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

Algorithmic Trading: A Practitioner’s Guide

henrikwarne.com

Post reply on HN
Print view View original

61–70 of 149 posts

Re: Algorithmic Trading: A Practitioner’s Guide

#61

Post by popinman322 »

c_o_n_v_e_x wrote: post #46
paulpauper wrote: post #24

Trading seems dauting, especially when your competition are HFTs and huge firms, but there are even very simple patterns that can be profitable, that does not require any advanced coding, APIs, huge troves of data, quant formulas, etc. Once such simple method, which still works, is to short BTC and go long QQQ/SPY during market hours if there is relative weakness of BTC before the market open, whilst going long QQQ/S…

>Pattern recognition, intuition are more valuable when to comes to trading than having more data or better tools. I have a few friends in the equities business and this topic always comes up over drinks. It would seem that in the age of GPU farms and open source ML tools, are we to a point where patterns are so subtle or short-lived that only a machine could pick up on them?

IIRC we still don't have very competitive neural network models for time series forecasting. It's a very active area of research.

Re: Algorithmic Trading: A Practitioner’s Guide

#63

Post by hackerlight »

nemo44x wrote: post #35

Is there a more socially unproductive, legit practice than this?

Many things, I'd argue. Meat industry causing animal suffering. Social media, tobacco, fast food industries causing human suffering. Oil and gas industry causing environmental destruction.

What direct suffering does algorithmic trading cause? It would take quite the narrative spin to argue that it's as bad or worse as one of the above industries.

Re: Algorithmic Trading: A Practitioner’s Guide

#64

Post by hackerlight »

nemo44x wrote: post #45
auntienomen wrote: post #39

Earlier quoted context omitted.

This is an article about how pension funds and other large players use algorithms to reduce their trading costs. You have a problem with that? Or did you not know you were in a Wendy's?

Yeah that’s maybe an edge case at best. It’s nickel vacuuming in most instances.

One of the largest uses of algorithmic trading is not an edge case.

Re: Algorithmic Trading: A Practitioner’s Guide

#65

Post by SamReidHughes »

TacticalCoder wrote: post #20

> There is also a special order type called post-only. It is designed to only supply liquidity, never take liquidity. If the market moves between the decision to send out an order, and the order reaching the exchange, the order will not cross. Instead, it will be hidden, or cancelled. This makes it easier for algorithm designers to get the behavior they intend (that is, resting orders will not accidentally be convert…

I would argue that it provides liquidity solely to whoever placed that resting order. But it takes it from the rest of the market.

Re: Algorithmic Trading: A Practitioner’s Guide

#66

Post by sarangab »

After using systems like Quantopian, Tradingview/pinescript, thinkorswim, metatrader etc for years, decided to build an own system from scratch.

We're building Invsto (https://invsto.com) to make algorithmic trading easier for individuals. Happy to chat with anyone interested in the space.

Re: Algorithmic Trading: A Practitioner’s Guide

#67

Post by Kranar »

Panzer04 wrote: post #60
Kranar wrote: post #52

Earlier quoted context omitted.

>Someone who merely bought and held tech stocks, like Apple & Nvidia, beat virtually all funds since 2009. This is a common misconception or a poorly phrased statement. It's not true that someone who bought/held tech stocks, or an ETF beat virtually all managed funds or that holding on to ETFs beats virtually every managed fund. It's true that passive investing, in tech or ETFs would have beat the average managed fun…

That largely depends on the type of leverage used (not all individual margin loans have the same conditions). It's possible to get pretty good terms as an individual in some circumstances (mostly if the loans are smaller and personally guaranteed), mostly by getting loans without margin calls attached. If you own a home, you can trivially borrow against it to invest without the risk of a margin call. It's all tradeof…

Nothing you've said has any relevance to this discussion. Regardless of what kind of leverage you use, if you're the one using it then you can end up with a negative balance putting you in debt. Case closed.

As for your other comment trying to be pedantic about funds owning three stocks, there are numerous publicly traded leveraged funds that trade just a single stock, one single stock [1]. They are known as single-stock ETFs and the purpose of these funds is specifically to provide an indirect form of leverage to investors. For example, IRA accounts are forbidden from using leverage, but someone can use an IRA account to purchase a leveraged ETF including a single stock ETF.

[1] https://www.direxion.com/single-stock-leveraged-etfs

Re: Algorithmic Trading: A Practitioner’s Guide

#68

Post by Galanwe »

synergy20 wrote: post #31

is this the same as 'quant trading'? I heard fresh outs who got offer from quant-trading companies making like half-million a year before their bonus.

The term "quant" has many meanings.

Originally it used to mean "quantitative", as opposed to "qualitative". It is a type of investment strategies where you target to be right "on average". Once you are right, say, 51% of the times, you aim to reach this asymptotic behavior by trading more and more. Either horizontally (trading more stocks) or vertically (trading more often). You want to keep the law of large numbers on your side to consistently earn that 1% edge that you found.

This contrasts with qualitative investment strategies (sometimes called "discretionary") where you target very precise and punctual events on which you have a very high (say 90%) chance of being correct.

You can imagine both these strategies with a coin toss game.

With a quantitative approach, you would try to find something that allows you to have just even barely more than a 50% chance of winning. Once you find that, you want to bet as much as possible. This is close to the strategy of a casino: they have games which all have an ever so slightly positive expected value, then they just have to make a lot of people play these games.

With a qualitative strategy, you would study very hard to find an event when you can predict at 90% chance the result of the coin toss. You don't play the game until this event is about to happen, and you bet big once it is about to happen.

Nowadays, the term "quant" has a broader meaning, which roughly encompass any kind of financial work which is heavy on math, or sophisticated. You can even find "back office quants", "pricing quants", etc

Re: Algorithmic Trading: A Practitioner’s Guide

#69

Post by rustamm »

Kranar wrote: post #52
paulpauper wrote: post #30

Earlier quoted context omitted.

Someone who merely bought and held tech stocks, like Apple & Nvidia, beat virtually all funds since 2009. There are always ways to make money even when your competitors have such advanced tools. The world of finance is big enough that there are opportunities for players of all sizes and resources. Look how badly AQR has done despite hiring from such a qualified talent pool.

>Someone who merely bought and held tech stocks, like Apple & Nvidia, beat virtually all funds since 2009. This is a common misconception or a poorly phrased statement. It's not true that someone who bought/held tech stocks, or an ETF beat virtually all managed funds or that holding on to ETFs beats virtually every managed fund. It's true that passive investing, in tech or ETFs would have beat the average managed fun…

Just because a managed fund outperformed a market doesn't mean it didn't happen by pure luck. There are lots of managed funds and most of them are not profitable. If each chooses portfolio randomly, some of them will outperform the market.

Re: Algorithmic Trading: A Practitioner’s Guide

#70

Post by Galanwe »

siftrics wrote: post #49
porb121 wrote: post #41

Earlier quoted context omitted.

if someone bids 10000000 shares for mid, and I hit their bid, who provided the liquidity?

You can argue about this philosophically, but if you talk to literally anyone in the industry, they will understand that - "taking liquidity" is taking existing orders off the orderbook and - "making [liquidity]" is opening new orders that rest on the orderbook It's the terminology of the industry.

To add on your comment: I don't even think the philosophical discussion makes sense.

At the end of the day, what counts as the truth is how much fees you pay when sending a limit order that immediately crosses.

And the answer, for every single market on the planet, is "you pay taker fees". Period.

Post reply on HN