first working at Jane Street in your 20s, retire by 30, and then set up your own smal fusion/cancer research lab where you can do research without being tied to government funding and politics. By 30, many cancer researchers have barely finished their PhDs, so you won’t actually be that far behind scientifically, but you’ll be far ahead financially.
Signals and Threads[0] is a podcast featuring interesting conversations from engineers at Jane Street<p>[0] <a href="https://signalsandthreads.com/" rel="nofollow">https://signalsandthreads.com/</a>
There's one thing that always baffles me about this kind of market work. Let's for the sake of argument assume that HFT and other sophisticated market making activities are crucial for price discovery and other great social benefits. Then why does this amazingly important social good get mostly turned off over 80% of the time[1]? Even as a retail buy-and-hold investor in boring ETFs not being able to trade outside normal office hours is an inconvenience. Surely the world economy has even more uses for trading at all hours than me?<p>[1] <a href="https://www.nyse.com/markets/hours-calendars" rel="nofollow">https://www.nyse.com/markets/hours-calendars</a>
The first argument made in favor OCaml is very similar to Paul Graham's Beating the average (<a href="http://www.paulgraham.com/avg.html" rel="nofollow">http://www.paulgraham.com/avg.html</a>), if anyone wants to read more about that kind of reasoning.
Most of this article is pretty interesting and unusually insightful about quant firms. But I find this sentence to be utterly disgusting.<p>> Trading is a lottery operated by market makers, and like the lottery its social function is to convert mass innumeracy into funding for better causes.<p>It’s basically saying “we should have your money because we know better than you”. Maybe true, but I find it a very weak argument for the social function of a quant firm. Especially when lotteries are essentially a form of regressive taxation.
Curious how people are so interested in Jane Street, ostensibly because they do technically challenging work, but much less so about other places where the work is at least just as challenging, but the money sucks.
> the winners get a job from which people routinely retire rich in their 30s<p>I'd love to know where this claim comes from! I'm not sure how much/what supporting evidence there is. It doesn't fit _my_ experience of people in Quant Finance. Of course, what does "routinely" mean here? 20%? More?
Why do all these companies with pretentious attitudes exist when the average* software still takes 30 seconds to show a paragraph of text? Ironically whenever these companies take the security test, they not only fall flat on their face, but are proven to not have the slightest clue how to do basic stuff like string escaping. I actually looked at the code for one of the top Haskell companies with the same attitude, and the story I just wrote is precisely what happened. Jane street sounds like a level 2 company like these Haskell companies and Cloudflare: They can escape strings, but only in places that have been famously exploited thousands of times, like SQL; they don't know how to actually know when the problem presents itself in a different unusual context, which they may have made themselves. It seems you need a million dollar employee income to reach level 3, and for level 4+ you simply need to be someone who is genuinely interested in the topic (as there is no monetary incentive) and have spent 10 years reaching it. Also, this applies to all aspects of tech, not just software security.<p>* not even average, this describes almost all software made in the last 20 years.
-- only managed to make it half way through the article so apologies if this was covered later - re: the intern game - if the game is confidence in the probability of your answers - couldn't you deliberately get the answers wrong and just bet low confidence in your answers? - this seems like a really stupid comment on my part so I presume I'm missing something important --
This is wrong.<p>The 100 poker-chips interview thing = interview BEFORE internship.<p>As you would expect, a job offer is based on full internship performance.<p>I don't have personal experience but I have friends who interviewed there.
From the article:<p>> One example of this is NAV trading (Jane Street has a paper here), where an investor wants to place a large trade in an ETF and agrees to buy it at some future point at whatever its net asset value is, less some small fee.<p>Should this say "agrees to buy it at some future point at whatever its net asset value is *NOW*"? Otherwise, I'm confused, why wouldn't the investor just buy it later?
Something I don’t understand:<p>Why haven’t their gains been arbitraged away? Conceptually what they do seems simple enough; and presumably you just need capital to do it. Hell, their own former employees could theoretically compete against them - as could many traders who would pay to learn those strategies.<p>So why are they still making so much? I don’t understand why their “advantage” hasn’t been arbitrated away into a commodity business.
I first heard about Jane Street taking the first computational linguistics course that used OCaml. I followed them over the next decade, every couple years seeing them pop up.<p>The given reasons for using Ocaml outlined in this article is interesting. It makes reasoning about complex systems easier. And making an "esoteric choice" earlier has advantages in hiring.<p>Assuming you give back to keep it going.<p>I don't have the advantage of secrecy where I'm at. But I'd be willing to guess that strategy could apply to other languages and markets.<p>Have fun first movers.
Quite a few inaccuracies in there. The ones that jump at me:<p><pre><code> - OCaml does type inference, so you don't actually declare the types and have the compiler check them, as stated in the article.
- Investors are not market-makers, the two words actually refer to the two types of opposed participants in the market.
- OCaml is the language used for research, but they actually have a lot of developers working on the compiler and on libraries for OCaml which are themselves implemented in C or C++.
- Jane Street is hiring massively and not nearly as exclusive as advertised here, though they do indeed pay slightly above the average. Most likely they had a few good years and are investing the cash they made into hiring expensive staff.</code></pre>
This reads like a very rosy picture of what these firms actually do. Surely they are secretive and tight lipped about all the money being made in low risk trades.<p>There are known loopholes that market makers get to exploit since they help keep the casino going. No need to make its a noble profession or compare to impact to actual economy or mankind.<p>These are the worst of the worst when its comes to exploitative and manipulative behavior to make money over retail trades just as a Hedge fund selling CDO's to pension funds.
> The other mitigation strategy is: just buy some puts.<p>I wonder who are the counterparties selling puts to Jane Street. My cynical view is that they are losing overall but the traders don't care because they are winning in short term (when nothing happens) and may have already changed their job when the market crashes.
1. It's realllyyyy hard to get hired. So many stories are along the lines of "I applied...blah blah... didn't get in"<p>2. You have to solve over the phone very hard math questions to make it past the initial screening stage. I dunno what comes after that.<p><i>The highest-stakes gambling events in the world are typically very discreet, invite-only affairs. One that might be close to the top in terms of available winnings happens at the end of Jane Street internships: interns get a stack of 100 poker chips and spend half a day getting asked brainteasers and then betting on their confidence in the answers. Some of these questions might be pure math and probability questions, some might be more abstract bets on making a market in some outcome, and apparently one of the questions is a tough probability question where part of the prompt is to bet on how long it will take to get the answer.1</i><p>I dunno why brain teasers are so important. I increased my account by 5x since the lows of covid to present with simple large cap tech and etf strategies (tesla ,tqqq, tecl, amazon, and others ). I don't need to mentally visualize 3d shapes intersecting 2-d planes or count colored vertices of hypercubes to make money or develop good strategies. Just some basic calculations and some other analysis..maybe advanced high school level. It's like if you want to find good traders, look for people with good track records.<p>If I were going to start a fund, I would do away with the puzzles. Instead what I would do is look for people who seem to have good track records on reddit or elsewhere and some decent risk management, like on wallstreetbets. There are thousands of users there and then I would try to find the best ones and try to quiz them on risk management to see if they are relying on luck or have a system. Recruiting from reddit or twitter is harder than linkedin, but I think the quality is better because you are seeing actual traders in their element. instead of hoping that puzzle skill will lead to trading skill, you just pick people who are already good.<p>Continued:<p><i>The other mitigation strategy is: just buy some puts. Markets usually don't crash upwards, but they do have a habit of crashing downwards. And for a market-maker, a crash is a uniquely interesting situation: volume is high, spreads rocket up because people are afraid to trade or don't have the liquidity, so an active participant can make a staggering amount of money. (I liked this Reddit AMA: "Yeah, 08-09 was insane. I've heard stories. No one knew what the fuck was going on and everyone was on edge. Then it all turned out fine and everyone got PAID.")</i><p>Puts bleed out a lot. Even during bear markets they lose money if the path dependency is unfavorable. The covid crash would have been perfect, but the 2022 bear market has been much more gradual, so puts would have done more poorly.<p>Yes, the 'crashing down' aspect is captured by the skew or smile. That's why a 20% ITM put will have a much higher IV than a 20% OTM call or be much higher than predicted by the volatility of the underlying. In some cases it will be massive...like a 38% IV compared to 11% for the underlying. So many people have tried to make put strategies work, and I have yet to see anyone do it, but if someone actually could I imagine they would not tell.