How I made $500k with machine learning and high frequency trading • Hacker News Board

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

#151

Post by OldSchool » Tue, Nov 06, 2012, 6:31 PM UTC

beagle3 wrote: post #48

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…

It is only gambling in the sense that any business is gambling: Your customers might stop coming tomorrow because the fad wore off, or a competitor provides a better/cheaper/hipper alternative. (And indeed, living is gambling. It's all just a matter of the risk/reward portfolie). But jspauld has apparently made $2/trade after fees on 250,000 trading, with a very small standard deviation (I would guess less than $2/tr…

Very true, every time you hire someone or spend money on Adwords you're gambling that you'll net more in $$$ or at least lifestyle improvement (in the case of a new employee) than your cost. That is, if you're in business to make a profit, not just spending OPM to build your brand. Even acquiring a customer is a gamble.

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

#152

Post by nesu » Tue, Nov 06, 2012, 6:32 PM UTC

Wait. This is not HFT. There's a huge difference between automated and high frequency trading. What he does is only automated scalping at best (or at the fastest).

Automated trading is more on strategy, while HFT has more to do with volume and speed. With automated trading, you predict price movements. HFT involves being a liquidity provider. You don't use market technical indicators in HFT, you wait for some really huge orders.

HFT firms won't bother him. Those are dealing with an entirely different set of algorithms. He should have contacted brokers instead.

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

#153

Post by HockeyPlayer » Tue, Nov 06, 2012, 6:33 PM UTC

stcredzero wrote: post #124

Is there a way to do this with Python or Ruby? I could just as well program this in C#, but I have a friend who can code a little, but doesn't really need everything in C# to do what he wants. The value add of offering the simulator, including the taking into account the bid/ask prices and a stochastic model for latency. Combine this with a web based code editor and easy hosting, and I think this would be a viable pr…

Sure, his program could easily have been done in almost any language. He wasn't competing on speed, which might have excluded languages like Python.

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

#154

Post by wyan » Tue, Nov 06, 2012, 6:36 PM UTC

hnruss wrote: post #90

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…

You mean, something like Instagram?

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

#155

Post by mattyppants » Tue, Nov 06, 2012, 6:41 PM UTC

dschiptsov wrote: post #101

mattyppants wrote: post #91

Earlier quoted context omitted.

This is not even close to an accurate summary. He never stated that he had any employer backing, and he wasn't collecting market making fees. In fact he was paying brokerage fees which is the exact opposite.

He used his own money to test his algos? Come on. Taking a Machine Learning and Statistics course does not make you a trader. You need something else - access to the system. This is why my guess is that he was an employee.

No, he said he backtested them against data he collected from the feeds. He paid for the feeds just like everyone else does with his own money.

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

#156

Post by HockeyPlayer » Tue, Nov 06, 2012, 6:42 PM UTC

I run a 12 person HFT group in Denver. This is an excellent description of how an individual can bootstrap themselves into success. Great story, and nicely presented.

The one interesting point that he glossed over is what his indicators were. He wrote, "The indicators that were most useful were all relatively simple and were based on recent events in the market I was trading as well as the markets of correlated securities."

Which doesn't really tell you much.

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

#157

Post by mattyppants » Tue, Nov 06, 2012, 6:42 PM UTC

unreal37 wrote: post #123

iandanforth wrote: post #72

I'm glad to see a healthy respect for investment among the hacker community. It's traders like this who commit to nearly a full 10 seconds of ownership that are the backbone of economic growth for this country.

I think there is a healthy respect for computer code that can let loose upon the world and make money.

Your missing the point that owning part of a company for 10 seconds brings zero social value to the economy. Which I also tend to agree with. Slightly off-topic from the point of the article, which was that he achieved it without backing of a fairly large institution.

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

#158

Post by HockeyPlayer » Tue, Nov 06, 2012, 6:45 PM UTC

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…

I run an HFT group, and what he describes isn't what we'd call "retail". He was doing a number of things that professional shops do, including making markets to avoid paying the spread and paying attention to queue position to predict execution

With a bit of luck and a good partner, this guy could have built a sustainable business.

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

#159

Post by owyn » Tue, Nov 06, 2012, 6:45 PM UTC

stcredzero wrote: post #124

Is there a way to do this with Python or Ruby? I could just as well program this in C#, but I have a friend who can code a little, but doesn't really need everything in C# to do what he wants. The value add of offering the simulator, including the taking into account the bid/ask prices and a stochastic model for latency. Combine this with a web based code editor and easy hosting, and I think this would be a viable pr…

Nope, there is absolutely no way to do this with python or ruby. :)

Seriously though, there are some existing frameworks and products that you could check out. I haven't used this myself, it's just in my bookmarks:

http://www.rapidquant.com/features

A C++ library with python bindings: http://quantlib.org/index.shtml

And a low level library for data analysis: http://pandas.pydata.org

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

#160

Post by marvin » Tue, Nov 06, 2012, 6:50 PM UTC

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 is a very mean and unconstructive comment to someone who made the impressive achievement of building his own automated trading system and actually making money from it. I've started calling out comments like this one, because they cause a bad environment for useful discussion.

The only argument in your comment that isn't your own unfounded opinion is that market makers make money from people who execute trades. But this is true by definition.

The traders who "gleefully picked off all those trades" weren't outsmarting anyone, they were simply profiting from the difference in the asking and offering price in the market. This is the role of a market maker, and actually makes it cheaper for people like OP to execute a large number of trades. So even though this comment sounds like a sensible rebuttal of the linked article, it doesn't really say anything at all.

Again, sorry for creating a negative reply and contributing to a bad tone, but I really the right thing is to call out these kinds of replies. They discourage honest sharing and discussion.

yajoe's comment is an example of how to criticize a post like this in a useful way. http://news.ycombinator.com/item?id=4748989