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

NautilusTrader: Open-source algorithmic trading platform

nautilustrader.io

Post reply on HN
Print view View original

81–90 of 134 posts

Re: NautilusTrader: Open-source algorithmic trading platform

#81

Post by lerchmo »

OutOfHere wrote: post #74
mtlynch wrote: post #65

Earlier quoted context omitted.

> I used to trade options and had about 99.5% success on all my trades Do you mean this literally or is that an exaggeration for effect? I'm not sure how you'd do that unless all your trades are like selling a put with a $50 strike price expiring in a month when the stock is trading at $100.

> Do you mean this literally or is that an exaggeration for effect? As a fellow options trader, I can confirm that it is spot on what is seen with SPY; it is not an exaggeration at all.

assuming this is something like covered calls?

Re: NautilusTrader: Open-source algorithmic trading platform

#83

Post by lotsofpulp »

jjangkke wrote: post #43

I used to trade options and had about 99.5% success on all my trades The problem is the 0.5% of the time, it erases all the gains made on the successful ones. I'm convinced without information edge or some capital sunk cost edge (for HFT) you are literally just flipping coins when it comes to trading. What's dangerous is fixation on strategies that form after a period of success. All in all, I think just buying stock…

> I'm convinced without information edge or some capital sunk cost edge (for HFT) you are literally just flipping coins when it comes to trading.

This applies to everything in life.

Re: NautilusTrader: Open-source algorithmic trading platform

#84

Post by flessner »

jjangkke wrote: post #43

I used to trade options and had about 99.5% success on all my trades The problem is the 0.5% of the time, it erases all the gains made on the successful ones. I'm convinced without information edge or some capital sunk cost edge (for HFT) you are literally just flipping coins when it comes to trading. What's dangerous is fixation on strategies that form after a period of success. All in all, I think just buying stock…

That's a common pattern in trading strategies with negative skews or tail risks. Even large hedge funds, like LTCM, can fall into this same pitfall.

For anyone interested, I can recommend the book "Systematic Trading" by Robert Carver. You don't have to be into algorithmic trading, the sections on risk management and positive vs negative skews are already worth the read.

Re: NautilusTrader: Open-source algorithmic trading platform

#85

Post by mhh__ »

andrepd wrote: post #75
jjangkke wrote: post #43

I used to trade options and had about 99.5% success on all my trades The problem is the 0.5% of the time, it erases all the gains made on the successful ones. I'm convinced without information edge or some capital sunk cost edge (for HFT) you are literally just flipping coins when it comes to trading. What's dangerous is fixation on strategies that form after a period of success. All in all, I think just buying stock…

Option trading is notoriously difficult. HFTs simply don't take a position at all: they build portfolios that are as neutral as possible in every measure (not only delta, but in vol, its higher order terms, higher order delta terms, etc). This is very difficult to do unless you have substantial capital to absorb short term fluctuations, but if you do you can capture a consistent and very low risk profit.

HFTs are typically market makers so they don't really build a portfolio as per se anyway (in practice you can target a certain position obviously)

Re: NautilusTrader: Open-source algorithmic trading platform

#86

Post by mhh__ »

SilverElfin wrote: post #59
iav wrote: post #36

This is pretty comprehensive. I work at a quant firm, and we don't even have some of this implemented in code. The tricky part is always going to be the integration. Nautilus has its own OMS system, but so does IBKR, and there is no guarantee that they are going to match. For very small funds, running entirely on IBKR platform (or Alpaca if you can live with their constraints) makes sense. For very large funds, you i…

Since you work in a quant firm, I am curious if there are any good public resources to learn about the techniques used by such firms. Aimed at a novice I mean. I’ve always had a curiosity about it, but I feel like whatever I can find is more basic than what is truly being done in production.

You need a good background in mathematics and some knowledge of why you should be paid a premium by the market (e.g. carry, trend, various premia and so on)

Anything else requires millions of infra spend.

Re: NautilusTrader: Open-source algorithmic trading platform

#87

Post by chollida1 »

mathiaspoint wrote: post #42
iav wrote: post #36

This is pretty comprehensive. I work at a quant firm, and we don't even have some of this implemented in code. The tricky part is always going to be the integration. Nautilus has its own OMS system, but so does IBKR, and there is no guarantee that they are going to match. For very small funds, running entirely on IBKR platform (or Alpaca if you can live with their constraints) makes sense. For very large funds, you i…

I've heard ibkr will ban you if they can detect algorithmic trading.

The fact that they have an api says otherwise:)

Lots of small hedge fund use them as a stop gap until they get bigger, or fail.

Re: NautilusTrader: Open-source algorithmic trading platform

#88

Post by mapontosevenths »

OutOfHere wrote: post #71
adamiscool8 wrote: post #63

Earlier quoted context omitted.

This doesn't make sense. If you had a 99.5% success rate, you could simply use a fixed $ stop and clean up. I agree you need an information edge, but the difficulty of finding one inversely correlates with your scale. Buying and holding is the best strategy for almost everybody - but that is orthogonal to whether trading is a "coin flip" vs mining an edge.

It does make sense. A small fixed stop loss is just asking to be triggered. A large fixed stop loss will result in a zero-gain zero-loss scenario over many trades. More generally, there is no optimal amount of stop loss. It ultimately gets auto-stopped out at 3:30 pm, although by then it could have gone nearly to zero anyway. Either the strategy works or it doesn't, and with Trump manipulating the market on random da…

OutOfHere gets it. Think about it in terms of probability. The closer to the current price your stop loss is the higher the odds of it being triggered by random market fluctuations.

The same applies to your profit target. This means that if your profit mark is $10, and your stop loss is $5 you will lose roughly twice as often as you win, all other things being equal.

What you actually CAN do is use smart money management, (something like the Kelly Criterion) to ensure that you properly capitalize on any slight edge you do manage to find without going broke in the long term.

That of course requires you to find a bet you can win 51% of the time, and that you be made of iron when it comes to sticking to the plan. Most folks can't.

Re: NautilusTrader: Open-source algorithmic trading platform

#89

Post by adamiscool8 »

mapontosevenths wrote: post #88
OutOfHere wrote: post #71

Earlier quoted context omitted.

It does make sense. A small fixed stop loss is just asking to be triggered. A large fixed stop loss will result in a zero-gain zero-loss scenario over many trades. More generally, there is no optimal amount of stop loss. It ultimately gets auto-stopped out at 3:30 pm, although by then it could have gone nearly to zero anyway. Either the strategy works or it doesn't, and with Trump manipulating the market on random da…

OutOfHere gets it. Think about it in terms of probability. The closer to the current price your stop loss is the higher the odds of it being triggered by random market fluctuations. The same applies to your profit target. This means that if your profit mark is $10, and your stop loss is $5 you will lose roughly twice as often as you win, all other things being equal. What you actually CAN do is use smart money manage…

All of these points are irrelevant. If you have the data that shows 99.5% win rate, you would have the data on your expected win size, and could compute a fixed loss stop that would keep you on-side. In the example, which I'm assuming must be hyperbolic, your average win would only need to stay >1/199th of your average loss. I would agree that if a person cannot manage their positions to this, that person should stick to investing.

Re: NautilusTrader: Open-source algorithmic trading platform

#90

Post by OutOfHere »

lerchmo wrote: post #81
OutOfHere wrote: post #74

Earlier quoted context omitted.

> Do you mean this literally or is that an exaggeration for effect? As a fellow options trader, I can confirm that it is spot on what is seen with SPY; it is not an exaggeration at all.

assuming this is something like covered calls?

That isn't it for me, but that's not to say it couldn't be for others.
Post reply on HN