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

I might be a complete failure (after 8 years of work)

pathdependent.com

Post reply on HN
Print view View original

61–70 of 118 posts

Re: I might be a complete failure (after 8 years of work)

#61

Post by chasingsparks »

krosaen wrote: post #9

I find it inspiring that he's stuck with it, and wonder how's he's made a living in the meantime

I take short-term jobs and contract work for a few months, then quit everything for full-time research. Usually the jobs I take happen to require learning something tangential but possibly useful. It's a happy coincidence that I am perfectly happy with the absurd level of frugality required. (I live in a place that most people call the "House of Squalor. It's really not dirty, but it is semi-dilapidated.)

Re: I might be a complete failure (after 8 years of work)

#62

Post by chasingsparks »

bbers wrote: post #32

To beat the market you HAVE to predict what trades others will make in the future. It is easy to use math to make it appear to be a numbers game but that is just abstraction. The reality is that people buy and sell stocks for 1000s of different reasons.

I use much less math than your average quant. I prefer simulations with a lot of agents. Think Game of Life, not partial differential equations.

Re: I might be a complete failure (after 8 years of work)

#63

Post by chasingsparks »

wangwei wrote: post #41

Even people with very high level math knowledge fails to predict the market (remember the two noble prize scientists who failed?).Since you're out of school when you were 17, I doubt you'll able develop complex schemes (unless you are a very very rare genius).

I was not out of school at 17. What caused you to infer that? (I'd like to correct it if it sounds that way.)

I am actually starting a masters/Ph.D. program in the fall in computational social sciences. Like I mentioned in the post, I am interested in social systems in general, not only markets. (However, I do find markets to be the most interesting.)

Re: I might be a complete failure (after 8 years of work)

#64

Post by chasingsparks »

arthurdent wrote: post #46

I find myself excited over the preliminary results of my increasingly sophisticated simulations This. The vast majority of the money being made in the market is being made by people with what you may consider a laughable lack of "sophistication". Start simple. Complex modeling of stock market behavior only makes sense if you think the stock market has a complex underlying structure.

By sophisticated, I meant refined and cultivated, not devoid of simplicity. It was a poor choice of words. At some point in my development, I started to eschew more complicated forms of AI and machine learning, in favor of simpler agent-based models with "laughably" simple assumptions. I think it's the proper direction, then again, the title I chose for the post was not arbitrary. ;)

Re: I might be a complete failure (after 8 years of work)

#65

Post by jrockway »

muhfuhkuh wrote: post #42

Look to none other than Long Term Capital Management[1] back in the 90s to see how dangerous a game it is to either algorithmically predict or automate a thing as emotionally wrought and, at times, institutionally manipulated as the market. [1] http://en.wikipedia.org/wiki/Long-Term_Capital_Management

What? Their time-tested bond arbitrage algorithms worked great. They beat the market month after month.

Only when they tried speculating on equities did they fail. And even then, their positions only took short-term losses that they would have recovered from had they not been leveraged 100-to-1. Their algorithms were correct, but they didn't have enough capital to see their positions through to the end.

The LTCM story is about greed and leverage, not algorithmic trading. Plenty of money can be made trying to beat the market: it pays the salaries of the hundreds of thousands of investment bank employees around the world. The problem is, you personally can't do the same things they do, because you don't have enough capital.

Raise a billion dollars and your days of commuting to work are over.

Re: I might be a complete failure (after 8 years of work)

#66

Post by chasingsparks »

Loy wrote: post #37

The term "complete failure" is more revealing about your draconian idea of success. You should take a look at your own words : "they have never found their Holy Grail" [therefore, they joined the ranks of Academics as a testimony of failure]. If you're running after the Holy Grail, you have no choice but to be a hero. What kind of pressure is that ? In fact, most of the answers are in front of you, in the way you exp…

The title was unnecessarily provocative, a reflection of my mood at the time of writing it. (It's Sunday; I'm hung-over; and I was looking at the results of a four day simulation run that were ambiguous.)

My priority is in "beating the markets" because it's a fascinating challenge. I also have interests that lie beyond satisfying my intellectual curiosity but require wealth. I find the idea of using my own wealth to solve problems (e.g. avenues of cancer research) far more appealing than fund-raising or requiring public funds.

Re: I might be a complete failure (after 8 years of work)

#67

Post by majma »

nandemo wrote: post #53
edw519 wrote: post #19

After eight years, I have nothing concrete to show for my efforts. So sad, so true. This is what happens when you play a zero-sum game. Even if you win, someone has to lose. Here's an idea. For the next 8 years, why don't you do something that helps others by making the pie bigger for everyone instead of just trying to game a bigger piece of a smaller pie for yourself. When you do for others, you always have somethin…

So sad, so true. This is what happens when you play a zero-sum game. Even if you win, someone has to lose. When you play football too, if you win someone has to lose. And yet there are positive externalities, to wit millions of people watching the World Cup. In trading, the externalities might not be so obvious but they do exist.

Could you explain more please? Bonus points if you do it without using the phrase 'provide liquidity'.

Re: I might be a complete failure (after 8 years of work)

#68

Post by Kaizyn »

motters wrote: post #20

Many smart people have tried to "beat the markets" and failed. From what I've seen of it, this sort of activity is at best a form of numerology. Also, in trying to move money around slightly more efficiently than the next market analyst you're not really generating wealth as such, and your ideas and activities will leave no lasting value for society. You may not be a failure, since the skills acquired in analysing tr…

If I'm not mistaken, I believe the idea is to set up some automated trading bots that just simply make money for you on a continual basis. This constant revenue stream allows you to forget about making money and allowing you to use those resources to either fund a life of ease and luxury or other pursuits such as "making the world a better place".

Re: I might be a complete failure (after 8 years of work)

#69

Post by 10ren »

Question: has anyone managed to do this? (eg. do hedge funds do this, or do they combine algorithms with their own ideas and quasi-insider knowledge, plus massive amounts of timely data - inaccessible to an individual).

My old supervisor (ACM fellow) comforted my failure to create AI by saying that lots of other clever people hadn't done it either. I have a feeling that you are attempting something more difficult that AI - ie, algorithmic inference without the data that your competitors have. Because of their better information, they will find a regularity (eg. 0.01% dip on Wednesdays) quicker than you - and trade it away.

It would be awesome if you could do it though - forget the philosopher's stone aspect of gold-for-nothing, this would be a very cool intellectual achievement.

Idea: focus on small markets/exchanges (unpopular metals, unpopular countries, even non-financial markets, like eBay) that don't attract the attention of big players. Easier to find arbitrage there; you can make some actual money, and then move onto a bigger market with your learnings, earnings, and confidence. Sort of a niche strategy.

Personal Note: My AI research depressed me, and I (wisely) decided to shelve it. It's been incredibly more satisfying to work with others, to achieve doable goals, that benefit others. Fortunately, AI comes up in everything; and my current approach (of not doing AI directly) may well be the most effective way to approach the goal. It might be the same with you - both markets and inference are universal.

EDIT So, Tantalus = low-hanging fruit + moving-goalposts http://en.wikipedia.org/wiki/Tantalus#Story_of_Tantalus

Re: I might be a complete failure (after 8 years of work)

#70

Post by mistermann »

lowkey wrote: post #49
jeromec wrote: post #40

Earlier quoted context omitted.

> This is what happens when you play a zero-sum game. Even if you win, someone has to lose. How is it always a zero-sum game to win? Fundamentally, trading stocks is not that at all. If I buy one of the first issued shares of "Tasty Ice Cream" for $1 because I want the company to spread their goodness as well as get a return on my money, no one has lost. If I lose my job the next week and need to sell my share which…

I think this is the way capital markets were originally supposed to work; matching companies in need of growth capital with investors seeking to earn a risk-adjusted return on their long-term savings. Unfortunately, the financial markets have morphed over time such that now, traders trade for the sake of trading. Instead of our work being our wealth, trading securities somehow became an end unto itself. What's worse…

>Retail investors are taught to 'invest for the long-term', and 'not try to time the market' meanwhile hedge-funds and high-frequency traders play a different game with different rules.

Exactly....that the financial services industry constantly advises the retail investor to "invest for the long term", when they do the exact opposite (while generally using the funds provided by these exact same retail investors), to me is a pretty strong indicator that the markets are something entirely different than how they are sold to people.

The markets were created to provide funding for companies in need of capital, and a branch of the financial services industry was legitimately required to facilitate this process between companies and individual investors, but now almost everywhere I look people are focused on finding new ways to skim money out of the system, under the guise of "providing liquidity", as if there was some problem with liquidity 10, 20, 30 years ago.

Post reply on HN