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

What I Learned from Losing $200M (2015)

nautil.us

Post reply on HN
Print view View original

101–110 of 136 posts

Re: What I Learned from Losing $200M (2015)

#101

Post by johnvonneumann »

ryandrake wrote: post #13
anonu wrote: post #10

As a junior trader navigating the markets during that time I noticed that nobody has a clue about anything. Pundits, researchers, analysts, Junior guys, Senior guys... They all pretend to know. I'm not suggesting the markets are truly random. Systematic profits are feasible year over year... But only to the select few who are in the right product or looking at the market in the right way. I suppose it's those guys th…

> Systematic profits are feasible year over year... But only to the select few who are in the right product or looking at the market in the right way. You seem to be falling for the same myth: That "certain wizards" can get +EV. Every casino on Earth makes money from this myth. I believe that there are no wizards in the stock market (or in business in general), and everyone's gains and losses vs the total market are…

So he is falling for a myth, yet what you're referring to is the Efficient Market Hypothesis, does the word "hypothesis" ring any bells for you? Even more so, does the fact that a theory that originated in the 1960's, still bares the name EMH. Almost 60 years on and it still hasn't been proven.

Your thought process seems incredibly flawed, the idea you have about coin flipping being at all similar to stock picking is laughable. And this to me would indicate that whilst you may believe that your analogy is a good indication of survivorship bias, it is clear to me that you are suffering from confirmation bias. You're looking for examples that suit your belief, and ignoring some basic elements. Let's look at this one seriously, coin-flipping is an statistically independent event, whilst if we apply even some fundamentals of finance and EMH (which I don't agree with, btw), we will quickly find ourselves in the realm of correlation. I never stuck around Uni long enough to get deep into this stuff, but you could do with some reading on the CAPM model. To quote from wikipedia on the CAPM model:

"Financial correlations play a key role in modern finance."

You only have to look at the fact that many companies have been put out of business by companies like Amazon, which means that at the very minimum, you have at least ONE external factor that has an influence on the performance and price of other securities.

And this is all without even starting on the even more hilarious notion that information across the market is symmetrical, that's right, I as a lowly consumer exchange peasant, have the exact same information available to me as Jim Simons and the team at RenTec. Who's Medallion Fund, by the way, returned a 35% CAR over a 20 year period. You have to be mental to believe the shit they teach in undergraduate Finance.

Re: What I Learned from Losing $200M (2015)

#102

Post by Bromskloss »

purplepotato wrote: post #90

I was working in the crude oil / nat gas options pit at the NYMEX during the summer of 2010 when these trades went down (where much of Mexico's hedge was traded but not necessarily the author's portion.)

Please tell us stories, if you have some.

Re: What I Learned from Losing $200M (2015)

#103

Post by yters »

the_cat_kittles wrote: post #96
yters wrote: post #85

Earlier quoted context omitted.

Two sequences of coin flips: HHHHHHHHHHHHHHHHHHHHHHHHHHHHHH THTHTHHTHTHHHHHHHTTTHTHHHTHHTT Which was generated by fair coin, which by 2 headed? How did you decide?

assuming you want a bayesian approach? take a flat prior and look at the MAP? though you know its just going to be the observed rate. if you know you only have a fair coin or a double head coin just compare their likelihoods? what is the point you are trying to get at?

Solomonoff induction and universal probability.

Re: What I Learned from Losing $200M (2015)

#104

Post by prewett »

inthewoods wrote: post #88
Danihan wrote: post #36

Earlier quoted context omitted.

I'm sure you're correct, all stock picking is dangerous though. I'd assert that my style is the least dangerous. > look around your little bubble I didn't (and don't) have a bubble. In fact, I frankly despise the Bay Area and SV because it constantly tries to put you in a liberal / wealthy bubble. >If you can see signs that Chipotle is doing well, so can everyone else. Apparently not, though. They doubled on IPO but…

>I'm sure you're correct, all stock picking is dangerous though. I'd assert that my style is the least dangerous. Why? Seems to me that all stock picking has the exact same level of risk.

If the stock market were truly random, then yes, all stock picking would be equally risky. But the stock market is a voting machine voting on the future of companies, and is definitely not random. So you could look at two stock picking strategies: 1) throw darts at the Wall Street Journal and buy whatever it hits, 2) buy new companies that are being used by lots of people in your area (the parent's strategy). #1 should give random results centered around the performance of the entire market. #2 will probably give better results, because it incorporates some information relevant to the voting, namely that customers like the product. Which company would you rather have: some random company or a company that people are willing to stand in a visible line for? If there are lines, either people need the product or they love the product, and either way that tends to correlate with profits.

Re: What I Learned from Losing $200M (2015)

#105

Post by prewett »

martijndwars wrote: post #65
Danihan wrote: post #26

Earlier quoted context omitted.

Because I don't look at the internals, I look at the outcomes. Eg - Everyone using google, everyone eating at Chipotle, everyone being addicted to cell phones, etc.

Picking stocks based on "everyone uses..." is a great example of a bubble. What about all the companies that are not consumer facing?

A bubble is "everyone buying because everyone knows someone will buy even higher." "Everyone uses..." is a great example of a revenue generator. I assume you are meaning that you buying based on popularity of places you frequent is an example of an echo chamber?

Regarding non-consumer facing companies, I don't see what the problem is. If you have no information about the company, then it's gambling to buy it. So you miss out on some companies, no big deal. Most people only have enough money to make a few bets, so you should bet on what you know about. It's not like a test where you lose points if you don't answer a question; you don't lose money if you don't buy a stock you don't know about.

Re: What I Learned from Losing $200M (2015)

#106

Post by interestingcpp »

sokoloff wrote: post #100
kovek wrote: post #67

Earlier quoted context omitted.

> That should make you want to invest in the #1 company that manufactures satellite components. I understand the idea, and I might agree. I just wanted to point out that in the book "The Intelligent Investor", there was the idea that you could think about investing into the second best player in a certain space. I think the reasoning was that there is more opportunity for growth for a second-grade company than a firs…

I'm not sure I'd want to invest in the second-place search engine or second-place social network.

Well, they’re just ad providers, and plenty of people invest in nth tier ad providers.

Re: What I Learned from Losing $200M (2015)

#107

Post by TuringNYC »

yellowstuff wrote: post #55
lordnacho wrote: post #49

As a derivs trader myself, I find much wrong with this account. First of all, he came away net positive, so the title is a bit of a humblebrag. Aside from that though, he should have known beforehand how models work. David Hume mentioned it hundreds of years ago; you only have the past, and the past might not contain all the dynamics of the future. Lest you think this only happens to social science models, there have…

Agreed. This part seemed like an exaggeration to me: > Hitting a market’s ceiling like this was something that none of my methodologies accounted for. I worked at a bank in 2008. The biggest blowup in recent memory then was LTCM in 1997 (seems quaint now), which blew up partly because they had positions too large for the markets they were in. It is simply not credible that someone would put on a huge position in 2008…

>> It is simply not credible that someone would put on a huge position in 2008 and not give any thought to the impact their own trading would have on prices.

And yet it keeps happenning, see "London Whale": https://en.wikipedia.org/wiki/2012_JPMorgan_Chase_trading_lo...

Re: What I Learned from Losing $200M (2015)

#108

Post by the_cat_kittles »

yters wrote: post #103
the_cat_kittles wrote: post #96

Earlier quoted context omitted.

assuming you want a bayesian approach? take a flat prior and look at the MAP? though you know its just going to be the observed rate. if you know you only have a fair coin or a double head coin just compare their likelihoods? what is the point you are trying to get at?

Solomonoff induction and universal probability.

how about something specific. how would you answer the question you proposed used bayesian methods?

Re: What I Learned from Losing $200M (2015)

#109

Post by yters »

yters wrote: post #103
the_cat_kittles wrote: post #96

Earlier quoted context omitted.

assuming you want a bayesian approach? take a flat prior and look at the MAP? though you know its just going to be the observed rate. if you know you only have a fair coin or a double head coin just compare their likelihoods? what is the point you are trying to get at?

Solomonoff induction and universal probability.

P(all heads|double headed coin) = 1

P(all heads|fair coin) = 2^{-L}

Applying Bayes gets P(double headed coin|all heads) >> P(fair coin|all heads) for a long enough sequence.

Re: What I Learned from Losing $200M (2015)

#110

Post by softwarefounder »

rb808 wrote: post #7
CalChris wrote: post #6

I crewed on a sailboat in YRA races leading up to and during the 2008-9 crash. It was mostly people from Lehman and Barclays. I was the only SV guy on the boat. No one partied harder than those guys but then they really only partied with themselves. It was kinda like Boiler Room. The Dot Com boom is the stuff of legends but these guys left nothing on the table. Barclays had a riff and ordered a string of cabs to take…

> It was mostly people from Lehman and Barclays. I was the only SV guy on the boat And now its mostly SV guys on the boat and the occasional trader? :) "Highly paid, quite sure of themselves but wrong... It was just other people’s money so who cared."

Haha, brilliant point. Fortunately, it's probably a lot of VC money instead of individuals, or publicly traded companies; at least one would think.
Post reply on HN