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

How I made $500k with machine learning and high frequency trading

jspauld.com

Post reply on HN
Print view View original

81–90 of 320 posts

Re: How I made $500k with machine learning and high frequency trading

#81

Post by jspaulding »

yajoe wrote: post #43
tarr11 wrote: post #23

This is just glorified gambling. I am not sure what special insight or advantage he had, other than his own model. Every trader has a model. It could have easily been called "how i lost 500k with machine learning". Like gambling, it's easy to manipulate statistics to show that you did well in some period of time. I worked for a large investment bank about 10 years ago, writing trading programs for quant traders who w…

This. I love crazy projects and Show HN's until the cows come home, but this one is dangerous that I must repeat the warning to others. I cannot emphasize how important it is to understand that people who trade using price action ( http://en.wikipedia.org/wiki/Price_action_trading ) are just speculating based on where they expect the price to move. It's no different than people who play Texas Hold'em online and specu…

Hey, I didn't actually intend this to be a course. I do not make any money in the market right now so am certainly not qualified to teach a course on it. And of course, if I was making money in the market I wouldn't have posted this at all. So please everyone remember that. These comments have made me realize it's probably for the best if I do not post the source code. Basically you are competing against armies of PHDs who are buying buildings next to the exchange so they can get their executions slightly faster. It is indeed surprising to me that I was able to make money in the first place. But I do know for a fact that I did make money and I also know that I was not at risk of losing a bunch of money. As mentioned the most I lost in one day was $2000. That's all I was risking.

Re: How I made $500k with machine learning and high frequency trading

#82

Post by downandout »

tarr11 wrote: post #23

This is just glorified gambling. I am not sure what special insight or advantage he had, other than his own model. Every trader has a model. It could have easily been called "how i lost 500k with machine learning". Like gambling, it's easy to manipulate statistics to show that you did well in some period of time. I worked for a large investment bank about 10 years ago, writing trading programs for quant traders who w…

"Glorified gambling" is a pretty accurate description of this. However, that isn't necessarily a bad thing. If you have a situation where you can gamble with a long-term positive edge, then the proper strategy is to play with as much money and for as long as possible. The determining factor here is whether or not the combination of a particular investor's strategy, algorithm, and ability to execute will give them a long-term edge over others in the market - not whether or not this may be a risky activity in the short-term.

Re: How I made $500k with machine learning and high frequency trading

#83

Post by fr0sty »

beagle3 wrote: post #73
crntaylor wrote: post #63

Earlier quoted context omitted.

The difficulty is in identifying what is a 'flash crash' (i.e. a temporary downward blip in prices caused by computer or human error) and what is a genuine downward price movement. If the market dives and you quickly get into a big long position, and then it dives some more - what do you do? You can either close out your losing trade and take the loss, or hope that the market comes back up, all the while holding on t…

> The exchange could nullify all trades in a certain period of time, which would completely wipe out your upside potential. This is the most important thing: In every single "flash crash", the exchanges have retroactively canceled trades, in a rather arbitrary manner (e.g., "every trade between 16:30 and 16:38 is null and void"). There is some underlying justification, but it is also arbitrary (e.g., "anything below…

Nitpick: When exchanges have busted trades there is a "at or below $XX.XX" condition as well as the "between XX:XX and YY:YY" condition.

In the Flash Crash as well as the Knight Capital incident "up/down 30% from the Previous Close" was the price collar (anything outside that was busted and anything inside stood).

Of course there is no guarantee that the same criteria will be used the next time around so caveat emptor.

Re: How I made $500k with machine learning and high frequency trading

#84

Post by demione »

beagle3 wrote: post #50
yajoe wrote: post #43

Earlier quoted context omitted.

This. I love crazy projects and Show HN's until the cows come home, but this one is dangerous that I must repeat the warning to others. I cannot emphasize how important it is to understand that people who trade using price action ( http://en.wikipedia.org/wiki/Price_action_trading ) are just speculating based on where they expect the price to move. It's no different than people who play Texas Hold'em online and specu…

The Instagram guys found an edge. It won them 730m. Fantastic. More than any edge ever won by me. But the market has changed so much since then, please be careful before you follow this course. You are not wrong, but what you wrote here is applicable to any success story posted on HN. Caveat lector. Always.

Agreed... in essence this is why we're all here. As entrepreneurs we all educated risk takers, and we realize any venture is essentially gambling if there is no edge. At any time, there could be a new idea that pushes any one HFT algorithm (or mobile photo sharing app, or words with friends clone) past the established mindshare into blue ocean territory. When that time comes, do you want to be caught with your pants down, lumbering under the excuse that you thought the oceans were too red for you to bother?

Re: How I made $500k with machine learning and high frequency trading

#85

Post by baq »

arbuge wrote: post #62

I am curious as to exactly why the profitability decreased steadily and rather rapidly all the way to ~zero. The article doesn't seem to expound on that unless I missed something. Is this a result of bots on the other side adapting in some way to what you were doing? I would have thought you would be too small a player for them to notice.

another comment explained: http://finance.yahoo.com/echarts?s=%5EVIX+Interactive#symbol...;

Re: How I made $500k with machine learning and high frequency trading

#86

Post by briancurtin »

kami8845 wrote: post #32
jspaulding wrote: post #21

Earlier quoted context omitted.

Yes, however, I forgot to mention that pretty early on I converted my program to use an API from https://www.tradingtechnologies.com . It's odd, but I can't remember exactly why. I think it was simply because I found a broker who could offer me a lower commission rate and they only supported TT. With regard to posting code yes I may do that. We'll see.

>With regard to posting code yes I may do that. We'll see. What does it depend on?

If I was correct in guessing that he used XTAPI (via X_TRADER Pro), IIRC that costs around $1k/mo depending on who you go through for the front end license. I was only a developer so I'm not sure what licensing costs are for gateways, but those are the servers you connect to in order to place orders on the exchanges. He was trading the Russel on ICE, and DAX on Xetra (maybe Eurex?), so he would have needed two gateways.

If he posts the code, you're a long way from running it.

Re: How I made $500k with machine learning and high frequency trading

#89

Post by hafabnew »

I've been considering trying HFT myself for a while. I'm competent with Machine Learning and am a Software Developer by day, so I can program and can sysadmin well enough to get something up and running without any trouble at all.

But, every time I've tried to actually get started, I've always found the amount of research required before being able to begin is just staggering.

It seems like the logical course of single-programmer HFT trading being:

- Find sample data

- Build your trading program using sample data

- When you're happy: connect to live API and set your trading program loose

- Iterate.

However, the first step and the third step seem like the ones which require the most research.

Is there somewhere which has a straightforward dump of timestamped market data available to download (free or not), in order to actually develop a working program?

Likewise, figuring out what to actually trade with, and which service to use is also pretty taxing.

Re: How I made $500k with machine learning and high frequency trading

#90

Post by hnruss »

Just because you CAN do it doesn't mean you SHOULD. Even if you don't think of it as "gambling", you're still taking in tons of money without providing any tangible benefit to society.

If you want to make money from investing, why not do so in a socially responsible way? Invest in companies that are changing the world for the better. You might not bring home as much money, but at least you'll be able to sleep well at night.

Post reply on HN