Regarding the last point in working at Jane Street versus research on fusion/cancer:
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.
That sounds as if you can jump into a field without spending 10-20 years of learning and do cutting-edge research. I'm not sure whether someone who has done quant finance can make meaningful contributions to the actual science. So if your role ends up spending money and doing top-level management, why not just fund companies that do and stay in finance?
[edit] To add one prominent example - it's doable, as Jeff Hawkins demonstrated. Founded Palm, made money, then created his own brain research lab. As far as I understand, the neurological community has not embraced his ideas with open arms, but they are at least intrigued by his universal computational model of the brain. So that's a pretty big accomplishment.
I think Hawkin's work (including things done at Numenta, his company) is highly controversial in neuroscience. I have read some of the articles and claims and the biggest issues is that he produces a lot of vision papers, that contain no actionable models and thus ultimately are useless.
So either he's withholding whatever concrete insight he found, or he hasn't found it yet - I believe the latter is more likely.
> as if you can jump into a field without spending 10-20 years of learning and do cutting-edge research.
I know multiple people who have done this.
If you want a famous person, look at Paul Erdős. Always jumping into new areas of mathematics and solving problems at the cutting edge.
Or, to get to the current subject, Taleb using his background as a trader to jump to a career in academia, where many consider(ed) his research cutting edge.
I agree that a smart, ambitious person can become expert in more than one field in their life — but jumping to a new area of mathematics is on an entirely lower level than jumping from algorithmic finance to cutting edge biomedical, especially at the time Erdős was working.
He already had a PhD so it's not surprising he was able to enter Academia. But outside his books I'm not aware of anyone talking about his Research much.
I spent most of my career in finance including a hedge fund and several international banks.
Personally I really enjoy his books and recommend them to my friends and family.
Professionally, I have never heard his name being mentioned.
You find the same 'fluffy' management books on trader desks (varies between firms).
- Some translation of "the art of war".
- Tribal leadership and their whole "Tribe of tribes" nonsense
- its name escapes me at the moment, but some leadership book written by a US marine? Because this maps directly to finance?
- six sigma
Usually they are "management" focused chanting type material.
I'm not away of Erdos being at the cutting edge. There are far better examples of people moving between fields and doing very high-level stuff. For example, David Mumford, who has a great blog by the way.
I can't judge that from an outside perspective, but to be fair, having nothing to show for two decades of work is an intrinsic occupational risk of science.
> I can't judge that from an outside perspective, but to be fair, having nothing to show for two decades of work is an intrinsic occupational risk of science.
How? In today's academic environment you won't survive even a few years without publishable research results.
Jeff Hawkins probably also had a several Hundred Million Dollar payout. That's in a totally different class than a Quant who cashes out at 30 with MAYBE $10M in the bank.
You’re overestimating the technical skills needed to do cancer research: there’s a reason why many wet labs allow high school students to come and help with research. It’s mostly grunt work and whatever technical skills can be learned by a high school student over a summer.
I would venture to say the average Jane street worker has done more good for society than the average cancer researcher or Alzheimer’s researcher.
For fusion research, the quantitative skills from many Jane Street people can easily transfer to make meaningful contributions to fusion research.
I do cancer research. You're right, the technical skills for basic wet lab stuff are not hard to learn. The hard part of cancer research is not learning how to pipette, it's learning how to ask and investigate worthwhile questions. Experimental design is subtle and requires wide ranging knowledge of sources of biological and technical confounders. Growing cells might be easy, but do you know how to debug a fluorescence microscopy experiment; in fact, do you even know how to recognize that it needs to be debugged? This is stuff you can learn over the course of a few years in a phd program, but it takes longer to become a true expert. Finally, what do you investigate? Cancer? What tissue type, which cell type, which proteins, DNA structures, or RNA structures are most important and least understood? Just absorbing a small fraction of the literature so that you don't ask stupid or boring questions is a lifelong task.
So you can do break-through research after having done a little bit of wet work in a cancer research institute, as a 30-year old who has been coding quant systems for a decade, and then become equivalent to a PhD in the field?
Or you can take your ~10M or whatever in earnings during a decade as an OCaml programmer at Jane Street and fund a cutting-edge cancer research lab?
Also I said the average researcher, meaning taken from the global population, ie not from a top institute. Jane Street is highly concentrated in talent that produce real results in the world.
The average cancer researcher has probably done more good than the average trading firm worker though.
It's wildly inaccurate to say that much of Alzheimer's research has been shown to be fabricated. The article you link to is about one person's work. Even though he is a very influential researcher, there must be at least thousands of people doing research on Alzheimer's disease, if not more.
It seems none of the people that posted here understood the intention of my comment: The parent comment by ChadNauseam implies that Madagascar citizens somehow would not benefit from cancer research (they would!), which could only be true if they don't get cancer. In my comment I pointed out this last end of this logical chain of conclusions to show how ridiculous the parents' post was.
Nope, just no treatment (not fact checked of course, but it sounds likely that they wouldn't get top notch treatment that an average person say in Western Europe could expect)
> It’s mostly grunt work and whatever technical skills can be learned by a high school student over a summer.
Who do you think is supervising the high school student? Where does the idea for the project come from? Where does the money supporting the high school intern' experiments come from?
But that's not the interesting part, right? That would be like saying that tech is simply assembling prototypes. The really interesting decisions are the strategic ones that require both a high-level overview of the opportunity landscape and some foundational knowledge of its feasibility. Or am I mistaken and it's simply brute force trials?
Yeah you’re right. My point is that the technical bar for entry is low, and can be attained maybe by spending one year in a top lab. You then skip the hazing ritual that is the PhD and postdoc and directly start your own small lab.
Spend your money attending conferences to make connections and get yourself updated in the field.
Hire technicians to help with your grunt work. Spend your days reading research papers, discussing science at conferences, and setting up new experiments.
The bar for doing some experiments, in an environment with tons of logistical, technical, and intellectual support, is indeed pretty low. A high school student could certainly learn to run a gel in a week or two; patching a neuron might take a few months.However, the physical "act" of collecting data, especially in the happy case where all of the conditions have been worked out and the results look "as expected", is a very small part of being a scientist.
More often, you are trying something that hasn't been done before and you're getting results that don't quite make sense. Here, experience and background knowledge seem key, and I'm not sure that you'll pick up much of that in a year, even in a "top lab" because experiments are slow. On top of that, you'll need to learn how to design experiments and analyze/present their results in ways that your peers find convincing, which is in itself a non-trivial skill. All this presumes that you're even able to find your way into a "top lab", but that's not a foregone conclusion either: these places can be incredibly selective even among people with a decade of experience in the same field.
Put another way, your answer assumes there's a lot of fat to trim in the PhD/postdoc stages. What is it and can it really be cut down by 90% as you propose?
Long postdocs are a modern phenomenon due to the oversupply of biomedical researchers. Historically people did shorter PhDs and skipped postdocs. There are stories of old timers saying they got their faculty position based on just one or two papers, unthinkable these days.
If you’re independently wealthy, you don’t need to go through the modem hazing ritual and you can start your own lab much earlier.
I’m not saying you can start immediately and be an effective researcher, you will initially suck like everyone else. But you will have a better time learning how to fail if your career/livelihood is not on the line.
Not to mention that tightening spreads, deepening books, and equalizing prices across regulatory/financial/geographical regimes is a pretty serious social good in its own right.
I understand that (as the article mentions) these folks clean up when the wheels have already come off anyways, but day-in-day-out, the spread on AAPL is one tick ($0.01) nowadays, rather than the 1/8ths that you'd get quoted by some loud guy from Jersey 30 years ago.
Citations on this stuff are hard to come by, but it does seem at least directionally true that these advanced actors are making less money over time even as the problem becomes harder. If that's true, it's money not going into the pocket of a middle-man somewhere. Multiply that by everyone's retirement account and we're talking real money.
If you're going to hold AAPL longer than a quarter, then the tick vs. 1/8 doesn't matter, and if you're not, your trade doesn't need to happen to support the core goal of financial markets which is to finance companies.
You're the best kind of correct, which is technically correct. But what I said is that "multiplied by every retirement account we're talking real money". Which is no-qualifiers correct.
That ETF that you should have your roll in? It's buying and selling securities all the time, and encountering friction along the way. And whether people have ETFs or individual equities in their (hopefully tax-advantaged) retirement account, across everyone with a retirement account it adds up.
I know that people often have a low-key axe to grind about advanced market actors being "bad", and I know that politicians go to the well with this narrative all the time, but it's misleading at best and usually just demonstrably wrong. And with nothing but respect, I tend to bow out of conversations where people push the issue past a comment or two.
There are exceptions: Citadel paying 2x for PFOF on Robinhood vs. Schwab to get optionality on internalizing against dumb flow? Yeah, that's pretty iffy. But in general advanced actors are slicing strips of meat off of each other to the benefit of 401ks everywhere.
> That ETF that you should have your roll in? It's buying and selling securities all the time, and encountering friction along the way
I really hope it isn’t, or it would be making lots of tax events! I hope it’s doing in-kind transactions like all the normal ETFs.
> Citadel paying 2x for PFOF on Robinhood vs. Schwab to get optionality on internalizing against dumb flow? Yeah, that's pretty iffy. But in general advanced actors are slicing strips of meat off of each other to the benefit of 401ks everywhere.
No, robinhood average order is just smaller so less risk of adverse selection.
I didn't think that a digression about the tax optimization opportunities of in-kind securities swaps would be enlightening to a general audience.
As for your second point, I guess neither of us works at Citadel so we're both guessing, but if you agree that there is a market for order flow and that the most charitable interpretation of why that would be the case is because of the attached optionality for internal netting, then you're sort of making the assertion that all PFOF is equally valuable, which would be a hell of a coincidence.
As a sort of side note, I regard the cherry-pick the parent with ">" prefixes and go after snippets of what they posted as basically the lowest form of discourse on HN in spite of how popular it is, and I would encourage anyone to try to reply to someone's entire comment rather than just trying to find the weld points and lean on those spots. It's pretty weak.
> But in general advanced actors are slicing strips of meat off of each other to the benefit of 401ks everywhere.
I am not sure I understand. That would mean that the number of advanced actors would stay stable or go down over time (generally much research shows that markets tend to concentrate even in pure random trading, so the number of advanced actors should go down). Is that actually the case?
As I've mentioned elsewhere in the thread, it's notoriously difficult to get citations on this stuff so take with a grain of salt, but I've heard that in 2019 the "HFT" industry (defined some way) had cumulative annual profits in the US of somewhere between 2-4 billion dollars. That's a long holiday weekend for Google or FB. I've also heard that this (inflation-adjusted etc. etc.) this is down sharply from ten years before, when spreads were wide and undocumented order types were winked at.
To wildly oversimplify, market makers will tend to drive the spread down to the tick size, and arbitrageurs will tend to put themselves out of a job.
For the industry as a whole to be growing either in distinct actors or cumulative top-line, the number of markets and instruments and general financial activity has to be growing faster than the big dogs are eating each other. This is my (semi-informed) guess.
> HFT industry annual profits 2-4 billion. That's a long holiday weekend for Google or FB
Nitpick: that’s 10 to 37 days for Google or FB. 2B for Google is ~10 days, 4B for Meta is 37 days. Alphabet net income for 2021 was $76.033B, Meta Net income $39.370B.
Haha fair play, but to nitpick in return: when I worked on FB Ads, revenue was not even remotely uniformly distributed across days of the year, and Thanksgiving? Lots of family photos going up and being looked at relative to your typical Thursday.
Morbidly funny anecdote: when I first moved to the NYC office I obviously didn't have many friends, let alone a significant other. Myself and a few of the other sad saps who failed to notice the office clearing out a bit early around us were borderline freaking out over what looked like very anomalous behavior in the IG ranking system.
It wasn't until we raised it in a slack group that included some coupled folks that we realized it was Valentines day. Womp Womp. ;)
Is Citadel paying 2X for PFOF on Robinhood vs Schwab because it is dumb retail flow or due to increased optionality because it is inherently built in slow transmission flow to reach the execution platform?
You are really going to need a citation to back up that the core goal of financial markets is to finance companies.
That is, in my view, at best an ancillary goal (notice that most money in the markets doesn’t participate in buying shares from the company itself).
That may be what you want the markets to be about but every other participant has other desires from the markets and the great thing is they can all get what they want from them.
Markets in equities exist because companies want finance enough to be willing to sell the equity. Not sure why something that elementary needs a citation.
It's also true there are many other participants with many other strategies to extract that value created by the companies from acquiring and merging them to collecting dividends from a balanced portfolio to day trading, but the reason the market exists in the first place is because companies that create value need their capital in order to do it. As you correctly point out, most of the actual trades are secondary market ones involving companies not in the process of fundraising, but those trades are still positive sum inasmuch as without liquid secondary markets, companies that create the actual value might have found it too hard to raise funding. The difference between being able to sell TWTR on IPO day or shortly afterwards and being forced to hold it until an Elon Musk comes along and follows through with its existence has a huge impact on its ability to raise funds and grow. On the other hand reducing the time between trades down to smaller sub-second microsecond intervals is - whilst useful to people trying to win at essentially zero-sum trading games and inflating asset prices very slightly - going to have a pretty minimal impact on whether companies create more value by raising more funds.
I’m not sure you’re reading that correctly, since the response concerned the idea that the core goal of financial markets is X, and you respond with a statement on equities markets, which by definition only concern financing companies.
Equities markets are a piece of the pie. The largest, by far, for the kind of HFT we’re talking about, but a fraction of financial activity.
The core goal of financial markets is to gather and match prospective buyers and sellers so that trading can occur.
I was replying to a post talking about secondary market share trading replying to a post about the value of liquidity in AAPL, so in context it seemed clear equities were the market under discussion
Obviously true that financial markets for commodity futures etc have different functions, though a similar logic applies to them (the extra liquidity in commodities futures markets is useful to the extent it facilitates real world production decisions)
If the only people who trade AAPL plan on holding it for 10 years then by the time they want to sell it there will be no one to buy it from them, since the probability that someone else will want to make their once-every-10-years trade at the exact same time is zero. In order for long term investors to function they need liquidity to enter and exit positions.
Further: Suppose you had an auction on shares once an hour, or even once a day, but no continuous trading. Would that make the world any worse off? (except for high frequency shops?)
I mean, what you're describing is how to create a black market in securities for people who want to act on news, knowledge, or sentiment before tomorrow's auction.
Sooner or later (spoiler alert: sooner) someone would put it on the Internet, and it would be unregulated, at least at first, and the insiders would do even better than they are now.
You should read up on the early days of ECNs, Island and Archipelago and what not. Alternatively, you could look at this exact script being played out in crypto right now.
I doubt it. I've never heard a coherent explanation how liquidity on sub-second scale is a great social good, while at the same time the largest equity markets in the world are closed 2/3rd of the day, plus all weekends and holidays.
Because real world information has sub-second resolution and a healthy market should reflect that. It’s a continuous auction. Some markets are open longer such as FX.
The NYSE is closed between 16:00 and 9:30, plus all weekends and bank holidays. AFAICT real world information doesn't stop in the closed hours. So how is it credible that it's super valuable that trades can happen during the open hours at sub-second resolution, but we're suddenly ok at 16:00 with a 17.5 hour resolution?
You are begging the question. Why should a market reflect that? Why is that preferable to a daily auction (throw in a stochastic cut-off time to thwart HFT even more)?
Because you can’t make sure that all auctions happen at the same time. Consequently, news will disproportionately affect stocks having an earlier auction. That’s just one reason I’m mentioning.
I would suspect that almost zero people do this because working in an environment changes who you are as a person. If you spend a lot of time around cancer researchers you will think that cancer research is the most important thing in the world. You would need almost monastic mental compartmentalization to work at Jane Street for N years and remain singlemindedly focused on cancer research.
I thought about that often. Why are highly intelligent, passionate young activists not working into the industries they despise to change the system from within. But you're right that the system you're in changes you, corrupts you. You'd have to sacrifice a lot to climb the corporate ladder, doing things you despise, just for the chance to blow things up years, decades later. Your idealists friends would distance themselves from you. You'd have to constantly remind yourself that you're doing wrong things but that it's ok, because it is part of a bigger plan. It would be very isolating. You'd probably sacrifice your sanity for a foolish plan.
Most people think that their values come from some immanent "Self" but in reality our values mostly come from our surroundings. The same baby with the same genetics raised in Broken Arrow, Oklahoma vs. Upper West Side New York will have extremely different values, go farther than that and the difference in values is almost unrecognizable. If you or I grew up in wealthy families in ancient Greece, we would think NOT owning slaves was weird and suspicious.
Tangentially related is the movement of Effective Altruism, which is basically espousing taking a high paying job and then using the rewards to make a change.
The existence of a "movement" makes it much more likely to succeed. If you are working at Goldman Sachs during the day, you need to be talking to Effective Altruism friends at night, or you will almost certainly be washed away in the current of Goldman Sachs value systems.
By now, finance and FAANG have minted a very large number of tech millionaires. So, in practice, you have you ask yourself why ~nobody has actually done what you say. Personally I don't think it's because of a lack of funds.
>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.
A lot of people trying to change their careers would like to have a talk with you about this, I bet.
Ph.D in CS is 5-10 years ahead of newbie trying to learn computers
The article should not have delved into this as its a can of worms - stick with the industry, programming and math.
Its funny that the article uses a “chess champion” and a “concert pianist” as examples to to argue that you don’t question their occupations when it comes to being a benefit to mankind or not. I mean, the huge fucking salaries, where does the money come from? from fucking trees?
WTF! Seriously? I predict this will be one of the first article/HN post that has negative publicity for Jane Street .
For some reason articles on hedge funds bragging about intellectual abilities if their employees always seem to invoke these professions (chess champion, concert pianist).
I'm still waiting to see them bragging about having top proctologists in their team ;)
I think you are wildly underestimating the cost of running a world class lab.
Even if you retire with 10M by 30 you aren't going to run a world class research center with that kind of money for very long, or at all. Hell Land, buildings and equipment probably eat most of that right out of the gate. MIT's Plasma Science and Fusion Center had to shut down when their funding dropped from $28 to $14M per year. So at 10M you could Fund, forget about building, a center 1/3 the size of MIT's for 1 year.
I know we're supposed to assume the best interpretation, but I'm struggling to imagine how can someone think this is a reasonable comment other than through self-deception.
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.