Ask HN: How to get into quantitative trading? • Hacker News Board

Re: Ask HN: How to get into quantitative trading?

#61

Post by bluepizza » Sat, Jan 27, 2024, 8:02 AM UTC

TheAlchemist wrote: post #51

bluepizza wrote: post #49

Earlier quoted context omitted.

It was closed to external investors in 1993, and had gains of 98% in 2008, in the middle of an economic meltdown. How come they figured it out for one single fund, but aren't able to make any alternative strong strategies except the secret and closed one?

Strategies have limits on how much money they can manage. By limits, I mean - either it's impossible to put in more (could be because there are no more counterparties to trade with above a certain amount) or it would significantly decrease performance.

And in the 30 years since closure, which has seen China, USA, and Europe boom and multiply in size, somehow the strategy still has limits?

My questions are valid, and there might be valid answers. We don't know, because nobody knows how Medallion works. All we know is that nobody else can replicate it, not even RenTech itself. Which makes asking questions very important.

Re: Ask HN: How to get into quantitative trading?

#62

Post by Imanari » Sat, Jan 27, 2024, 9:01 AM UTC

Hit me up if you want to chat. I‘ve been a hobbyist algotrader for some years (with mixed results).

I’d say this: It is very hard to beat the market consistently. It is even harder to statistically prove and convince yourself that our new strategy actually now beats the market. There are a lot of gotchas and caveats to watch out for when backtesting.

I spent most of my time with time series techniques as this was most fun to me. My current stack is ccxt, binance, polygon.io and self made backtesting in python.

Re: Ask HN: How to get into quantitative trading?

#63

Post by affyboi » Sat, Jan 27, 2024, 2:14 PM UTC

jairuhme wrote: post #45

chronic74930 wrote: post #21

Earlier quoted context omitted.

> You should know this. Anyone who worked in quantitative trading knows you can consistently outperform the S&P500/VTI/SPY over 20+ year horizons. Buy and hold passive US equity index funds gives a Sharpe ratio of ~0.7 with annual return of 8-10%. Meanwhile, high frequency trading does upward of Sharpe 10+ with 40-60% annual return, with 10+ year track records, of course. Hell, even Citadel hedge fund, after their ri…

Citadel isn't making the bulk of their alpha from simply finding an edge, its making trades in dark pools based on trade information they receive for executing trades. The individual trader won't be able to make trades at the speed required when finding small price discrepancies, nor would they have the capital to really earn anything.

Citadel securities is the market maker, Citadel is the hedge fund and doesn’t have access to data from the other side

Re: Ask HN: How to get into quantitative trading?

#64

Post by affyboi » Sat, Jan 27, 2024, 2:15 PM UTC

chronic74930 wrote: post #21

worik wrote: post #17

You should know this is a really bad idea, unless loosing money is your goal "Automated" trading, even if there is an edge (and unless you are a crook, there is not) will sink you with transaction costs You should know this.

> You should know this. Anyone who worked in quantitative trading knows you can consistently outperform the S&P500/VTI/SPY over 20+ year horizons. Buy and hold passive US equity index funds gives a Sharpe ratio of ~0.7 with annual return of 8-10%. Meanwhile, high frequency trading does upward of Sharpe 10+ with 40-60% annual return, with 10+ year track records, of course. Hell, even Citadel hedge fund, after their ri…

> Hell, even Citadel hedge fund, after their ridiculous 50% performance fee

Do you have a source?

Re: Ask HN: How to get into quantitative trading?

#65

Post by bmitc » Sat, Jan 27, 2024, 2:58 PM UTC

bluepizza wrote: post #61

TheAlchemist wrote: post #51

Earlier quoted context omitted.

Strategies have limits on how much money they can manage. By limits, I mean - either it's impossible to put in more (could be because there are no more counterparties to trade with above a certain amount) or it would significantly decrease performance.

And in the 30 years since closure, which has seen China, USA, and Europe boom and multiply in size, somehow the strategy still has limits? My questions are valid, and there might be valid answers. We don't know, because nobody knows how Medallion works. All we know is that nobody else can replicate it, not even RenTech itself. Which makes asking questions very important.

> All we know is that nobody else can replicate it, not even RenTech itself.

Several firms have done similar things.

Re: Ask HN: How to get into quantitative trading?

#66

Post by bluepizza » Sat, Jan 27, 2024, 3:12 PM UTC

bmitc wrote: post #65

bluepizza wrote: post #61

Earlier quoted context omitted.

And in the 30 years since closure, which has seen China, USA, and Europe boom and multiply in size, somehow the strategy still has limits? My questions are valid, and there might be valid answers. We don't know, because nobody knows how Medallion works. All we know is that nobody else can replicate it, not even RenTech itself. Which makes asking questions very important.

> All we know is that nobody else can replicate it, not even RenTech itself. Several firms have done similar things.

Do you have examples? Googling is not yielding any results for me.

Re: Ask HN: How to get into quantitative trading?

#67

Post by tim333 » Sat, Jan 27, 2024, 7:10 PM UTC

I've recently been studying up on it with a view to unleash the bots soon. I can recommend a couple of Udemy courses, details below.

There's a guy here explaining how to make trading bots in Python https://youtu.be/sUQmuL95_oY He has free code available on github to try out. He also explains how to backtest. There are a couple of popular packages like backtesting.py out there.

ccxt is a handy python package if you want to trade crypto. There are various free tutorials for it.

I also recommend playing around with Tradingview. It draws graphs and you can make your own indicators with their language Pinescript and even trade off that although Python probably gives you more flexibility.

Udemy courses https://www.udemy.com/course/algorithmic-trading-strategies-... and https://www.udemy.com/course/complete-algorthmic-forex-tradi... I paid about £14 each - they keep chopping and changing with the discounts

The details vary a bit depending what you want to trade - crypto, stocks fx or whatever.

One tip - chatgpt4 is surprisingly helpful and can almost code a whole bot for you if you know what you want. One slight pain for me has been finding brokers with the what I want. I'm the wrong nationality or they don't have futures or there isn't enough liquidity or they are sketchy or such like.

Re: Ask HN: How to get into quantitative trading?

#68

Post by tim333 » Sat, Jan 27, 2024, 7:24 PM UTC

worik wrote: post #17

You should know this is a really bad idea, unless loosing money is your goal "Automated" trading, even if there is an edge (and unless you are a crook, there is not) will sink you with transaction costs You should know this.

Transaction costs can be surprisingly low these days. Like in crypto you can get 0.02% maker fees or less and in stocks it's sometimes free. See Alpaca for example https://brokerchooser.com/broker-reviews/alpaca-trading-revi... - "Commission-free stock and ETF trading"

Re: Ask HN: How to get into quantitative trading?

#69

Post by MuffinFlavored » Sat, Jan 27, 2024, 7:41 PM UTC

nurettin wrote: post #57

MuffinFlavored wrote: post #39

Earlier quoted context omitted.

Slightly related: what are "big players" more likely to do to get exposure to the S&P directionally: 1. SPY shares long/short on margin/leverage 1. SPY options 1. SPX options 1. /ES e-mini futures 1. a blend of all Does one trump another in popularity?

it's usually front month ES, depending on volume. And MES for smaller accounts.

I'd love to talk to you more. I tried to make a Mastodon account to reach you. I just have questions about how you might know this/what else I could learn from you. I don't trade or anything like that, just interested in learning if you'd like to chat more.

Re: Ask HN: How to get into quantitative trading?

#70

Post by jleahy » Sat, Jan 27, 2024, 8:09 PM UTC

MuffinFlavored wrote: post #53

jleahy wrote: post #48

Earlier quoted context omitted.

The spread between SPX and ES is purely mechanical. It’s a function of expected future interest rates and dividends over the remaining life of the future. There is no such thing as support/resistance in reality.

> There is no such thing as support/resistance in reality. Where would you say 80% of the daily trade volume comes from on average? The powers to be that I can think of: institutional investors / fund managers slowly reallocating (selling stuff off, buying stuff) daily high frequency trading algorithms trading shares back and forth to each other in an artificial way to generate synthetic volume/movement market makers…

In my experience a large % of the daily trade volume comes from me.