[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]
2023-11: Warosu is now out of extended maintenance.

/biz/ - Business & Finance

View post   
File: 72 KB, 260x260, Line-QR.jpg [View same] [iqdb] [saucenao] [google]
692262 No.692262 [Reply] [Original]

Why aren't you writing a computer program that trades stocks for you?

>> No.692264

Because of the latency to the NYSE. High-frequency traders have already done that.

>> No.692273

>>692264
This.

You'll never be able to beat Wall Street when they're literally renting space in the NYSE server rooms.

>> No.692428

I'm writing an algorithm futures trading platform.

>>692273
>>692264
Algorithmic trading != HFT

Unless you're competing in the HFT space (which you can't, unles you have hundreds of millions of dollars), you don't need < 1000 microsecond latency.

Linode's east coast datacenter gets you within 1ms of 90% of brokers/exchanges there for $20/mo.

>> No.692438

>>692262
Because I already have.

I have an auto rebalancing (based on Sun et al's risk adjusted return model) vanguard etf algorithm on Quantopian.

I found a simple stock API and ported everything to Node and made a simple private Android app that it talks to. I use GCM to ping me whenever my vanguard contributions clear, and it tells me what to buy and in what quantity. When it's an opportune time to rebalance, I'm notified with everything that I need to do.

How much would people pay for this?

>> No.692443

>>692438
You're a fucking idiot if you think that quantopian and backtesting prove anything of value

>> No.692490
File: 24 KB, 333x333, tumblr_inline_n4pegaqRW41r9vdbu.jpg [View same] [iqdb] [saucenao] [google]
692490

>>692438
>How much would people pay for this?
Heh. If it actually worked, you wouldn't need people to pay you for it.

>> No.692492

>>692490

Which is why we still don't know about the black-scholes formula.

>> No.692515

unethical

>> No.692520

Because I sell credit spreads in RUT and /ES futures that require my attention about once a week.

>> No.692522

>>692492
What don't we know about it?

>> No.692583

>>692443

The rebalancing algorithm is from MIT, cunt. I didn't come up with it. Super nerds did. I just transcribed it.

>>692490

It does work. It's not a super trading algorithm. It's not an "OMG watch me beat the index by 15%" algorithm like 90% of shit on Quantopian.

All it does is do a better job of timing rebalances than a band (%) rebalancer or a fixed interval rebalancer. And it makes you not have to do any math once your contributions clear. You just enter what it says.

My portfolio is otherwise lazy. I only hold VTI, VXUS, BND, and BNDX

>> No.692585

>>692583
Optimal Rebalancing Strategy for Institutional Portfolios - MIT

http://ssg.mit.edu/group/fan/papers/SSRN_opt_rebal.pdf

^It just implements this. And notifies you when trade/what to trade.

>> No.692613 [DELETED] 

>>692583
Are you sure it works? The markets have been bullish for years.