Hacker News new | past | comments | ask | show | jobs | submit login
Day Trading for a Living? (ssrn.com)
541 points by ak39 6 days ago | hide | past | web | favorite | 366 comments

They have a reason to make that article: there is a lot of companies fooling the Brazilian people over this.

I'm a Brazilian software developer for the financial market. I also made some financial courses on the same university that the authors.

Brazilian stock exchange market has exploded in the last years and most of my friends decided to 'work' as traders. None of them with prior knowledge or experience in the market. All of the posts on Instagram every time they earn some money.

The cause for this fever was the marketing around 3 or 4 companies who sell the dream of financial independence through high-risk financial market lessons. Normally they offered courses to famous people for free and use their image to convince the rest of us.

One interesting thing: they normally sell the course with a private platform to trade the stocks included. I believe that they use that platform to collect data and operate against the traders that use that platform. They know the strategy that they are going to use (because they teach them), so it is easy to operate against.

> One interesting thing: they normally sell the course with a private platform to trade the stocks included. I believe that they use that platform to collect data and operate against the traders that use that platform. They know the strategy that they are going to use (because they teach them), so it is easy to operate against.

You’re assuming a complex strategy is needed when a simple one can easily suffice: simply pretend that you executed the trades, give the clients their profits, and keep the losses. In other words, the broker can take the opposite side of every deal and keep the spread and commissions. If, indeed, the client is trading based on no information whatsoever, then the broker wins and the client loses.

To improve this, the broker can throw out the best performing clients, since they might genuinely know what they’re doing.

I’m not suggesting that this is what any particular broker does, but it’s a strategy that works very well, at least in principle.

A business using this approach is called a Bucket Shop - it's both fairly well-known and explicitly illegal in most places: https://en.wikipedia.org/wiki/Bucket_shop_(stock_market)

It's also very similar to 'spread betting', explicitly legal in the UK and a large industry.

I have a hard time telling the difference between the “bucket shop” described in the lede there and a modern options trading market.

I guess the difference is that in a market third parties eat the spread by market making whereas in a bucket shop the market itself profits off the spread?

I'm not sure if it's the only difference, but according to the wikipedia definition, in a bucket shop there is "no transfer or delivery of the stock or commodities nominally dealt in", which would be one difference from the option market.

The bucket shops described in https://en.wikipedia.org/wiki/Reminiscences_of_a_Stock_Opera... operated by offering incredible leverage and using actual ticker tape quotes, which, at the time, were delayed slightly.

They offered the 'service' of being able to trade large nominal amounts at the last market price, without market impact, but had incredible levels of margin -- 50-1+ or something like that.

This meant that customers often hit margin limits, incurring additional 'commissions' to close out their positions.

Operationally, the bucket shop netted long and short positions, and either actually traded the residual or, occasionally, would use market orders (e.g. aggressive sells) to push the punters into margin calls, which they would then execute. They would then buy low and cover their earlier positions.

Jesse Livermore managed to beat them through astute technical analysis and through the installation of a direct telegraph line from the exchange, thus beating his 'brokers', who promptly kicked him out.

I don’t think most commodities traders actually transfer or take delivery of the commodities though. Sure there are some firms that actually are engaged in the commodities business, and use the futures markets to offset price risk, but most traders in the commodities markets are basically just betting on commodities prices.

The nominal dealing on a future's market will be on some paper that entitles the holder to the commodity delivery at a future date. When you buy oil futures, you are NOT buying oil since the oil doesn't exist (well I guess it exists underground somewhere, maybe), you are buying a contract to deliver crude oil and that needs to really exist. If the contracts don't exist that's a Bucket Shop.

It matters when things go badly wrong, because with a Bucket Shop you're left penniless, it was all imaginary anyway. Whereas with a futures market even if the exchange blows up those are real contracts you can sell to somebody who wants the commodity you were trading in, at worst a mild inconvenience and a small haircut to your final profit.

You may be confusing the exchange with a broker. With a serious commodities broker, you’re using the services of the broker to trade on an exchange, e.g. the CME. The broker does not particularly care whether you make or lose money, so long as you don’t actually go negative.

In the US, securities brokers are very heavily regulated — see “regulation NMS”.

I don't think so (I am not an expert though so it's possible), my point was that even if the exchange, not just your broker, blows up, the thing you were buying and selling is a real thing anyway, unlike in a Bucket Shop.

Interesting, thanks for the link. I wish there were some comprehensive list or taxonomy on these kinds of scams.

There was a really interesting talk I saw a few years back that dealt with this sort of thing. Very much worth watching:


When I worked in Forex (~1999-2003) this was common practice, at least where I worked. Bring in customers, teach them the bare minimum, and let them trade with 100:1 (and sometimes 200:1 leverage). Even better, because 99% of customers lose, we would "assume the risk" of most trades, accepting their trades but not offloading them to larger banks (ABN, Deutsche, etc), so their loss was all profit. If someone demonstrated they were a decent trader (very, very rare), we would set a special bit on their account and, from then on, offload their trades to major banks at better spreads than we offered to the customer, making 5-10 pips per transaction. Most of the "decent" traders traded in huge lots (10-100 millions), so making 5-10 pips per transaction was substantial. Long story short, you make money on the crappy traders and on the good traders, and the good traders were even better because they would keep coming back.

> To improve this, the broker can throw out the best performing clients, since they might genuinely know what they’re doing.

In a parallel space this is happening in the UK with bookmakers. Winning punters either have their accounts heavily restricted, i.e. only able to place small value bets, or their accounts closed entirely.

Certain states in Australia had to introduce legislation to guarantee a bet size the bookmakers must stand. From memory it was $500 on country race meetings and $1000 on metropolitan meetings.

  To improve this, the broker can throw out the best
  performing clients, since they might genuinely
  know what they’re doing.
This practice is pervasive in sports betting. It's a common practice to ban winners, especially in the US and UK.

https://augur.net is an Ethereum-based betting market that will help traders/bettors avoid these kinds of malicious practices. Augur v2 is being developed now, you can expect it to be Really Cool by about April 2020. Millions of dollars will be wagered on Augur for the 2020 election.

Betting could very well be the killer feature for Ethereum.

So far betting seems to be the ONLY feature for all things blockchain. In China quite a few of my friends went to make games on the blockchain during the wave of cryptocurrency mania. That didn't work out so they then went on to make (illegal) gambling games on the blockchain, although that didn't really workout quite as well either (apparently something to do with generating random numbers on the blockchain)

I worked on a Bitcoin sports betting site. Been there, done that.

You don't have to throw out the clients that perform well! You charge them fees and send their orders to market.

Maybe front-run them a bit if you're into that.

>If, indeed, the client is trading based on no information whatsoever, then the broker wins and the client loses.

If this were the case, wouldnt both sides just make 0$ on average over a long period of time? (or whatever the avg growth of the market was during this period)

I guess the broker would earn fees, but that still doesn't exactly sound lucrative

For the kind of trader that clicks on a computer all day, fees and spreads (which are not always easy to tell apart) can easily dominate price movement.

I urge people who care at all about how stock exchanges actually work -- it's substantially worse than what you describe -- to read 'Flash Boys: A Wall Street Revolt' by Michael Lewis.

Presumably this strategy would land the operators in jail if enough of their traders successfully predicted a large enough market move.

The best way to get rich in a gold rush is to sell shovels.

Every time someone comes up with a "get rich with this training/tool/etc." it's very unlikely to make you rich, because if it would work that great then why isn't the seller of the training/tool using it to get rich, instead of selling whatever they're selling?

Well, I’ve seen a few people blogging about how they make money by blogging about how to make money.

There's a classic one where one places a classified ad that says: "How to make money at home! Send XXX $10 for information."

Send them $10, and you get a letter that says: "Place a classified ad that says: How to make money at home! Send XXX $10 for information."

This was probably 30 years ago--make $100's "Stuffing envelopes."

Exactly. This goes double for when we are talking about investments.

If the proposed strategy really really works, then you would get investors and offer them a share of the pie. "Teaching" them accomplishes nothing and reduces your profits.

Bitcoin ASIC manufacturers are known to mine Bitcoin before selling their machines.

Thats what i did.

Welcome to 1999 :-) Or China 2010 or so. It almost feels like there are people who look at these events, figure out who made the money, and then work to recreate them in a new environment where they can be in the position to make all the money.

My experience from the dot com day trading frenzy was that there were three kinds of investors;

Those that had no clue and just followed a bunch of investment "tip" sources and did what ever was the investment of the day (these people were essentially gamblers, and like gamblers were up one day and down the next, only to eventually bottom out).

Then there were the folks who had read up on how the markets worked, maybe taken a finance class or two, and read a bunch of Morningstar reports on various funds and their strategies and tried to create some sort of meta strategy that was a mix of the funds they found impressive. These people had good days and bad days and over the course of a couple of years basically matched the S&P500 or other widely diversified stock indexes in gains. What I learned from them is that if the market in general is up 9% and you're up 10% you are only doing slightly better than the market, even though you feel like "hey I'm getting 10% a year, I'll double my money in 7 years!"

Then there were the very serious folks, these folks read annual reports and 10-Q statements and prospectuses. They kept a databases of people who were executives, board members, and advisors of different companies. They consumed four or five different regional news streams (usually London, Tokyo, New York, Chicago, and Washington) They kept indexes and stock price histories in their own databases and mapped current events to stock motion. They broadly characterized every stock they watched closely by who managed the entity, what markets it was most effected by, and least effected by, and what government policies could help or hurt it. I'm sure if they could afford a Bloomberg Terminal subscription they had one of those too. They did well for themselves but they invested 80 hours a week into doing well.

Not surprisingly there are lots of people in the first group, fewer in the second, and fewer still in the hard core group. I personally see myself in the second group, and over the years my own portfolio has done slightly better than the market.

A unique strategy one person I knew took during the dot com bubble was to convert gains into "things" on a regular basis. At one time he had about a dozen different Porsche sports cars, two houses, and some acreage in the Livermore valley that he rented out as a vineyard. He did pretty well selling that stuff to recover losses from the crash.

Haha, I invested in the "shovel" companies that were making real profits (such as Cisco). They got caught in the secondary die-off, when the dot-bust companies stopped buying shovels.

Yeah, I’m in the second group. I use a diversity of buy and hold strategies (picking individual stocks based on fundamentals or personal brand experience, buying certain sector ETFs, etc). The efficient market hypothesis index ETF promoters think I can’t beat the market, but if they’re right it doesn’t matter, since I make sure to keep my transaction costs low and only use low expense ratio ETFs, and I have a diversified portfolio so I’ll just end up matching the market on average.

So, let's see: we have a surge of stock market speculation driven by day traders seeking to get rich quick, in a country with a growing influence from Protestant Evangelicals preaching the prosperity gospel.

Brazillians, you can ask any American about this. We've been there. It does not end well.

It's no reason to discount this study of course. It just means it should try to be reproduced by following large samples of day traders in American, European and/or Japanese markets. My hypothesis is that you'd get pretty similar results. Day trading always seemed like gambling to me. It's just fancy Vegas.

What you believe the platforms are doing seems similar, if not the same, to “front running”, which is not allowed in the US

Hey. Brazilian here too. How can I get in touch?

For anyone interested in a great story about a world class mathematician who built a company that has used algorithms to beat the market for the last few decades, James Simons has an impressive record:




One other thing, the Simons Foundation funds Quanta Magazine, whose articles frequently appear on HN:


Renaissance Technologies did not use algorithms to beat the market... they used tax schemes to do so:


When you factor in the money they saved from dodging taxes and how they managed to compound the growth on that money, their returns end up being no more impressive than any other hedge fund.

Even if it turns out that what they did wasn't illegal, the bulk of their profits beyond what would have been made just investing in an index fund came from their tax scheme rather than from any kind of insight into the financial markets.

You have absolutely no idea what you’re talking about. RenTec is the most profitable money making machine the world has ever seen (and likely will see). RenTec’s Medallion fund has an annualized return for over 30 years in excess of 35% net of fees. I’m not sure what their management fee is, but their performance fee is like 45%. This means they are making like 80% returns YoY before fees.

I suspect you don’t work in the industry, because what you’re saying is absolutely ludicrous. The S&P 500 lost like 40% in 2008. Meanwhile Medallion doubled up after fees. So a total return of 145%. What you’re saying is simply not possible with a tax loophole.

There’s not a single hedge fund analyst or PM on the Street who doesn’t think that RenTec is anything less than amazing. No one thinks they are as good as those guys, no one. If you want the real story about Simmons and RenTec, check out the MIT Sloan fireside chat with him on YouTube.

I'm not in the industry, but I wonder: did people have the same kind of reverence for Madoff?

Madoff was a ponzi scheme that relied on new investment (since it was simple pyramid). Medallion fund limits outside investment (so there is no outside source of money) and has operated for many decades (ie more money was withdrawn from it than put in).

I know nothing of Medallion, but I'm curious if only private insiders can invest, how do you know they're not just lying about how much they return?

Such returns would not be possible if they grew to a certain size. By closing the fund, they can keep it small enough to limit scaling issues.

That's illogical because people don't get paid in ROI. They get paid in absolute dollars. Earning a 30% ROI on 200 billion dollars is better than earning a 30% ROI on 100 billion dollars.

But they do have other funds. It's simply they want to keep their money in higher return fund and other people's money in the bigger, lower return fund.

My point is that once a fund grows to a certain size, ROI at that scale becomes impossible. I wouldn't be surprised to learn that some of their funds have investment restrictions and mandatory distributions.

It allows family and former employees, so it is actually A LOT of people. Kind of hard to keep a conspiracy with 100+ people for 30+ years. Why would all these people lie?

All the family, employees, and former employees have a maximum investment. Every year they withdraw funds so that the principle never exceeds a certain threshold.

You can't just conjure a 45% ROI out of nothing.

It either means inflation is incredibly high, productivity growth is incredibly high, there is some exotic "dark market" that most investors cannot access (probably made out of dark matter) or other people are losing a lot of money.

Except inflation isn't high, productivity growth isn't high and I just made up the "dark market" so only one explanation remains.

Like they’re very very good at what they do? Simmons is a genius ans the people that work st RenTec are some of the best and brightest minds in the world.

Even so, I must admit, their profability and success is outstanding. But then again, so are Apple’s and Google’s. The only difference is thst no one knows how RenTec makes buttloads of money.

This is another reason individual day traders lose - it's a zero sum game and in liquid markets you are competing against outfits like RenTec.

Beta (by definition) is not a zero sum game, but alpha is.

> Renaissance Technologies did not use algorithms to beat the market...

This is false. RenTec generated outsized returns for a decade prior to entering into the derivatives with Barclays and DB beginning in 2000.

> the bulk of their profits beyond what would have been made just investing in an index fund

This is also false. For example, RenTec generated 99% return in 2000 net of fees, while the S&P 500 lost ~8%.

From what has been published, secrecy within Renaissance is so extreme that perhaps only 2-3 people actually know what is going on. So for an outsider, I say it's pure speculation.

And I speculate that their outsized gains are due to market manipulation, front running, or other insider activities (as in, illegal/unfair).

No one is smart enough; no algorithm or model is future-proof; nobody gets returns like this unless they define the scenarios themselves.

> their outsized gains are due to market manipulation, front running, or other insider activities (as in, illegal/unfair).

as a former insider at a different, less successful-than-rentec-but-still-successful firm, I doubt it. They have better data, technology, and employees than most other market participants. So then the question is how did they get there:

> No one is smart enough; no algorithm or model is future-proof

I agree with the sentiment, but the way it worked was in the early days, it was much easier for one person or a small team to get those advantages. Then, as the market matured and simple algorithms became less profitable, they made new, more complex ones. It became harder for new entrants to jump-start the whole process from scratch. Simons started rentech in 1982. Teleport that Simons to today and his fund probably wouldn't even get off the ground. He would still be a genius, but a billionaire? I don't think so. As you say, no one is smart enough. It was right place/right time, and they built on their advantages.

Question since you are informed about them and the space...

I had read this[1] and seen some other things about his activities.

Since Mercer and his family seem deeply wrapped up in something that has some worrying ties to hostile foreign nations, a company who has been recorded saying they've done REALLY shady things along those lines, etc...

How feasible is it that RenTech used similar technology and god knows what data to directly impact the geopolitical landscape, and traded off of that? If you can for example...create major disturbances in certain areas, or draw certain attention to various things at a certain scale, could that impact markets?

I want to clarify that I know absolutely nothing about the regulations and mechanisms in place to catch such a thing with the SEC, etc., and don't want this to come off as a conspiracy theory. It's just...with what's come out so far the conspiracy is kind of writing itself, so I wouldn't be surprised if something like this were at play.

[1] https://www.institutionalinvestor.com/article/b17q91wjnnr68x...

It is feasible, although I would be surprised if it something like this were at play. For one thing the other CEO Brown and Simons himself are on the opposite side of the political spectrum. That's not any guarantee but my point is that these organizations aren't monolithic...if there were geopolitical impacts hopefully someone would leak. But even if rentech were completely amoral, they probably still wouldn't do it just because the risk/reward asymmetry is too great. Sounds like it'd fall more to the FBI than SEC; much higher penalties. Too much of the rest of the business is profitable. These people have a lot to lose. And when hedge funds engage in bad behavior, it tends to look more like this [1]. Perfectly legal, yet still harming society, in my (less informed) opinion.

What keeps me up at night is not rentech but state actors, like you said, hostile foreign nations. They have the resources to pull it off and plenty of motivation...trading profits would just be icing on the cake, really.

Here's an amateur-level manipulation example you may find interesting [2]. Could rentech do this more subtly with 10x impact? Like I was saying, it may be feasible...still kinda doubt it. If this happens my guess it would be a fund that was no longer profitable, on it's way out, they have nothing left to lose, all this computing power laying around...

[1] https://www.bloomberg.com/opinion/articles/2019-02-27/windst...

[2] https://www.bloomberg.com/news/articles/2017-04-21/german-so...

> So for an outsider, I say it's pure speculation.

And you assert this on what basis? There's enough that's been published to get some broad idea of what happens within the firm.

> And I speculate that their outsized gains are due to market manipulation, front running, or other insider activities (as in, illegal/unfair).

If they were engaging in illegal activities, it's very likely the firm, which is undoubtedly scrutinized by regulators, any such activities would have been discovered.

> No one is smart enough; no algorithm or model is future-proof; nobody gets returns like this unless they define the scenarios themselves.

RenTech constantly updates its algorithms. Alpha decay is a well known phenomenon. The firm utilizes sophisticated risk management techniques in order to avoid drawdowns, in all likelihood. Assuming this is the case, they're able to more effectively compound returns while simultaneously levering positions. Here's an example of a risk parity strategy [0], which may help explain how risk management works.

[0] https://towardsdatascience.com/ray-dalio-etf-900edfe64b05?_b...

There was a video interview with the founder where he explained pretty much how they do it. The employ a lot of bright PhD maths/physics types, get them to come up with all the algorithmic strategies they can think of, run tests with historic data and live trading to see which ones work and then scale up those.

There isn't one smart guy or one great strategy - there are dozens of smart guys and loads of strategies and they can win because they outsmart the city types.

That sounds too much like this:


As I remember, beaucoup back testing, hilarity ensues anyway.

The issue with LTCM wasn't just that they didn't test sufficiently - the problem is that traders could take positions without supervision and any accountability. My understanding is that any firms that survived that period (and I believe they were all pulling the same shenanigans as LTCM, just maybe to a lesser degree) now have risk management because unlike Lehman/Bear they can't just foist that risk off onto the public market. No one at the head of a multi-billion dollar hedge fund wants to stop being there if all they have to do to endure is pay for risk management.

*edit: I'm not a serious student of this aspect, but I recall that none of the top independent hedge funds took a bath in the 2007+ collapse (that is, those that weren't in-house funds from a major wall st. company). They all had their risk management in order and all did pretty well in buying distressed assets. Some like Bridgewater really managed to grow non-stop right through that.

As I remember the book, they back tested quite a lot. It's just that whenever they went live with trading reality changed on 'em. Go figure.

They were exposed to a margin call. Trading reality changing or no, if you can get taken out by your prime broker at the end of the day, you need to account for that too and they didn't.

One thing backtesting misses is the other market participants may see what you are up to in live trading and try to take advantage. That was a big factor with LTCM.

They could lie about that too and just use linear regression.

A linear model on a heretofore unknown predictor is basically how all hedge funds make money.

Coming up with the predictor is often the hard part. For example, take the tweets of a (sane) president and run sentiment analysis on it. If it is positively correlated with mentioning an equity, the sentiment of the tweet might be a good linear predictor of the stock price.

The math is simple once the feature is well defined.

Feature development is the current frontier, as I understand it.

That's my point. They hire the smart people so they can really abuse the simple stuff.

No mention of leverage?

I'd take leverage for granted but it only works if the underlying bet you are leveraging is in your favour.

The article says that this scheme occurred during specific years.

During the 1990’s they also beat the market.

“Medallion, which is open only to current and former Renaissance employees, has generated returns of about 40 percent after fees for decades by using computers to spot market patterns. It’s distinct from the funds Renaissance makes available to outsiders, such as the Institutional Equities Fund.

Medallion earns most of its money through short-term trading of securities and other assets. Such earnings typically get taxed at the same marginal rate as salary. The tax code rewards longer-term investments with a preferential, lower rate.

The dispute centers on transactions the firm carried out with Barclays Plc and Deutsche Bank AG between 2000 and 2015 that had the effect of transforming short-term trading gains into long-term returns. Rather than own securities directly, Renaissance instructed the banks to buy and sell them within a portfolio of assets. It then bought an option from the banks tied to the portfolio’s performance”

I’m confused. Doesn’t the recipient pay these taxes, so it wouldn’t show up when comparing fund returns?

I'm not a finance guy, but I think the idea is this:

Ren says "we bought an option and held it for a year". The option's counterparty was a bank, I guess, so they just form a company to hold the bag, and the company that is created reports to the IRS that they don't have any assets, just a basket of assets to offset the option they're responsible for.

At the end of the contract (and there can be more than one of these going on at the same time) the assets can be sold, and would be sold, in order to pay for the option contract which was now held for a year. Some money would go to the broker for executing the trades over time, etc. but much less than the taxes that would otherwise be paid.

So, the part that has the tax man scratching his head is that if Ren was directing the buying and selling of the assets in the basket constantly, is that really holding on to an asset, or is it just a lie? They say they got a tax lawyer to sign off on it, so it goes to court to see if it's a defensible position. If not, they'll probably pay a slap on the wrist and won't get to do it again.

It's not done in isolation IIRC. This sounds a lot like how I understand swaps are done, so it may be that this is only slightly unusual.

> When you factor in the money they saved from dodging taxes and how they managed to compound the growth on that money, their returns end up being no more impressive than any other hedge fund.

That's a bold claim that you should be prepared to defend. If not, you should remove it.

you've been rightfully skewered by other people but I just want to point that this is what happens when someone reads one article about a controversial thing they don't otherwise know anything about and then goes around presenting that finding as if they're some kind of expert.

Someone will always beat the market, because that's what we're looking for.

I think it was on "Thinking Fast and Slow" that I read it's pointless to analyze the stock market winners, since it's a random process. Imagine there are 30 million entities that own stocks in the US - individual, companies, funds, etc... If on a given year half of them did better than average (with some rounding liberty):

  1st year: 15M better than average
  2nd year: 7.5M better than average
  3: 3.25M
  4: 1.6M
  5: 800k
  6: 400k
  7: 200k
  8: 100k
  9: 50k
  10: 25k
And so on... after 20 years, probably 25 traders will have beaten the market for 20 years straight. Next year, same thing, 25 traders will have beaten the market, and so on. In the US, there's always someone who has beaten the market over the last few decades - there's nothing special about them. Randomly picking stocks would give you the same end result. But since we're storytellers, we build an explanation and go with it.

Okay, let’s do the math buddy. Let’s assume the null hypothesis (the market) has a mean return of 10% and vol of 10%. RenTec’s Medallion Fund has made 80% annualized return before fees in the last 20 years. Assuming the market is a random walk (it’s not), the chance of RenTec making >50% (I don’t think they’ve ever made less than that in 20 years) every year for 20 years is...

3.486 x 10^(-91)

The actual probability is a lot smaller btw.

Why do people keep posting this tired argument over and over? It’s not true. There’s no evidence that it is true. The math strongly suggests (see above) that you can beat the market.

>Why do people keep posting this tired argument over and over?

Sour grapes. Easier to pretend _nobody_ beats the market than honestly admit you're not willing to put in the effort to constantly discover new alpha. And it's fine to not be willing to do that -- it takes a lot of effort and skill to discover alpha and most people will be better off improving their skills in their day job and just tracking indexes.

Doesn't mean it's impossible to beat the market, especially as a small investor where you have some advantages over larger players (your orders aren't large enough to move markets and reveal valuable information)

Who said it’s a random process? Stock prices don’t go up randomly over the long term.

Joel Greenblatt had 50% yearly returns for a decade.

Renaissance Technologies uses algorithms.

Also, I don’t think you would expect someone to beat the market every year, but over a ten year period, for example.

An important difference. You would expect someone to beat the market every year for ten years just by chance alone, you wouldn't expect it to be any specific person. I expect someone is going to win the lottery, I just don't expect that someone to be me.

Yes, given enough traders, you very much would expect some of them to beat the market consistently, over a decade or more.

I’ve got a roulette “system” too. Trust me, it even has algorithms and machine learning. I’ll sell you the book.

The discussion reminds me of a fun exercise we did in one of my B-school classes: everyone stands up and flips a coin. If you flip heads you sit down and stop the game. Everyone who flipped heads flips again, with the new tails flippers sitting down. At the end when there’s one person left, the prof interviews him, asking “how did you become so skilled at flipping heads?” and “what advice can others take to get as good as you!”

This is what we are doing when we admire stock pickers and try to figure out what their secret is.

Well Claude Shannon and Ed Thorpe did figure out how to beat roullette. And BlackJack. And then Thorpe figured out Black Scholes and made a killing in the markets trading warrents.

See my above comment about how statistics does not bear out your fooled by randomness theory.

I don’t have a dog in this fight but Ed Thorpe has one of his funds liquidated via RICO. He’s usually used as an example in the positive case for “only bad behavior gets outsized market returns” argument.

Scholes & Merton got famously wrecked in the markets.

This is the perfect example of broken telephone. (If you don't know the full story and fail to post any source)

The reason why they got the RICO.

"From 1969 to the end of 1987 the amount invested rose from $1.4 million to $273 million. Its limited partners saw their wealth grow at 18.2 percent annualized after fees. PNP had no losing years — and not even a losing quarter."

The regulators thought that it was a Ponzi scheme based off that information or did they...

"But most knowledgeable observers believed that the charges — filed under the Racketeer Influenced and Corrupt Organizations (RICO) Act — were also intended to get the Princeton Newport principals to testify against Michael Milken, the controversial “junk bond” trader who had upended Wall Street conventions and who had dealings with PNP."

The regulators thought that it was a Ponzi scheme, when in reality, they mastered the tax code and created losses via hedging, which was legal at the time.

"The resulting charges were related to trades that PNP had made to create losses that would offset corresponding gains that arose during the firm’s hedging maneuvers."

"The Princeton Newport attorney, Theodore Wells, a Harvard Law and Harvard Business School alumnus known for high-profile white collar criminal defense, argued that the trades were allowed under IRS regulations. They even had a former IRS commissioner, Donald Alexander, prepared to testify in their defense. The judge would not permit it."

>I don’t have a dog in this fight but Ed Thorpe has one of his funds liquidated via RICO.

First off, his name is Edward O. Thorp https://g.co/kgs/jfpjL1 not "Thorpe"

>Ed Thorpe has one of his funds liquidated via RICO.

Please post a source because almost all of the charges were overturned.

"Ed Thorp, who was not greatly affected by the criminal charges, was able to restart his hedge fund activity. In 2012 Thorp’s net worth was estimated at $800 million."

All of this information came from this source.


See also the Eudaemonic Pie

If you think everything is just random, you should educate yourself the impact of superior information sources or technology on producing returns. Are you telling me someone who runs the fastest market data feed between Chicago and New Jersey, doing arbitrages between S&P futures and S&P ETFs is just getting lucky over and over again, and it will eventually be revealed that they just won 30,000 coin tosses in a row?

Correct. There are numerous studies showing professional traders (individual or companies) are in fact no better than randomly picking stocks. Just the sheer number of players and variables will eventually produce winners, then we will justify why it happened in the first place.

Do you understand that 'studies' are not necessarily reality? I know the "I fucking love science" crowd has replaced priests with professors, but professors, their data and their logic are also highly fallible.

How about you go ahead and mock up a probability model about how a company like Jump Trading can make 100s of market neutral bets every day for ten years, and end up being lucky to make money on 95% of those days. The probability of that happening due to randomness around a 50% probability on each trade is probably lower than 1 in the number of atoms in the universe.

I just watched a video today via this thread talking about 10k SEC documents.

Pull up Goldman Sachs, go to page 95ish:

https://www.goldmansachs.com/investor-relations/financials/c... (p. 95)

https://www.goldmansachs.com/investor-relations/financials/c... (p. 94)

look at the distribution chart. That's days where rev was positive vs. days not.

In the video they covered 2014/2015/2016 which I why I went a looked at 2017/2018. They are right a LOT. And year over year.

I don't think it's magic. I think they are paying attention and using information to their advantage.

This is a contradiction. You first assume complete market efficiency (assets are priced in) to support that the market is random and unpredictable. Then you say that a hedge funds' payroll is no better than random guessing (something you could scale up without hiring anyone). This means that quants are grossly overpaid (assets don't reflect their true price) and that the trader's market is highly inefficient... That's having your cake and eating it too: a rational irrational market.

And it is a hypothesis that can't be disproven or proven, since, in a round-about way, it concerns non-computable concepts like Kolmogorov Complexity/Randomness (The efficient market hypothesis basically states that the stock market is an optimally compressed computer program, there are no discernible patterns left to reduce the "file" size).

There are numerous tried and tested ways to beat the market, including:

- Have more information than other players - Have better models than other players - Make faster decisions than other players - Have enough energy to perturb the system and predict the outcome (this is the big one that is out of reach of most individuals and non-Physics PhD's).

Survivor bias is of course a very real phenomenon, but like other incomplete information games that can be won, or even solved, algorithmically: Poker, international diplomacy, ..., we would not conflate pure luck with the real skills required to play those games consistently well.

Think about how the underlying market works though. Is success or failure actually a random process? Or are cause and effect relationships happening? Isn't it possible to identify traits of a good business that others aren't recognizing.

The math, at least when it comes to the Medallion Fund, says that you are wrong. The odds of being able to achieve 40%+ returns (net of very steep fees) over 20+ years are astronomically low. In fact, the odds of simply being in the top 25% of hedge funds for 20 straight years, through “randomly picking stocks” as you put it, are 1/(4^20), which is roughly 1.09 in 1 trillion. There have been less than 25,000 hedge funds in US history.

Yours is the kind of folksy explanation that appeals to the masses. It seems like the kind of thing Bernie Sanders would say in a speech. But in this case it leads to an astoundingly inaccurate conclusion.

This is the correct response.

For whatever reason, people seem to think "50/50 chance of beating the market" = 50% chance of generating 40% return on any given year.

Try randomly picking stocks over the last 20 years, and run 1 billion simulations and see if you get anywhere near that (even letting you have survivorship bias for free)

> people seem to think "50/50 chance of beating the market" = 50% chance of generating 40% return on any given year

Yeah, quants like to coyly say they're figuring out where pennies might drop, and then they bring in leverage. Without the leverage, they'd be beating the market by less than a percent in so many cases.

It's not really random, the value of stock has more to do with popularity of stock than the data people like to analyze.

As long as people keep buying a stock, it's price will continue to go up. Same thing the other way.

Now also keep in mind I can sell stock and you buy it, and we can both make money as I bought my stock lower than you.

Also, the people I've known who call themselves day traders, were inevitably people who weren't too bright and were unemployed with independent income (mostly trust fund babies, but some others who had supplemental income) - From my own personal experience, these people don't even know what EBITDA adjusted earnings are.

If a world class mathematician says he will do something, I see the probability of success wildly higher.

It's like when you go to a writing class, and someone inevitably is admonished for doing run on sentences, and then they say "well Charles Dickens did it". Why yes, he did. But you probably aren't going to have Charles Dickens levels of talent.

In my entire life, as far as I know I've only ever met one person who played the stock market manually and did well at it. It was almost a hobby for him. He enjoyed analysing individual businesses very carefully, and would look for clearly financially sensible opportunities that were effective at relatively small scales but overlooked by the big players.

A favourite type of investment that came up now and then would be a stock that had seen a sudden sharp drop in share price triggered by some announcement that sounded scary but actually made perfect sense if you understood the industry, so it panicked investors who were only looking at the numbers and left the price-to-book low or sometimes even below 1 at the time he bought. That's probably as close to a certain profit as you're ever likely to find trading stocks.

In any case, he almost always invested in "real" businesses and for the long term. He would set stops to protect himself, but he wasn't really a day-trader in the sense we're talking about here.

The last I heard, his average annual returns were something like 15-20% before taxes and fees, which is a pretty impressive record given he'd maintained it for many years and through both bull and bear markets. I doubt most people could do what he did -- the amount of patience and discipline he had were extraordinary -- but it was fascinating discussing it with him, and strangely satisfying to watch him do well by going with common sense and considered choices against a market that was temporarily confused about something.

This is warren buffet style value investing.

"Day Trading" is actively taking positions with the intent to sell in the "short-term".

Yes, exactly. My point was just that anecdotally, literally no-one I know who has attempted "day trading" style investing has made good money from it over the long term. In contrast, the one person I know who has attempted "value" style investing and done it seriously actually has made quite good money from it consistently. This is in line with the article's conclusions, but in another context: it looks like an individual trying to day-trade in today's markets is probably setting up to fail.

I don't doubt the headline, but "We observe all individuals who began to day trade between 2013 and 2015 in the Brazilian equity futures market" seems like a weird way to come to that conclusion.

20,000 people trading a broad based equity futures market over two years isnt a bad study

But yes a study of individual stocks and certain sectors would be better

Or following those individual accounts to see how they grew based on everything it traded would be interesting

Why weird?

Because that is not a representative sampling of the day trader population. Assuming the techniques of the study were sound, at best we can conclude whether or not day trading Brazilian futures is a sustainably profitable undertaking.

Wouldn't you need a basis for believing that day trading Brazilian futures has unique properties not present in other markets to assert this? E.g. the volatility or spread is significantly larger than European futures or the information flow is more tightly de/regulated etc.

Is there a basis for believing this?

Making an assumption about the population based on a small non-random sample is usually the belief that needs to be justified by evidence.

20,000 people is not a small sample. Sampling bias is still an issue with large populations, but generally significantly less important.

That belief seems like it would be obviously justified. The same would be true for pretty much any geographically restricted set of day traders on a specific exchange tied to the geography, likely even for huge exchanges. There are huge effects from correlation with that country’s equity markets, currency policy, political events, etc. to such a degree that it would be unreasonable to assume they aren’t present until they are proven present. That’s reversing the burden of proof in a terrible way.

The number of confounders you need to control for before you could confidently not make this assumption is staggering.

There is a distinction between structural confounders and the variables which you are literally backing yourself to predict when you're a day trader. You've mentioned two in the same sentence which I would consider to be polar opposites: political events and currency policy.

For example if market data operated on a significant lag in Brazil for amateur investors, but didn't in the UK, this would represent a structural disadvantage which might make it impossible to beat the market except through dumb luck in Brazil.

It wouldn't mean it's definitely possible to beat the market in the UK, but it would be something one would want to control for in a study which makes a global prediction based on data from a sample in one nation.

However political events, as I understand you to mean, are one of the things most traders are specifically backing themselves to predict in a futures market.

So my question is: what's the basis for believing that there are structural impediments to retail day traders which are present in Brazil but not present in the rest of the world? Everyone in this thread is absolutely certain they exist, but nobody seems to be able to outline what disadvantages are present for a retail Brazilian futures trader which are not present in the UK outside of red herrings like "Brazil is less stable".

You are making a large misunderstanding. The confounders are not about what the day traders themselves are using in their own models. The confounders are things that affect researchers studying whole cohorts of traders.

Day traders may well believe they can predict political events in Brazil effectively. And yet, there may be some quality about Brazilian political events that impacts the cohort of Brazilian day traders systematically differently than say Korean political events affect Korean day traders.

For example, maybe a new cabinet member was just appointed in Brazil who likes to leak new political announcements early to an old business partner with connections at several large banks and trading shops, and so regular day traders are at a significant disadvantage in just this one market.

Or maybe during the time period of the study, there were political protests that disproportionately involved young people, meaning the average age (and perhaps skill level) of the observed cohort of traders was artificially too high just in the window of time of the study. Or any number of possible things like this.

You can’t assume these effects don’t exist when looking at different cohorts that have real reasons why they might not be probabilistically equivalent to a more general population. You have to actually account for the possible sources of confounding (in this case the country of the exchange), either by proposing some prior counterfactual model, or by collecting data across different cohorts and explicitly controlling for the confounder with whatever type of model you are fitting.

To be clear: none of this requires you to know in advance what the confounders actually are, and indeed in most causal inference models you literally cannot know what they are. The best you can hope is that you measures things like, e.g. country of exchange or age of the traders, that correlates with the hidden confounders well enough that they allow you to control for the confounding effects in your model.

It is not a thing in professional statistics to expect researchers to already know the sources of confounding before postulating that there could be sources of confounding that need to be controlled prior to believing the results are generalizeable.

Kind of not all exchanges are equal, professional investors still don't trust the main Chinese exchange.

As per the top comment at the moment of writing, most of these day traders were likely to be clueless people roped in by scammy "day trading courses".

Yes. Brazil is a relatively high-friction market, and you'd need to know the actual trading costs. The studies I've seen in the US are that most 'active' traders lose money, but 10-30% do well. There is an entire industry of day-trading hedge funds, so many people believe this.

Yeah what you are doing is nowadays known as the "Trump approach".

Asking what the basis is for disbelieving a scientific study with a statistically significant sample size cannot reasonably be described as analogous to anything Donald Trump does.

Obligatory joke about the three scientists visiting Scotland in the first time:

Biologist: "Whoa, you see that in the distance? That's amazing! All cows in Scotland are brown!"

Physicist: "Hold on -- all we know at this point is that there's a population of cows in Scotland that's brown."

Mathematician: "No, all we know is, there exists at least one cow in Scotland such that this side is brown."


With that said, I agree that it's an excessively narrow sample for the kind of conclusion they're trying to draw.

wouldn't the Physicist first need to assume a perfectly spherical frictionless cow?

Isn't it the case for most studies that we read about online? Anything that ranges from "X can cure Cancer" to "Sleeping X hours is better for you" has a sample mostly consisting of American individuals. That doesn't stop them from generalizing nor does it make us question said generalization.

That being said, this seems like a double-standard.

It's very informative for Brazilian traders, however ;-)

Because it would represent an extremely narrow slice of traders. It's possible traders of these equities find it quite hard to make a living but a larger number of other day traders do just fine.

Also it says "impossible" but the article actually shows it's difficult and rare.

Glancing at the abstract it looks like 97% lose money, what larger number in a more suitable sample size would make a difference?

I've heard low single digits of traders and investors deliver Alpha. Impossible? no. Highly efficent? Yes

The paper's claim is "virtually impossible".

The post's title is click bait.

lots of things are virtually impossible to do for a living and people still manage it though.

Swing trading with baskets as opposed to daytrading seems to be a better way to go according to historical performance, for example https://www.elastic.co/blog/generating-and-visualizing-alpha...

Anything that worked consistently yesterday will no longer work tomorrow as the HFT's arbitrage that market inefficiency away. That's one of the reasons why even extensive back testing isn't a reliable indicator that a trading strategy will be successful.

HFTs don't arbitrage the single person day trader because those guys are are like a mouse farting in a hurricane. There is no point trying to identify and scalp someone who trades at most some $ millions a day, unless the person is doing it on really swampy stocks with no liquidity. Even then, the capacity in those markets for HFTs is also extremely limited.

HFTs will target big insitutional traders who have to shift a billion dollars at rebalancing time in a single day.

Even that is a dying business today because they themselves have been arbitraged away to a degree.

That's not so simple because swing trading is not looking for Arbitrage. 90% of traders lose but a good portion of us make money. It is about building a system which works in bull and bear markets.

My firm uses a similar system that hedges long and short baskets and can confirm that hedging is really the only way to swing trade profitably in the long term.

I do well with options trading for a fairly long time now. Anything over 25% a year is good for me but I aim for 100% per year. I am a hit over that this year at this point. Firms have a huge amount of money so they have to use other strategies. For someone with less than $1M to trade with going long or short depending on the market conditions is adequate. Also money management is a large factor.

I assume you trading US markets? The US markets have been very volatile the last 1.5 years which makes for more optimal swing trading. If the market was either going down or up in a linear line, most hedge funds underperform in that scenario.

Not for me. Volatile is bad as I made more money the years before. End of last year sucked and parts of this year had unusual losses. I prefer trending markets.

I only trade US equities and in particular large caps or soon to be large caps.

That document shows that as algorithms might decay, you can swap out the datasets to out-perform. They also conducted a year long study on the results.

If you take Jan 1, 2013 to Dec 31st, 2015, the Brazillian stock market did very, very poorly. That probably explains why it was nearly impossible since most traders are probably long, not short the market. So even taking 1 day positions in stocks, still on average over 3 years will experience signifigant downward price pressure.

That said, the fact that the market is bad, is just randomness and shows one of the major difficulties with relying on investment income to support your life. Sometimes stuff just sucks, for very long periods of time. Meanwhile you must pay your bills.

That doesn't even get in to whether or not these people dramatically underperformed the market. They very likely did, but I also bet some of them outperformed based on luck or skill.

Stock market investing is something everyone should try at somepoint in small doses just to understand how difficult it is. Its probably the hardest way to make money in the world and even the best investors usually have a bad few years which ends their careers.

> Its probably the hardest way to make money in the world

Just no. Certainly it's not easy. But the hardest? I'd say there is a long list of occupations which most people would consider harder on various dimensions.

I think what parent commenter means is that it would be unlikely for almost anyone to make money that way, no matter how smart they are or how much time they put in.

I think they aren’t saying that it’s a hard profession by itself in the way that you are talking about — only that the expected returns are negative for almost anyone.

So in that case it’s not calling day traders smarter or harder working than others. It’s saying that no matter what your level of intelligence or other personal attributes there is likely some other job that would have much higher expected returns.

For person A it might be to be a firefighter, for person B, to be a gymnast, for person C, to be a brain surgeon, for person D to work as a maid. Etc etc.

At least that’s the way that I read it.

Really well put.

I just left a 4am part time delivery job. The constant interruption and sitting then walking then sitting was hurting me more than daily half marathon or power tools injuries + full time car repair work. And it's slightly below minimum wage. So your health is torn, and you're not even able to plan long term because you earn so few. I can quit (it was a planned interim thing) but my colleagues can not. And they have two jobs.

Since that, I consider any job with a desk a god damn luxury. Bonus point: you have your own toilets.

Right, but the point was, day trading isn't a "job" -- a day on the tough job you described would never end with you needing to write a check to your employer.

Isn't it just a matter of discipline to avoid too large losses ?

Even if you can avoid too large losses, you can't avoid the possibility of lots of little losses.

Agreed. If the sample was California-based day traders specializing in SAAS software stocks between 2016 and 2018, would the sub-headline be that 97% of traders made great profits?

agreed. i noticed that as well. however, it says started between 13-15 so theoretically some of those on the tail end could have been in a up market.

not only that but futures can be shorted too so i don't think its unfair to use "are they profitable" as the criteria in a down or sideways market.

i do wish they showed a breakdown of under vs overperformance relative to the bovespa during their trading period.

Putting aside the potential quality issues of the study, the headline here in HN isn't accurate and made a lot of people comment about how wrong it is.

The study doesn't literally say "impossible" but "virtually impossible" and they do say that a very small percentage was able to make about $310 a day which is about ~110k a year.

0.0051% of the study group, to be precise.

The more interesting trend, which isn't covered in the abstract, is that profitability is inversely correlated with how long you stay at it. It was 30% of people who day traded for one day and gave up on it, and steadily dropped down to 3% for people who kept at it for more than 300 days.

As a stats person, that struck me as a very evocative distribution. The paper put it nicely: "This peculiar pattern is similar to what we would fi􏰄nd, for instance, in the casino roulette."

The next thing I would want to know is, if you model the stock market as a machine that simply gives random payouts drawn from some distribution, what is the probability that a set of 1,551 groups of 300 random draws from that distribution would contain at least one set that averages more than $310? If the paper ran such a model, I didn't see it.

> if you model the stock market as a machine that simply gives random payouts drawn from some distribution, what is the probability that a set of 1,551 groups of 300 random draws from that distribution would contain at least one set that averages more than $310? If the paper ran such a model, I didn't see it.

The abstract says the trader who made $310/day had a standard deviation of $2560 (presumably per day). Over 300 samples, that's a standard deviation of 2560/sqrt(300) or $150/day. So if he really had a breakeven strategy and got lucky, he was 2 standard deviations from the mean, or in the top 2.5%, and you'd expect 40 other traders to have done just as well, and it would be vanishingly unlikely that nobody else did.

Not quite what you were asking, but I don't think there's enough information at least in the abstract to answer your question. The payout distribution will be largely dependent on the size of the traders' positions and riskiness of their strategies.

Yes, it certainly looks like what would be dictated by statistics with random chance that eventually reverts to a slightly negative mean.

In the short term it's not that unusual to flip a coin and get 5 heads in a row, but the more you throw the closer to a 50-50 split you'll see.

I'd also be very interested to see what would happened, hypothetically, if everyone just stopped trading after some fixed number of hours, held that position until the point when they decided to give up trading, and then liquidated it.

My hypothesis is that their average returns would end up being much higher than the returns they realized by day trading that whole time, due to less money being lost to fees.

ICO bagholders disagree

> which is about ~110k a year

$310 * 5 days a week * 52 weeks a year = $80,600

110k pre tax in the USA is 80k post tax per year-- it is often helpful for us to think in terms of both numbers when comparing money making methods. That's my guess at least.

Probably more that they didn't account for weekends, $310 x 365 is $113k

They are also trading on a small account most likely. If you have $500k to trade with making $300 a day is much easier.

Yeah you’re right, I just multiplied by 7 days. Then again you don’t work 5 days every single week (depends of the location) but 80k is definitely closer to the actual figure.

If for each of those people, there are dozens of long tail traders who lost their shirt and tapped out after a few thousand, the point stands. It could be just random chance.

I mean you may as well say: Study shows it is impossible for an individual to build a tech start up as a living.

Just because something is hard doesn't mean it's impossible. Day trading, like many other risky ventures, is something that is very hard and unlikely but for the few who succeed ends up being very profitable.

The abstract of the paper seems to suggest otherwise: We observe all individuals who began to day trade between 2013 and 2015 in the Brazilian equity futures market, the third in terms of volume in the world, and persisted for at least 300 days: 97% of them lost money, only 0.4% earned more than a bank teller (US$54 per day), and the top individual earned only US$310 per day with great risk (a standard deviation of US$2,560).

My interpretation of this, is that those they tracked in this study, it's not even all that profitable for the few that beat the odds while being exposed to significant risk.

But what should we compare those odds, how are we supposed to interpret this?

Imagine we tracked a group of 1000 randomly picked individuals who wanted to be a brain surgeon. I don't know for sure but I wouldn't be surprised if of 1000 randomly chosen people, only 3 people managed to succeed at it after 4 years.

But no one would ever say "Study shows it's impossible for an individual to become a brain surgeon." or imply that it's just random luck to be a brain surgeon. Instead we conclude that being a brain surgeon must be very hard and require a great deal of dedication and commitment.

The reason day trading is singled out whereas brain surgeons are not is that the barrier to entry to become a day trader is very low, anyone can try to become a day trader and many people think they can succeed at it.

The proper conclusion to take from this is isn't that the stock market is pure luck or just a form of gambling... it's that one shouldn't be deceived by the low barrier to entry into thinking that they can succeed at it. It's really hard, soul crushing work to be a successful day trader in much the same way it's really hard work to do anything in life that pays well.

One major flaw in this reasoning..

Some group of people are guaranteed to become brain surgeons.. the demand is there, the training is there, so it is a near certainty. In fact in the US at least, you can know very closely how many brain surgeons there can be per year. Who will fill those slots isn’t too kind to all takers. It is dedication, hard work, and luck (ie ovarian lottery).

Also the majority of those that “fail” at obtaining a neurosurgery training. If they were on the path to medical training will end up falling back to any number of high paying jobs in medicine or surgery.

This study found no one did terribly well at day trading. Random chance implies there will be some outlying outcomes, but that is a fundamentally different objective than having too many people for few slots.

What would you say the difference is between trying to make money on short term investments and trying to make money betting on sports?

Also your final paragraph seems to imply a causal relationship between how hard you work and remuneration. Hard work doesn't hurt, but luck is very important.

There is definitely an edge to be found, but the game heavily favors players with structural (I am able to know things you don't) or institutional (I have hundreds of subject matter experts and deep pockets) advantages, and individual day traders have none of those things.

Which isn't to say day traders can't get those things, it's just that you stop calling yourself a day trader and start calling yourself an investor when you start to intentionally build up structural/institutional advantages.

Isn't it possible that there is simple not enough information in the system to successfully day trade (at least in this particular market). Considering that the daily movements of stocks are like a random walk, I don't find it hard to believe.

It's often difficult for me to discuss this since I do automated trading. I've been doing this for over 10 years now, have grown a company out of it that has 30 employees, and I maybe lose money 2 or 3 days out of the entire year and often times due to technical reasons...

Either daily movements of stocks are not at all like a random walk, or I am just one heck of an astronomically lucky guy. Now I do consider myself fortunate and privileged, for sure... but the amount of luck it would take to repeatedly make money on the market day after day that can fund a decently sized team of high paying engineers, data center costs/hardware costs, so on so forth... it's unthinkable.

The stock market, at least to the degree that I am involved in it, very short term market microstructure, doesn't behave anything like what economists say it does.

I know people who made millions day trading commodities, for no one in that cohort to have made any real money over two years is surprising to the point of disbelief. This may be something particular to the Brazilian exchange.

I’ve noticed a lot of successful traders trade commodities versus equities.

What makes commodities easier to trade?

Ed Seykota has been doing it for decades:


Commodities are generally harder to trade compared to equities. The reason successful traders do it is because you get access to much greater leverage compared to equities.


You’ll quickly get big or go home.

No inside information on commodities.

(However trading on embargoed government press releases is illegal. Though rather than kidnapping Clarence Beeks on the train, you can just interview the Florida farmers yourself.)

There is definitely inside information on commodities, it's just rare that insider trading on commodities is prosecuted.

You would definitely have an advantage in trading oil/gas futures if you had access to ExxonMobil's seismic exploration reports, for example.

Indeed! I found Matt Levine's post last Fall [1], particularly useful, as he believed you couldn't be prosecuted for insider trading of commodities (only securities). Then he received a nastygram from the CFTC :).

[1] https://www.bloomberg.com/amp/opinion/articles/2018-10-24/it...

Wow. But if Matt Levine can get it wrong too I don’t feel so bad.

But the insider trading you suggest above, is predicated on having material non public information of ExxonMobil’s.

Many who make millions are risky and lose it all, or more. How many of them are net positive over more than a year?

While I don't necessarily agree with Kranar's conclusions, I think his point is that the same can be said for many highly risky endeavors, like building a startup.

IMO the main difference is that, while in both cases you'll see a small number of successes and a huge number of failures, in startup land the net expected value is positive while in day trader land it is negative.

Playing the lottery is also very profitable for those who succeed.

Correct, and the statement "it is impossible to become rich by playing the lottery" is absurd. It's part of an annoying tendency to conflate "extremely low probability of success" with "impossible".

Is that the original goalpost? Because I'd say it is impossible to make a living playing the lottery. To make a living doing something, it needs to provide enough positive cash flow.

It's the goalpost of the post I was responding to. I may have misunderstood the point, though.

Ignore the link text here. They’re saying it’s “virtually impossible”, which would be true of the lottery as well.

I'm on the fence about whether adding the weasel word "virtually" is any better.

I think the term "for a living" is more important.

How would you interpret the phrase "It's impossible for an individual to play the lottery for a living"?

Technically we both know it IS possible, technically. But in the real world? I don't know, but I wouldn't be suggesting anyone out there try make a living playing the lotto.

It’s not a weasel word. A weasel word is used to mask a lack of sources or data, that’s not the case here.

Yes, the title should have been something like "impossible for an individual to day trade stocks for a consistently reliable living".

If you loose the lottery you end up with a piece of reciept paper, but if you buy a stock you have it for as long as you want.

The market always goes up. Even if right after this comment, if the sp500 tumbles and we enter a recession, your share of vix doesn’t go anywhere. Youd just wait for the market to recover and hit the price you bought it for.

However, day trading isn’t really viable unless you have a lot of money to take advantage of every dip and hill and get around day trading restrictions.

The problem with day trading, and stock trading in general, is that you can't predict the future. No one can. You can diversify your risk by buying the market (e.g. S&P 500 index fund) & hope the trend of going up N% per year continues, but even that isn't guaranteed. Essentially if you try to time/guess/predict the market, you're going gambling & the fees you pay on trading on the rake taken by the broker. Some gamblers do well. Most don't.

This is not correct.

If we consider a tech startup a single gamble, then living off trading is a continuous stream of independent gambles.

While individual gambles may turn out better than the average (even many in a row), repeated gambling will always see the total outcome return to the normal.

This is why the house always wins.

> Day trading, like many other risky ventures, is something that is very hard and unlikely but for the few who succeed ends up being very profitable.

> the top individual earned only US$310 per day with great risk (a standard deviation of US$2,560).

The very best day trader made around the same as an entry-level SWE, which does not scream immense profit to me.

I agree with your point totally. But just to put in perspective that U$310/day in Brazil would be the equivalent of the top 10% software engineer salaries in Brazil.

> an entry-level SWE

In a major tech hotspot, working 8 hours a day

Or you could say: it's impossible to win the world championship at anything.

It would be easy to find data to show that 99.99% of people aren't the best tennis player or fastest runner or best Scrabble player in the world; therefore, nobody can be the best at it. According to this reasoning, at least.

The point of the article is to dissuade people from pursuing careers in day trading based on marketing that it's an easy way to earn money (from people whose business is selling instructional materials... not day trading). If there were lots of people selling courses on "Become a WTA Tour Player, make millions winning tennis tournaments!" the same type of paper might be warranted for tennis playing.

On a high level I agree. The way I see it, "impossible" is the wrong label. What you must do, though, is have a rational basis to believe you're in (making up a number) the top 0.01% of people skilled in that area. If that's true, then it might be a reasonable plan. For the other 99.99%, it's a bad plan.

And when I say rational basis, that's its own challenge. There is a huge potential for emotional thinking disguised as rational and for self-delusion. Everyone wants to believe they are the top 0.01%, and most of them are wrong.

That would be a perfectly accurate statement. If none except world champions got paid, it would be an atrocious career choice

A better way to put it is that day-trading is like gambling. You always lose money in the long run, even if you have spurts of success.

The results are fairly obvious to anyone who has read "A Random Walk Down Wall Street".

Day traders rely on technical analysis which is the equivalent of astrology in the financial sector.

"Random Walk Down Wall Street" is somewhat outdated and a lot of the evidence from behavioural finance doesn't really support its central thesis. It's more dogma than anything.

The central thesis of short term market movements (say day to day or week to week or month to month) being a random walk is about 99.7% correct.

This is a statistical reality, you can literally crunch the numbers! Behavioral finance has nothing to do with it.

about 99.7%?

I've always hated the argument of this book: yes the markets are chaotic and short term movements appear random, but the appearance of being random is not equivalent to being caused by a random process. The book makes a big deal about how technical analysts saw patterns in random data so it must be bunk, but this doesn't really say anything about the predictive power of technical analysis when the data is generated by the aggregate actions of thousands of traders instead of a coin toss.

Price action is not the result of investors literally buying or selling at random. Each action is undertaken in response to external conditions and it is far from absurd to believe that patterns can exist in the price.

"The easiest way to make a small fortune in the stock market is to start with a large one."

Man this one is wrong lol

With big enough equity and proper risk management, one can do so. Problem lies with low equity & high expectation which often leads to improper risk management and a sudden margin call. i.e. it's really easy to make $1 with $1 billion account size with risk nearing zero, but trying to make $1 with $100 is exponentially hard & risky. These so called day traders opt in for later b/c they are trained by the course & society & influences to do so.

Every retail trader I've talked to, wants to start with $1000 or less and make $300-$400 per day. Literally everyone. After they lose their 50th $1k, they still aim for the same. It's like an addiction to gambling.

I always thought the problem with day trading is an absence of information: How does a single individual know enough about about all the various entities they are trading to really know when one is going up or down?

Seems like an institution with dedicated employees performing due diligence will know much more than an individual day trader, and thus will be able to make better decision about when to buy and sell?

Anyway, my bias has always been that day trading is really gambling.

All effective trading is, essentially, insider trading. Not in the legal definition, but in the sense that to beat the market, you have to know something that "the market" doesn't.

Or, just get lucky.

The only personal example I have is from buying NFLX stock about a decade ago. I saw a news (Slashdot I think) story that Netflix's servers were having an hours-long outage. I can't remember if it was mentioned in the story that their stocks took a dip, or I just happened to look. Either way, I noticed that their stocks were down about 30% on that news, which seemed silly to me. The market was treating this like it was some kind of disaster for the company, whereas I was pretty confident it was a temporary blip.

I bought the stock, just a few hundred worth since I was in my first barely-over-minimum wage job at the time. Netflix got their servers running again, and the stock recovered the next day. I decided to hold on to the stock because, why not? Today, it's worth about 20x what I bought at.

If only I'd bought more.

My point is, I knew that the market was wrong, and an hours-long server outage wasn't a reason for Netflix to suddenly be worth 30% less than it was the day before.

To successfully pick stocks, you have to have some kind of expertise or insider knowledge to know when the market's wrong, and you're right. Most of us, in most situations, don't have that kind of expertise or insider knowledge.

It’s not impossible, there are many people who make decent money from day trading. It might be more accurate to say “study shows it is difficult to earn a living trading stocks”

The abstract, of course, contains more detail than the headline regarding exactly what is claimed:

> We show that it is virtually impossible for an individual to day trade for a living, contrary to what course providers claim. We observe all individuals who began to day trade between 2013 and 2015 in the Brazilian equity futures market, the third in terms of volume in the world, and persisted for at least 300 days: 97% of them lost money, only 0.4% earned more than a bank teller (US$54 per day), and the top individual earned only US$310 per day with great risk (a standard deviation of US$2,560). Additionally, we find no evidence of learning by day trading.

The interesting bit to me is, “we find no evidence of learning by day trading”. This implies to me that the top earners are merely lucky, and have no special knowledge or advantages.

Insider info?

The Longer I live, the more I see people profiting off knowledge.

A personal example I see of this is in Tech. If you understand tech, you can find areas people don't understand. Those are what's worth investing or shorting.

Although if you are limited to day trades I'm not sure this is relevant.

It could also imply that they simply have related experience or skills before starting.

It's difficult to, for example, learn how to become a brain surgeon on the job.

Medicine actually involves a significant amount of learning on the job in all specialties - https://www.verywellhealth.com/types-of-doctors-residents-in...

We were taught explicitly in my (highly-ranked US) business school that stock performance is a random walk and success is luck.

I assume this means you were taught a very strong form of the efficient market hypothesis.

I believe there’s some good evidence for weak form efficient markets hypothesis in the most liquid markets, which implies you can’t make money by trading only on ‘technical signals’ in the price. This is what you might expect day traders to try to do, so in that sense it’s not a particularly surprising result this paper demonstrates.

But the consensus is, I think, that the strong form of the efficient market hypothesis might not be true in real life - being good at fundamental analysis can lead you to outperform the market.

Illiquid markets are a bit different again - you can potentially get ‘paid’ a lot more just for providing liquidity to the market there, which is different again.

This is an interesting concept to be teaching business school students.

Doesn't this amount to telling you your education is useless?

No, because running a business and stock trading are different things.

running a business is just the opposite side of the stock trading coin.

If the stocks themselves are random walks, then the underlying company performance must be random walks. If your education does indeed improve the outcomes of the business you manage and run then that would mean stock performance can be influenced by things other than luck.

Unless of course you are operating under the premise that company performance and stock performance are unrelated.

I don't know. But a business may perform predictably from the insider's point of view ans unpredictably from the outside.

And the article even moreso:

> "We follow all individuals who day traded mini-Ibovespa futures contracts for their first time from 2013 to 2015, a total of 19,646 individuals."

I know nothing about the Brazilian equities market, but it seems like it might be harder to for individuals to reliably make a profit on something heavily traded and fungible (Ibovespa is an index), where you're competing with much bigger players. Smaller stuff might work better--for one person--because it doesn't scale.

I don't know much about the Brazilian equities market per se, but I do know about trading in other markets.

Liquidity is a very important aspect of any market. If you are the largest player in a small market (i.e. it has low liquidity), you are market making, and the problem with that is information asymmetry comes into play: you're trading against people who know stuff you don't know, and they will only trade against you when you're wrong.

If you are in a highly liquid market, the market represents all the information out there, so you need to think about where your informational edge comes from.

Most day traders seem to focus on technical analysis in the belief that there are underlying probabilities that drive market movement and this alone is enough to find an edge.

The linked paper is not one I have read completely yet, but it seems to align with my own experience and those observed elsewhere: that might be true, but you need to find the edge nobody else has seen yet, and that seems to be very difficult and carry some risk.

Exactly. The odds that any random person (or worse, those with <2 years of day trading experience), has a meaningful edge on an index covering 80% of the Brazilian stock market are....not great.

The trick with a lot of smaller markets is that they're easily manipulated. You try to trend-follow and just get exploited by pump-and-dumps.

Probably very similar to results for people who try to make a living gambling in casinos.

Looking at how bad these results are, the career gamblers in the casino seem to have an edge. They have various strategies that actually do work to greater or lesser degree - e.g. Playing poker which is a game of skill against other people with a house rake instead of a house mathematical edge.

Poker isn't gambling, but a game of skill. Are there many people making a living actually gambling on something like craps or roulette.

Making a living on poker? Absolutely.

Making a living on craps or roulette? Not mathematically impossible, but highly improbable. How do they get their edge?

Roulette you just play the statistics: not really an edge. You run the Martingale system (double your bet after every loss), and my quick monte carlo simulation says I'd turn $1 into $2,250 after 5,000 iterations. But I'd need a bankroll of $8,200 to cover the 13 consecutive losses my model spit out. And if your bankroll is that big, there are easier ways to get the same return.

(Oh, and most tables have maximums, so you can't keep doubling your bet to infinity.)

Not likely. But there are some strategies that amount to exploiting casino promotions and perks. There used to be some tricks around slot machines - like those near the door had a payout ratio above 1.0. There are probably still tricks along those lines, but if they become well known the casino's change them. You'd be much better of sticking to games of skill.

Fair enough. I've heard of craps players teaming up to play huge on both sides of the pass/don't pass, which would still have a negative EV. But they racked up comps pretty quickly which supposedly gave them a slight overall return when factoring that in. This might be good for a while to accumulate free rooms and whatnot, but not sure how one would actually make a living doing this.

Yeah, seems like a bad bet, but still better than day trading judging by the study results.

Not craps or roulette (without cheating), but certainly with blackjack. By finding favorable conditions and counting cards, it's possible to extract a slight advantage over the house (say 1% edge). Of course this is a grueling enterprise with an extremely high variance and thus requiring a large amount of capital.

Not impossible but highly unlikely probably means impossible, practically speaking.

Unfortunately this falls into the "too good to be true" category. A very unscientific categorization but surprisingly accurate in real life.

To me the worst conclusion is that "Additionally, we find no evidence of learning by day trading." Which leads me to believe that despite your best efforts at obtaining knowledge, luck in timing still plays too much of a role in the outcome. That doesn't add up to a viable alternative to a regular day job.

"Making decent money" doesn't mean crap it's about risk, expected value, and opportunity cost.

If you've got a million dollars and put it into an index fund you'll make ~$100k/year. If you day trade and make "Decent money" of say $70k/year you're still at a loss vs if you took the less 'risky' approach. Even if you made $150k (50% over the index) it's still a loss due to lost opportunity cost. As the $100k index + a job at McDonalds making $30k/year still puts you ahead due to long term vs short term capital taxes. And if you've got the knowledge to day trade and earn more than 50% over an index you can probably command a better job than fast food.

If you don't have a million in capital it becomes harder and harder the less money you have.

Your friends would need to consistently beat the market by a percentage of their capital for it to be considered a 'win' imo.

While I agree that day trading anything is an extremely poor idea- why is trading by banks/hedge funds/Goldman Sachs etc. consistently profitable? (Or if isn't profitable, why do they do it?) Is it just the massive informational advantage that they have? I'm not advocating for trading per se (I certainly don't do it!)- just curious about how large financial institutions presumably turn trading profits in light of EMH

Goldman Sachs isn't profitable because of prop trading. They charge fees for services. It's that simple.

So then why do they trade at all then? I'm not arguing for trading, I'm just curious.

It seems like the existence of 'trader' as a job at all of these financial institutions kind of disproves EMH if you have enough market data

They don't.

After the Volcker rule the big banks had to close their prop trading groups: https://en.wikipedia.org/wiki/Proprietary_trading

The idea that investment banks are filled with human "traders" is Hollywood nonsense.

Most trading activity (and even filling of retail order flow) is now done by quant-driven funds.

According to LinkedIn there are 573 people with the job title 'trader' at Goldman Sachs. I'm scanning through the list now- titles like Fixed Income Trader, Mortgage Trader, Rates Trader, Oil Trader, Equity Sales Trader, FX Trader, Emerging Markets Trader, Base Metals Trader, Interest Rates Trader, Investment Grade Bond Trader, Equity Derivatives Trader, Commodity Trader.... I found all of these on the first page of results (25 people)

By contrast, about 37,000 people work at Goldman. Those people with the name “trader” in their title are simply executing client trades as a service for fees—-not speculating with the banks own money.

I stand by my point, the idea that Goldman is full of traders in the Hollywood sense is nonsense.

There are not a lot of equity cash traders left in Wall Street firms; but it's not like equity cash is the only thing to trade.

Trump has undone the Volcker rule portion of Dodd-Frank [1], for what it's worth

Goldman Sachs, et al, would not lobby the government to get rid of a law that prevents them from doing something unprofitable.

[1]: https://www.politico.com/story/2019/08/20/volcker-rule-josep...

A big part of it is that on average the stock market goes up, so even if you are randomly buying stocks you'll come out ahead over the long run as long as fees don't eat it up. Another significant portion is that they have access to a lot of information you don't, and are able to buy stocks and derivatives that you cannot. This comes as a benefit of their size.

The type of trading that is consistently profitable at banks is mostly market-making, which is not a strategy available to a day-trader.

>why is trading by banks/hedge funds/Goldman Sachs etc. consistently profitable?

Not a level playing field. It's exactly because you're going up against guys like GS that's the reason for day trading being suicide.

In addition to what the others said, stock trading fees often don't scale with the cost of these stocks. If you're investing with low amounts, these static fees eat up a lot of any potential gains of day trading

I have a couple friends that do day trading on "technical analysis". As in, they only trade on the chart movements. Both claim to make money consistently. I generally believe them, but it seems like black magic to me. Especially when they explain what they're doing in detail.

Not sure I understand this study tho, and whether or not it relates.

One thing to keep in mind is that even a random number generator can trivially turn a profit when the market is trending up. It's when the market is volatile or going down that you get a real test of how well a strategy works.

Which isn't to say that technical analysis is unfounded. And I do believe that there are opportunities that exist for individuals that haven't been arbitrated out of the market by the professional quants simply because they don't represent enough money to be interesting on an institutional scale. But it takes something like a bear market to really know for sure if you've found those opportunities, or if you're just riding a rising tide that lifts all boats.

>I have a couple friends that do day trading on "technical analysis".

As someone that trades on fundamentals I like to think of that as reading tea leaves. Literally non sensical. Oh yeah sure that pattern is the upside down elephant so we should go short...

No smoke without fire though so I assume some are making money with it. Just doesn't seem like a sound idea to me.

I also have a few friends who do well day trading (which is usually done by mostly or only technical analysis). They have done well for themselves as they actually treat it like a job. Constantly trying to educate themselves, being at their computers paying attention to the market for the entire trading day.

Most folks who feel the allure of day trading don't realize that to make it work it has to be your full time job. You can't just make some trades, go off and do something else, and expect to make any money.

What i have seen this algos work mainly by having a high chance of winning a small payout and a very small chance of losing everything. So everything works fine for 1/2-1 year and ppl spread word of mouth how nice this steady income is. Meanwhile the owner of the algo/platform makes a nice commission. Then everything goes bust and starts anew somewhere else.

Technical analysis works the same way the weekly paper works to say that Virgo or Capricorn will have a good week at work.

Something that's probably not mentioned in the article is that it's really easy to get into day trading in Brazil. I think I read somewhere that in the US you need a US$ 25k account to day trade; in Brazil with a US$ 50 account you're in the game. There are many zero-fee brokers (though those are a recent development, I don't remember those in 2013), and capital gain taxes are 20%.

In fact I'm surprised that day trading in Brazil is not more popular than it already is. It's basically legalized gambling. I do remote work as a software developer in Brazil and I've been playing a bit with options swing trading on the side (which is even cheaper than day trading); I don't plan to make a living out of it, but it can be fun (and a bit addictive).

Applications are open for YC Winter 2020

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact