From the comments, how to hit it big?

I am not endorsing these claims, but I do enjoy a good rant.  It is an object lesson in showing how (some) people think about jobs, status, rivalry, and money.

First venture capital is generally consider where washed-out Wall Streeters go, when they can’t cut it in real finance. Very few b-school students start out trying to get into VC. And no, generally Silicon Valley people are not nearly as smart as HFT/algo quants. The type of kids who go to Google or Facebook are generally the Ivy CS students from the upper half of their class who are good at white-boarding problems (e.g. reverse a linked list). The truly brilliant kids, Putnam winners, math olympiads, core kernel contributors, etc. disproportionately go the quant route. (In which at least half will wind up in Chicago).

SV is generally a worse deal than HFT or quant trading. Starting comp is at least 50% higher than the big five tech firms, and goes up at a much faster rate. And definitely way higher than startups, which nearly always under-pay. It’s true in tech you can become a multibillionaire, but that’s extremely unlikely even for the most talented. In general SV is a bad deal for everyone except the small set of people lucky or connected enough to be at the top. Outside founder level, virtually no one gets rich from startups anymore. The equity and options comp is pathetic at best, if not outright fraudulent. (“You’ll be getting 1% of outstanding shares… from this round…”). Even founders have to live on 70k salaries in the Bay Area, then are frequently screwed over or cliff’d by their VCs. For every Google, heck for every Apigee, there’s a thousand no-name flame-outs, where no one but the VCs walk away with a dime.

Compare to quant trading. Compensation is cold hard cash, usually paid out annually, if not quarterly. Not lottery ticket equity with four year cliffs, unlimited dilution and byzantine share classes. Most comp is directly tied to individual trading performance, with clear results from trading everyday. No politics, extremely meritocratic, no being at the random whims of whether your app takes off fast enough to overcome your burn rate. Firms actually compete for talent and pay accordingly, instead of colluding to keep wages suppressed. Unless your ambition is to top the Forbes list, HFT’s a much better deal for someone extremely intelligent like a Math Olympiad. The probability of making “f-you money” before 40 is at least an order of magnitude higher as a prop quant than in the Valley.

That is from Doug.

Comments

"Outside founder level, virtually no one gets rich from startups anymore."

It would be interesting to look at this trend over time. How well did the 10th, 100th, and 1000th best-rewarded employees do off, say, Intel, Apple, Microsoft, Google, and Facebook?

I would like to see that as well. Good luck getting the data!

However, the OP was talking about startup "today". It is true that the startup scene seems considerably worse than in earlier years.

Yeah, I recall when Google went public, in 2004, everybody knew it would be a success, and the big investors were lining up for restricted shares. Now I'm sure it's different. I never got rich in SV (made about a half million) but I lived well and that's OK. The way my family got into the 1% was with DC area real estate, since we got in early (as in Eisenhower era! Back in those days the 'burbs was farmland almost).

Bonus trivia: Greek real estate has dropped so much that you can buy a concrete four story apartment building, with marble facade and brick interior walls for about the same price as a DC balloon frame wooden house, in similar 'good neighborhoods'. Location x3 (you cannot live here if you don't speak Greek). Who needs a college education when you have real estate? It's a joke but that's society for you.

When a start-up files with the SEC to go public, what does it have to reveal about who has been promised what?

You could also try to reverse engineer how much founders took. For example, Gordon "Moore's Law" Moore of Intel is today said to be worth $6.7 billion (and he's probably given away about that much as well). Robert Noyce of Intel went through a famous divorce in the 1970s that cost him half of his property, so there is probably a pretty intensive accounting somewhere of how much he got from Intel going public around 1970.

Mostly, S1s reveal executive compensation practices and actual salaries (and total comp packages) for the top executives. This is generally no more than ~5 people. They'll also contain information on board compensation.

If you read everything carefully there could conceivably be some cases where you could deduce an estimate for the 10th-highest-paid employee, though that is unlikely to be disclosed (and you'd need to have a very high tolerance for slogging through some very, very dry language). I highly doubt you could find anything useful for the 100th or 1000th highest-paid employee.

Employee #10 at Google was a couple years ahead of me in High School. I think he did alright.

"The Microsoft Millionaires Come of Age -By JULIE BICKMAY 29, 2005"

"While the exact number is not known, it is reasonable to assume that there were approximately 10,000 Microsoft millionaires created by the year 2000," said Richard S. Conway Jr., a Seattle economist whom Microsoft hired to study its impact on Washington State. "The wealth that has come to this area is staggering."

http://www.nytimes.com/2005/05/29/business/yourmoney/the-microsoft-millionaires-come-of-age.html

Seattle's an affluent town. If you assess the whole commuter belt, Miami, Dallas, Atlanta, Phoenix, and Las Vegas (among others) have, however, all grown more rapidly the last 40 years.

Obviously they have more than that number of VPs and they do very well. But by that time they were already large firms and it takes time to go through the ranks. You could do just as well at a non-SV firm.

It seems that most of the wealth in SV ends up in the hands of the people that work at the $10bn+ successes. This is a corollary of the VC rule that the only investments that matter return your whole fund. The companies that have small (<$1bn) exits don't add up to much. Just looking at FB's S-1, employees had 624mm shares in options and RSUs alone (so excluding Zuck and cofounders mostly) when they went public, worth $80bn today. They only had around 2,000 employees at the time.

I think if you talk to a wealth manager in the Bay Area, most of the clients are likely to be senior execs at big tech companies ($1mm+ salaries) or earlyish employees at companies like FB that are worth $10bn+. Most of the stories get written about founders at midsize companies because they're more interesting. But there are just a lot more people who own .01% of the $100bn companies than are founders who owned 10% of $100mm companies at exit.

All this is so sickening it makes me want to vote Trump.

Do people who have worked in or know people on both sides generally believe this? My experience is that finance programmers think they're gods, but that their code is actually quite poor. There are exceptions: for example, Jane Street has pretty good code. But in general, the quality of the code is very low and the level of arrogance is much higher in finance than in big tech companies (although startups can rival finance in arrogance). I find that people in finance are often very defensive about how smart they are and how much better they are. I've never been able to figure out why, though. Is it just insecurity?

From what I've seen of offers, I don't believe the comp claim. A "good" PhD with zero experience can get $200k+/yr in total comp at Google/FB/Apple if they know how to negotiate, and a "good" PhD in a hot area can get double that. Are finance companies paying fresh PhDs 50% more, or $600k/yr?

I agree that, as you move up, finance companies can pay much better, but when (admitted exceptional) entry level offers are in the $200k to $400k range, does it really matter if you make 2x that vs. 4x that if you're really senior and a company wants to hire you? I make > 10x what I spend. If that was > 20x, what difference would that make?

What if you only made 3x what you spend?

These spend to income multiples might make sense for a single 25 year old. But if you want to have anything remotely resembling a traditional American family in the Bay Area, even high-end tech comp doesn't cut it. I'm talking about a single-income household, a few kids, a decent sized house with backyard, good schools. Throw California taxes on top. Even getting to 1x, let alone 3x or 10x, that spend level requires 250k minimum. Probably a lot more. That means that senior engineers at Google are just barely squeaking by.

Hence the reason that startups have been able to sell dubious pie-in-the-sky dreams to starry-eyed developers. "Making it" in the Bay Area is only possible by winning the lottery, or screwing over your colleagues. The major tech companies have not kept compensation anywhere near on pace with Bay Area cost of living. Employee comp has to be suppressed, often through legally questionable wage-fixing, to keep delivering the insane profits that tech sector shareholders. Profits that mostly do nothing more than just sit as cash on the balance sheet. It's not really sustainable to expect 200 billion dollar companies to keep delivering double-digit growth.

Meanwhile the flood of brain-dead, bubble-chasing hot money into the VC space has completely corrupted startup-culture. The major tech companies of yesteryear couldn't even get funded in today's environment. YCombinator demands 250%+ revenue growth per month. Do you think early-stage Apple or Google was growing revenue at 250% a month? There's only a small subset of business plans that can deliver 250% growth per month, and damn few of them involve "meaningful" or "impactful" products. Building a world-changing product takes more time than 28 days. For everyone complaining about code quality, how do you think code quality looks at 250% a month? Because from what I've seen, it mostly looks like NoSql/node.js/Angular vomit.

The only real way to deliver those targets is to monopolize some platform with network externalities. Don't build something elegant, build something sloppy, insecure and debt-ridden to get to market first. "Move fast and break things". That's why social media, by far the most vapid and least impactful corner of the tech sector, is the darling of VCs. Think Instagram, not Sun Microsystems. If SV was actually about changing the world, then all these web startups would care a lot more about building protocols to defeat the NSA, rather than getting dog ear filters on people's profile pictures.

I wrote the original Skeeter comment (like from the show Doug, oh never mind). I agree with the gist of everything you said here and wrote my previous comment mostly to play devil's advocate. But, really, how much quant is there in Chicago?

"The other thing is that location is an issue. There are essentially no entry level quant positions outside of New York City. If you are in Chicago, there is basically only one employer that hires entry level quants (Citadel)." https://www.quora.com/What-is-an-entry-level-quant-job/answer/Joseph-Wang-9

I also think hedge funds/quant shares a lot of the same problems you identified with startups. The Quora post outlines a lot of the issues with quant programs these days. Graduates of top schools probably do well, but that's true of many career paths. My background is mostly cyber security, but I've taken some electronic trading, data mining and bayesian stats classes. I'm in New York if you're interested in grabbing coffee sometime.

Don't live in New York, but I'd definitey be interested in getting coffee next time I'm in the city. If you want to exchange contact info drop me a line at this email sometime in the next 48 hours:

kihicroslu {at} throwam [dot-com]

"But if you want to have anything remotely resembling a traditional American family in the Bay Area, even high-end tech comp doesn’t cut it. I’m talking about a single-income household, a few kids, a decent sized house with backyard, good schools."

A couple of things make me suspicious of the above claim:

1. Compensation is higher in the Bay Area.
2. Single-income may no longer be considered part of a "traditional American family" for the class of individual likely to work in the Bay Area. Male elites are more likely to marry female elites and female elites typically want to work.
3. What's considered "decent sized" is somewhat flexible across markets.

I realize cost-of-living is significantly higher than elsewhere and that compensation doesn't entirely make up for it. But it makes up for some of it, right? Then there's the intangibles:

1. Higher-than-normal odds of working at the next Google, Facebook, etc. Obviously extremely unlikely, but the odds are probably better in the Bay Area than they elsewhere. Can't win the lottery if you don't buy a ticket.

2. Surrounded by top talent, ambitious people, etc. Some people get off on that.

3. Plethora of tech jobs arguably makes finding "the next job" easier. Contrast this with someplace where tech jobs are scarce. Lose your position and you may have to move.

4. Natural beauty of that part of the country.

"YCombinator demands 250%+ revenue growth per month."

Huh?

I've worked at a small YCombinator startup (Asseta - 2013 class), and know a few who've gone through the full experience at other companies. I've never heard of this before.

Do you mean 25% per month?

The quality of coding in quant trading is extremely high variance, but it is important to understand that there is a good reason for that. Nearly every company has two types of code: infrastructure and pricing/order logic code. The former tends to be very good, or at least high performance and reasonably tested, while the later is intentionally thrown together by people who think of programming as their second or third job. I assume when you are talking about Jane Street, you are talking about their OCaml-y stuff, which is what I would call infrastructure . I wrote some truly abhorrent stuff on the pricing side for them for years, so I can vouch for the quality of the back end and the lack of quality on the front end.

There are good historical reasons for this setup: if your infrastructure code has a bug it is potentially company ending (see Knight Capital) or at least very costly. Also, infrastructure needs to change less frequently. Trading logic, on the other hand, has to move quickly. This was doubly true when automated trading was newly possible and it was just a race to see who could implement the strategies in code. When money starts falling from the sky, you don't go home to get your vacuum, you grab the biggest, closest basket and start grabbing.

All this is changing very rapidly. There are a plethora of reports that basically show that revenue in HFT/quant finance is falling, and costs are rising. Don't get me wrong: pay in bad years has still be way higher than any 25 year old needs to get by, but the FU money is harder to come by. I agree with a lot of the original poster's comments about SV vs Finance, but it's not obvious that finance (especially the finance practiced in Chicago, which is mostly HFT) has a lot of room to grow, and is most likely shrinking. Of course, some from Silicon Valley might express the same thing about their profession.

I've worked at IBM on a team that produced IT software, at a mercenary software consulting company doing random projects, at a fleet management startup, and at a couple start ups in the financial space (not doing anything related to HFT).

Most code is terrible. Like, objectively bad. That said, I'm not in SV, so I hold out the hope that all the "quality" devs are in SV.

My theory is that the "really good" coders (which is different from those who merely *think* they're "really good" but are, in fact, not) are, for the most part, working on "tough" problems. Mainly, things where either A. performance is important (e.g. HFT, browser code, operating systems, video game engines, etc.) or B. there is an abnormally low tolerance for bugs (e.g. NASA, operating systems, etc.)

95+% of coders are not working on "tough" problems. They're writing code that's more or less straightforward, and doing it pretty badly.

"95+% of coders are not working on “tough” problems."

That's true. It might even be low.

"They’re writing code that’s more or less straightforward, and doing it pretty badly."

That's obviously completely wrong. It most code was pretty bad, it wouldn't work. Instead a whole lot of code is mediocre.

In general code isn't elegant and often isn't commented. If it is commented it's in that bizarre method where you leave a comment for the obvious stuff (that no one needs) and neglect to write a decent meaningful comment for the important code.

And worse, I find that lot of code that is elegant, is too clever and sacrifices everything for speed. Which is (in most cases) poor coding. If you don't need speed, it's much better to make your code modular, readable and set up to provide a reasonable level of feedback.

"That’s obviously completely wrong. It most code was pretty bad, it wouldn’t work."

"Not working" is not my definition of "pretty bad". That is, I consider that code that, technically, "works", can still be "pretty bad". Not performant, not scalable, buggy, insecure, not maintainable, etc.

"If you don’t need speed, it’s much better to make your code modular, readable and set up to provide a reasonable level of feedback."

Agree. In general, I don't think "modular and readable" precludes code also being performant.

I think the world is full of crappy code, but that there is also a fair bit of really wonderful code written by someone who's a genuine craftsman, running in some control system (or similarly non-glamorous place) somewhere, performing about 20% better than you'd think was possible, but not otherwise changing the world. I don't think the best programmers all necessarily gravitate to SV or NASA or NSA or whatever--instead, I think there are a lot of really excellent programmers (as well as a much larger number of mediocre ones) working on unglamorous but important stuff, and known only to their coworkers.

I have a theory, yet to be tested and quite possibly wrong, that a software company could be wildly successful by pursuing a hiring strategy that filters out "mediocre" and leaves only "excellent". Some key insights that lead me to believe this:

1. The excellent guys don't make *that* much more than the mediocre (or "bad") guys. More, but not not commensurate with their productivity.

2. This is because managers tend to focus on the here and now and so end up discounting negative productivity incurred in the future by mediocre (or "bad") developers. Having to eventually rewrite everything to be more scalable, time spent fixing bugs, increased support costs due to buggy code, lower customer satisfaction, lower developer productivity because the code is overly verbose and convoluted, etc.

3. Many software companies' interview process is not-so-good at separating "excellent" from "mediocre".

4. Competent people enjoy working with other competent people. The prospect of a team consisting *solely* of competent people would be a big selling point and could lead applicants to accept lower compensation than they might otherwise.

So your plan for success would be:

1. Install an interview process that is good at identifying "excellent" candidates. Here "good" means "better than the processes used by ~90% of your competitors".
2. Don't hesitate to cut ties with staff who turn out to be not-so-excellent. At the same time, don't be a dick about it.
3. Along with #2, develop methods of accurately gauging which (if any) employees are performing not-so-excellently.
4. Compensate your technical staff excellently, though not crazily out of whack with the market.
5. Signal to prospective employees that, should they come on board, they'll be surrounded by top-notch coworkers.

"Most code is terrible. Like, objectively bad."

Apparently "objectively" is the new "literally".

@buddyglass - spot on. I program for fun, and I used to program even in OOP, 'hard coded' style, and I would add polymorphism "just for fun" until I realized the folly of my ways. Nowadays I do things the right way, using Gang of Four (programming patterns) and MEP is quite helpful for 'plug and play' functionality, see more here: https://msdn.microsoft.com/en-us/library/aa751829(v=vs.110).aspx

But most beginner amateur coders just dump junk into one big class, and it's not much different from procedural programming. C# is the language of choice for me, much better than Java as it has automatic garbage collection. C++/CLI might be the next best language (I used to program in pure C++, but it's hard to do smart pointers all the time, better to let the underlying compiler and architecture take care of it).

Ooh, buzz words!

I feel like that was a great troll by Ray.

Not *that* buzzy. I quit my last programming job in 2003, and there's nothing in that post that isn't understandable. Technical, yes, but even then, most people working in software, whether or not they're programmers, would follow it.

I make > 10x what I spend.

You're doing it wrong.

Not even sure how that's possible. Does he mean he spends 1/10th of his after tax income, or before tax? Either one of those numbers indicates some miserly levels of thrift at anything lower than 7 figures of income.

I'm 25 and spend ~40k/yr with ~100k gross income. There are guys at my firm ~5yrs older than me that are making triple+ what I make. If I don't let my lifestyle creep (big if), I could be around that 1/10th of before tax income range in five years.

This is highly misleading. Most programmers in finance are back office schlubs who make less and have lower status than the investment bankers and meatheads in sales and trading who can barely do algebra.

At a bank, maybe, but HFT and quant trading was basically founded on the smart people revolting against the meatheads. The companies he is referring to are run by quants and programmers for the sake of quants and programmer.

Most of the "smart people" don't work as quants. Most of them are in the back office of financial firms. Furthermore, quants tend to have less social status than ordinary MDs at ordinary banks.

Quant traders can get 10%-20% of the profits they generate. There's no way programmers in SV will get that kind of payout on average. For one reason most of the businesses they work on will never generate any profit.

But if the Efficient Market Theory is (largely) true, then it's hard to get money from quant trading, and arbitrage is going away now that 'speed bumps' are being installed in most exchanges.

Good rant however by the OT, and the replies.

"But if the Efficient Market Theory is (largely) true, then it’s hard to get money from quant trading"

You'd think.

The huge money in Silicon Valley comes from getting some kind of monopoly power, like in operating systems, search, or social media. How do you monopolize quant trading?

I'm not saying it can't be done, I'm just asking how do they do it?

There are a few possible routes to establishing a monopoly in quant trading. Here's one that seems to work really well. It's kind of hard to explain, so bear with me. Many trading signals reliably predict prices, but not strongly enough to overcome transaction costs (i.e. exchange fees, clearing fees, liquidity costs, etc.). A stock can moves up or down $.01 in the next period, your trading signals predicts the right direction 60% of the time, and transaction costs average $.0025. Unfortunately after costs, you'll end up losing $.0005 per trade, so it's not a viable strategy.

But let's say you've got three uncorrelated trading signals just like it. If you wait until all three point in the same direction, now there's a 94% of the stock moving in your favor. You easily clear the transaction cost threshold and neatly make $.0063 per trade. When you gather together multiple uncorrelated signals, the whole becomes worth much more than the sum of the parts.

This is basically how a company like Renaissance Technologies operates. It has hundreds of people working in silo'd groups. Each group contributes to the overall fund's strategy, but are largely unaware of what the others are doing. Alone any single group would probably not have a viable standalone strategy. That's a big deterrent to people leaving, starting from scratch is really hard. And there's only a few other major companies in Renaissance's league, where they're already strong enough to make money off of marginal signals. It's a chicken-and-egg problem. To become viable in the space you need to accumulate a whole bunch of signals, but to attract a critical mass of talent with signals you need to already have a viable strategy running.

Thanks.

Why didn't I think of that?

Uncorrelated, eh?

Snark aside, good comment. The lock-in associated with silos is interesting. Panopticon.

"Panopticon."

Not really, more like a classic clandestine cell structure.

@JWatts, agreed.

Keep in mind that this is also true of most hedge funds. The overwhelming majority of funds are a few dozen people working on some idea that fizzles. It's actually pretty similar to tech.

Big Finance makes similar profit-per-employee to Big Tech. But finance gives it all to the employees, which in tech most of it goes to the shareholders. Tech employees go along because they are promised that the employees can someday, if they are lucky, move up into that shareholder class.

VCs are really mostly MBA types chasing bubblish IPO markets.

Romney 2016!

I'm a Putnam winner currently living in Silicon Valley, working at Facebook, formerly started a couple startups and worked at Google, so I thought I should chime in to defend my industry. Maybe the average Putnam winner would make slightly more money working on Wall Street instead of in Silicon Valley. Who cares? Either career path is quite financially rewarding.

Are you going to live your life that way, optimizing huge career decisions to get a 20% increase in your expected net wealth? If that's your goal in life, perhaps Wall Street is right for you. If you think it would be more fun to work on something that millions of people use and love, or to try to build something new and different and ambitious even if the odds are stacked against you, then come to Silicon Valley. You will find plenty of smart people here, from Putnam winners to people who never graduated high school but have elite skills in their own way.

Looks like you went to Google moderately early, and then left to go the founder route. This jives with that the OP said.

The OP is 100% correct about "Not lottery ticket equity with four year cliffs, unlimited dilution and byzantine share classes." In general, startups are a suckers game, except for founders/exec brought in right before IPO.

The OP is being pretty elitist, in that Google/FB are in the top major employers in the US, but some niche quant firms are simply better deals.

Of course, in the final analysis firms don't matter too much, it is how great of a deal one gets from the firm one is in compared to deals one can get from other companies that matters. Being a partner at Microsoft is better than being a staff engineer at facebook.

Uncorrelated eh?

oops..ignore misplaced comment.

From the SV side, a lot of that rant was just that - a rant, exaggerated for effect. I've never heard of a 4 year cliff, especially on an initial equity grant. Typically it's 1 year.

My only critique of your comment, is that who actually likes FB? Most users dislike, if not despise it, and are only using due to some combination of path dependency and a trade-off of maintaining contact with old friends vs. the spying and annoyances.

@Kevin Lacker and some others,

You seem to be implying that much of what SV folks work on brings joy to the millions, while Wall Street is just about greed. Of course this is the standard stereotype that the Valley loves to sell. But honestly, a lot of what is done in SV is just as self serving as anything on Wall Street.

Facebook employs hundreds (thousands?) of smart quants whose main job is to put advertising in front of users, and to make as much money as possible while doing so. Apparently they even sometimes mislead their ad buyers to maximize fees, although I realize it was likely not a quant who made this decision. Is the world a better place because of a better Viagra ad placement?

Point is, taking our best and brightest and tasking them with selling advertising to maximize revenue is not on its face a better allocation of resources for the world than it would be if those same people worked in HFT, maximizing revenue while making the markets more efficient. People should do whichever work they personally find utility maximizing, but it is a bit silly to try to think that developing new trading algos is not a job where you try to build something new and different and ambitious even if the odds are stacked against you.

And feeding and maintaining these two tribes are why America can't have communist Hillarycare?

I think of this as The View From The Commanding Heights.

It genuinely depresses me how much smarter some people are than me.

I know that feel bro. But the ability to recognize what you don't know is a talent lacking in many people above and below you on the IQ scale.

Agreed, the first step in wisdom (as I might paraphrase Socrates), is admitting that in the long run, you don't really know very much in the scheme of things.

I assume you're of above average intelligence, or you wouldn't be reading this blog, which means there are steps you can take. If you want to feel better about yourself, learn a skill that is rare. In my case, I speak two foreign languages fluently. You'll feel a lot smarter relative to other people if you can suddenly break into Russian or Spanish when the situation calls for it.

Another possibility would be to master a musical instrument at a high amateur level.

Bill Gates (went public in 1986) was roughly an order of magnitude wealthier than Steve Jobs (went public in 1980) at the time of Jobs' death in 2011. Was this simply an idiosyncratic difference, or did expectations of wealth sharing from IPOs change over the course of the 1980s?

I dunno.

Joshua Rauh has been mining the annual Forbes 400 lists for insights. The first one in 1982 was dominated by oil men with tech guys only 3% of the list, but that soon changed.

https://www.gsb.stanford.edu/insights/joshua-rauh-what-forbes-400-list-says-about-american-wealth

That one is easy: Bill didn't mess up.

Jobs relinquished the CEO title, which caused him to be thrown out, which resulted in his emotionally selling a bunch of Apple.

Had jobs simply kept his shares of Apple from the IPO he (or rather his estate) would have a net worth of 87B dollars, which compares favorably with BillG's 81B.

I have worked in both industries. It is true that salaries are higher in finance but only in the short term. I had worked on two different quant strategies. Each of them vanished after 5 years. Code is bad because it is not supposed to last. Made good money out of them through bonuses. It got tough to find a third one. So I moved to start-up. Base salaries are lower but you can find a long term career way more easily. There is no barrier between both universes so no real arbitrage.

"I have worked in both industries."

I have too, and I have the math background.

The OP is massively overselling finance. It's not that different from tech. Failure is the usual end-state for quants and traders just as it is for startup founders. Most hedge funds start just like startups, and most end just like startups - in failure. In both sectors there are risky bets with high payoffs and there are stable jobs with decent money, but not private jet money.

I do agree with one thing from his post; VC is something you retire into, not generally something you aspire to.

I think it is important to understand the Q as well as the P. The Algo quant space (and hedge funds in general) is tiny in comparison to tech. Amazon alone hires hundreds of PHD's every year.

Which, I never disputed. The context of my post was that Peter Thiel made a comment about how no smart, ambitious kid would choose to go to Chicago. (Instead of San Francisco or New York). I actually agree with the overall point he's trying to make. But I thought the pedantic particulars were actually kind of ironic, since the 1% of the 1% of the cognitive elite by and large do go to Chicago.

Yeah, back in Chicago in 1987 my wife's boss's husband was a financial markets rocket scientist.

How big a deal are time zones for determining where you can be a quant? Chicago is only an hour off from New York.

Michael Milken moved from the East Coast back home to Encino, CA in the late 1970s, but he had to get to work in Beverly Hills at 4:30 am.

http://articles.latimes.com/1987-08-30/magazine/tm-5066_1_junk-bonds

A long time ago it was thought that satellite communications would allow financial markets people to live in Hawaii, but that hasn't happened probably because you'd have to get to work at 1:30 AM.

I met Milken in an elevator in LA once; I said nothing, it was after his conviction and he was going through a cancer scare that he luckily survived. I have no vested interest in what he did but it's clear he was, like Gov Elliot Spitzer, railroaded by his industry in connivance with the government.

As for location, I'm sure you're talking about 'in general' and not for HFT as per the now famous (and possibly obsolete) fiber optics line dug from Chicago to NY, as outlined in Lewis's "The Big Short".

Insider trading is a crime everywhere, Ray

https://en.wikipedia.org/wiki/Michael_Milken#Scandal

What % of kids are we talking about, exactly? How many fall into the category of "bright and ambitious"?

Since we're talking Math Olympiad or Putnam winner smarts, there are at most only a couple of dozen of those within the broad range (loosely defined). Especially once you subtract those who choose to become academics. Do the math.

If it's about Putnam winners, it's not true. It was hard to reconcile with the Putnam winners I know (I'm not in their league, but I cover seven years of recent winners), so I went looking for somebody who had data.

Turns out it's not even close to true. This is a list of Putnam fellows, rather than winners, but it's still a fairly qualified group.

http://www.d.umn.edu/~jgallian/PutnamCareer.pdf

Note that DE Shaw appears on this list a bunch of times, but that DE Shaw effectively hires these guys into their marketing department: they fund interesting projects that are only tangentially related to their core business. This is what happened to one of my friends, and I understand it's a nice life.

Wouldn't it be neat of smart people actually thought about how to make things people might want as opposed to allocating existing capital .00000000000000000000000001% more efficiently (if they're even doing that).

Indeed, would the industrial revolution have happened if quantitative trading had been a career option at that time?

It would be extremely neat.

And I await evidence that the HFT/quant types are actually doing anything useful.

How do you feel about true arbitrageurs? Are they parasitical, or do they add value?

I think they are justly compensated for creating significant welfare gains associated with the prevalence of the law of one price.

What quant guys are doing is not true arbitrage I don't think, but something kind of similar with a residual risk element for which they are also presumably compensated, but I find it difficult to quantify the risk pieces to make an overall assessment of the appropriateness of their compensation.

It was a generation ago that Lester Thurow offered that one of the problems of the age was the misallocation of human resources - so much to the counting house and not enough to the shop floor. That eventually ran its course after a fashion. The share of value added attributable to the real estate sector hit a plateau around 1985 and the share attributable to the financial sector around 1998 (the insurance business has been about the same dimension relative to the whole economy throughout the post-war period).

About 93% of the population lives neither in New York nor around the Bay Area, 99% of the population has nothing to do with tech in the Bay Area or finance in New York, and the majority of people who do (like the majority of people anywhere) are almost certainly hourly employees.

How long before Wall St quants are pushed aside by machine learning? Whereas information hungry, AI-driven trading seems to be a real threat to human coders, Silicon Valley depends on a much more human element in terms of understanding products that people will and won't want.

We're all luddites sooner or later...

Self-driving harvesters, construction robots assembling pre-fab products (driven to the site by driverless trucks) and, in the pipeline, management by algorithms which should cut quite a swath through the middle class. We clearly do not need near the levels of immigration which high-IQ people like Mark Zuckerberg say we do, and presumably he knows this. I think they're more driven by the need to import consumers, not workers. I have an acquaintance who helps structure derivatives ... out of pooled car title loans.

backslash - i. sorry about that.

Is this going to work better or worse than outsourcing "knowledge" work? My friend works at a firm that actually BANNED a certain offshore nation from participating in a project, due to quality issues. And this wasn't a high quality project: someone poured concrete without rebar. So getting banned from this project was really quite a feat.

That's a quality rant. Which reminds me, how is Tyrone voting this year?

I find this perfectly acceptable, and not even that ranty.

It is well known that finance has been skimming high IQ, but it might lead to a follow-on question: whether this is part of finance's misallocation of resources?

I see a few high IQ people above who tapped out, after achieving what? Curing a disease or just trying to shave profits off volatility?

Wouldn't *really* smart software types -- who are going to be very well paid regardless -- place more emphasis on factors other than max possible payout? If you're going to be in the top few percent of income and end up a multimillionaire, why wouldn't you put a lot of weight on the kind of work you actually enjoy most and the living environment you find most congenial? Pro athletes seem to have figured this out.

I think they do, and the OP is exposing an ignorance as to how top people are recruited and where they go. If you look at the best schools for technology, math, and finance, finance is not skimming the cream off the top. When they manage it, it's often for something unconventional and only tangential to their core business. DE Shaw is known for this, for example.

Can you give some examples of hiring Putnam winners into peripheral business at DE Shaw? Are there peripheral businesses besides the bioscience research?

Yes and yes.

Can you elaborate on what the ancillary business are? DE Shaw Research is the only one I can find.

I know a good school, IQ 145, CS grad who is sitting on the beach right now. I hope it's just a phase, but it seems in sync with a cultural thing, no? Falling labor participation.

Retiring by 30 or 35, pursuing hedonism (basic or abstract) rather than trying to foist some narrative onto your life when the world is already being subsumed by superorganisms?

In the realm of reality, though, quants are useless money siphoning criminals who, in the aggregate, contribute useless, pointless variation (risks) to the economy and society, at large. They, at best, contribute nothing, but in reality, exact destruction. The VC folks, on the other hand, make, at worst, a modest contribution to society. At best, they radically contribute making the world better.

Yeah, what kind of a world would this be without snapchat.

There's as much money-siphoning criminality in VC as there is in the rest of finance.

Re: "There’s as much money-siphoning criminality in VC "

Money siphoning VC:

Criminal level: no
As much as Quant class: No, and this assertion on your part is the stupidest thing ever said in all human history.

So, money siphoning VC: Sure, there's some.

I'm confused. It sounds like, in both the post and the comments, quant hedge funds are being compared to some conflation of venture capital, joining a true startup, and working for Facebook/Google, which are three wildly different things with wildly different skillsets and distributions of payout.

At this point Facebook and Google are big companies and attract Organizational Men (and Women).

There's nothing wrong with being an organizational type, even if they were in my day viewed with suspicion by good coders.

What I'm hearing here is that in order to get rich, you have to win the lottery.

For some definitions of "rich," for sure. The quoted text uses the old-time phrase “f-you money."

What's that today? 100M? Certainly higher than the few-M payouts common for workers in "successful" startups.

$7-8 million puts you at the $250K forever range. At that point, you don't need a job. You are only going to work, where and when you want to work.

If one owns his own house, and has $2mm - $4mm liquid (making allowances for local COL), only voluntary lifestyle extravagance would prevent him from considering that FU money.

So, to tie this all together oh Oracle of Arlington, should guys like Doug be raised in status? Lowered in status? Which is this post intended to do?

Completely false. Sounds like a post written by someone who's seen one too many Wall Street movies. In finance, the megafund Private Equity firms (the KKRs, TPGs and Carlyle's of the world) are by far the most prestigious. Much, much moreso than quant roles. Even out of undergraduate, investment banking M&A roles are where the top students go. Quant roles are seen as too restrictive, whereas on the banking side you can exit to whatever you want.

Other commenters have sort of touched on this, and I can't speak to the second or third paragraphs but I don't think Doug understands why people choose tech as opposed to finance. Everyone wants to be well compensated, but people don't idolize Zuck or Elon Musk simply because they're rich. Founders, programmers, etc. aren't daydreaming about a five bedroom house in Naperville, they're daydreaming about delivering commencement addresses at Ivys and keynotes at conferences, having 2M twitter followers, having the press hanging on their every word, to say nothing of creating something world-changing from scratch.

I also don't know a lot about the gritty details of HFT but I have a hard time believing that the core problems are an order of magnitude more difficult than the ones around machine learning, natrual language, etc, and require super-geniuses rather than ordinary geniuses.

Both financial and SV compensation (ultimately) relies on active equity markets.

From my experience, the "truly brilliant kids" tend to have other motivations than money. I've talked to plenty of VCs and people on Wall Street (because of my job), and there are plenty of bright people in both fields.

But the smartest people I've known (by miles) worked for: the NSA; in a physics lab; and themselves (writer).

I agree with a lot of what people are saying. However:
1). SV is one of the worst places to work in tech. You can work anywhere in the world, earn about the same, have half the cost and 1% of the BS.
2). You don't need venture money to launch a tech business, particularly with widespread cloud services. Just launch you business and collect ad money or whatever.
3). You don't need to join a firm to make money trading, nor do your trades need to be particularly high speed. I suppose you could benefit if people give you a lot of capital to work with, but not that much since you will start to move markets against you. Just work for two years in a stock tech job, save up, and you can double your income from that point forward exploiting the fact that markets are inefficient.

Dan is the most practical person here. Most of the other people commenting need to reevaluate how they measure their lives. Being status driven is just sad

The rant, right or wrong, reveals the idiocy of Thiel's comment.

There are, and will always be, specialized locations for all aspirations. Perhaps Thiel's comment holds true with a small addition: "San Francisco and New York are where the truly talented go... if they don't have intellectually specific ambitions, and are primarily motivated by money."

In the case of this rant, you don't go to SV or NY if you're interested in algorithmic finance, you go to Chicago.

If you're interested in robotics, you're going to Boston.

If you're interested in Medical devices, you're going to Minneapolis.

If you're interested in Aerospace, you're going to Seattle.

Even these categories have research center and company exceptions.

Thiel's definition of truly talented excludes those people who are the most talented in the 95% of subfields that don't cluster in NY or SV - anyone who has spent anytime with large groups of talented people, i.e. went to an elite college, know that the most talented people are almost always drawn to specific subfields, and rarely gravitate to NY or SV. Those locations instead attract B+ talent with A+ ambitions. None of the smartest people I know are in SV or NY, though most of the basic ones are.

Actually in robotics and medical devices the Bay Area is a better place, and is further along with sets of startups there. For aerospace, Southern California is the place to be.

In a nutshell, just about every technological area, California is the best place for research careers. Exceptions are air breathing aircraft engines in Cincinnati or Hartford and hydrocarbon chemistry in Houston or Wilmington. Everything else is a California strength.

Social has jpl, and a few others, but Seattle is still tops. That sf is the place to be for medical devices or robotics is a joke.

There are micro clusters nearly everywhere.

"The truly brilliant kids, Putnam winners, math olympiads, core kernel contributors, etc. disproportionately go the quant route." No, they do Math (or Physics, or sometimes Economics). Think Terry Tao, or Emmanuel Fahri.

This is consistent with my experience. In my circles, a job in finance is seen as a consolation prize for theoretical physics PhDs who don't make it in academia.

There is one relevant piece of informatiom in this rant. Wall Street pays 50 % more for same talent.

That is HUGE. Imagine how unpleasant or low status thise jobs must be viewed to create such a huge difference.

This rant is basically what it looks like when one group suddenly lose status in relation to another.

I can hear the world's smallest violin playing though. It's clearly much better for thr world that the smartest people are makig new things that matter to people, rather than putting ever more resources into capturing temporary price differences using technology.

This is exactly on point. We're seeing a lot about revealed preference here : people will take a large paycut to work on in SV rather than work on Wall Street.

The OP is right, and the difference between SV and HFT is the risk/reward balance - if you're young and smart and making your decision *today*.

If you want to get FU rich in SV (>$10M, for argument's sake), you have to found a company, get really lucky, and have a $50M+ exit. Or you have to join someone else's <10 person company that goes on to have a $1B exit. Joining Google/FB/Apple today isn't going to get you there, not at $300K/yr with SV cost of living.

Wall Street (i.e. a big bank) has a similar problem. There they have more predictability if you are very talented, but with a longer time horizon. You have to rise through the ranks of GS and then you can make millions by the time you're 40-50. But there's lots of political risks there, as those who get promoted aren't necessarily the smartest but the craftiest.

The HFT world, which is largely split between Chicago and NYC, employs many fewer people than the others but with much higher reward for the amount of risk. You can join as a Quant trader right out of school (undergrad or PhD) and if you're good you'll be making $500K/yr soon. If you're exceptionally smart and your trading strategies are top notch, you'll be making millions/yr. Your success isn't subject to the vicissitudes of viral marketing or corporate promotions, but just your own quantitative brilliance. You don't have to have founded an HFT shop, they hire lots of new talent and pay them a good portion of the profits they generate. That's why it is such a desirable destination for a Putnam/Olympiad winner currently deciding where to go.

There are few paths to becoming a billionaire through HFT but if you are the smartest one in the room wherever you go then it has the highest chance of getting you FU money. It's not for everyone but for the top of the top, which is what the OP was so right about.

https://twitter.com/samfaltman/status/720689042925301761

Sorry, last tweet is from a fake Sam Altman. Altman (of Y Combinator) only said this:

Sam Altman ‏@sama 14 apr.
Me: "Where do MIT CS students most want to go after college? Facebook? Google? Their own startup?"
Students: "Jane Street"

Any numbers for the percentage of putnam winners and IMO top ten for where they go?

I actually bet somewhere in academia and math/cs research where a lot are *really* poor, relatively speaking to where they could be guaranteed without the randomness of finance (its in medicine, actually)

"making “f-you money” before 40"

American Dream on crack.

In my view, it’s absolutely vital that we put our mark on making right investment since only then we will be able to gain profits. I am trading with OctaFX broker where I am able to gain great benefits in terms of investment which is with their mighty 50% bonus on deposit which is use able too, so that’s why it’s so good and it helps me big time with working and allows me to be successful and more importantly, I am comfortable with it.

Comments for this post are closed