Hacker News new | ask | show | jobs
by vostok 3499 days ago
Even as a liberal I think there's actually something to this idea. You can see it with things like the financial transaction tax proposals that are clearly meant to hurt nerdy HFTs rather than affect banks. I think part of it is that nerdy HFTs aren't eloquent enough to explain what they're doing and how it benefits the world.
2 comments

I would love to hear an explanation of how beating the next guy by 1ms is helping the world enough to justify obscene payouts.

Go for it.

So first, why it's important for me that we're able to trade continuously. Suppose I'm an HFT who trades the S&P 500 ETF. I put out a buy order at 217.75 and a sell order at 217.76.

If a trader comes in an buys the ETF from me at 217.76 then I need to be able to hedge immediately in other products. Without continuous trading I would need to quote wider.

Second, HFTs make less money per dollar traded than the people who used to do this. In a sense, HFTs "won" because they were willing to do this cheaper than anyone else.

This is different from investment banking, where it's my impression that they get paid these huge fees because CEOs hire their friends and pay them with shareholders' money.

Lay out the case for the payouts being obscene.

Probably keep in mind that the nominal value of the stocks changing hands on a given day is hundreds of billions of dollars and that an actively trading counterparty is facilitating the movement of that money. What's fair compensation for that?

How did markets function prior to HFT? They're only facilitating exchange between other algorithms, right? And in that case how is it helpful to the market?

Note I'm actually asking these questions - not just being rhetorical. The whole concept of it seems plainly ridiculous to me but I don't know much about this area.

How did markets function prior to HFT?

You had 10 very tall loud guys in brightly colored jackets yelling at each other in a pit. Mentally they were running the same algorithms as HFTs, just slower.

Later on you had some fast fingered guys watching charts and mentally running the same algorithms as guys in the pit, just a bit faster.

Except the people were exploiting inefficiencies in markets (and still do today). HFT algorithms exploit inefficiencies in servers, ISPs, fiber optic cable, a competing algorithm's implementation, etc.
And the tall loud guys exploited inefficiencies in trading pits and human perception. Ask yourself why I explicitly described them as "tall loud guys in brightly colored jackets"? Those are not irrelevant details.

HFTs do the exact same thing that pit traders did. They just do it faster and cheaper.

And if you really think they don't provide a useful service, you can very easily NOT buy their services - just change one flag in FIX. The question to ask yourself is why everyone chooses to trade with them.

https://www.chrisstucchio.com/blog/2014/how_to_not_get_rippe...

Those things can't be meaningfully separated from the market.

Exploiting inefficiencies in a competing algorithm is the same type of thing as exploiting inefficiencies in the mental model of the guy standing next to you. Taking advantage of better market proximity is the same type of thing as drinking less than the other guys in the pit.

I think viewing the HFT tax propoals as being motivated by anti-nerd rather than a out accumulation of wealth that is perceived to be at the expense of people like the tax supporter where the tax supporter fails to see the social utility of the function producing the accumulation of wealth is, well, a significant case of seeing what you want to see rather than what is actually there.
Go read Flash Boys. It's almost literally "Revenge of the Jocks". You have a jock (Katsuyama) who is naively bro-ing down and schmoozing clients while selling button pushing services to them.

Suddenly some evil nerds come along and use computers to "pick him off"! ("Pick off" is a wonderful phrase Michael Lewis uses but never defines, but it somehow involves computers doing evil things.)

Finally, at the end, Katsuyama builds a new exchange with no nerds allowed.

I was planning on reading Flash Boys at some point but if this is your overall description of the book my enthusiasm has been tempered.

Have you read, and if so, would you recommend I read Flash Boys: Not So Fast by Peter Kovac for a more balanced/informed perspective on the subject? Should I read both? Keep in mind my knowledge of HFT is extremely limited.

I have not read "Flash Boys: Not So Fast", so I can't comment on it. I do however recommend completely skipping Flash Boys - it provides almost no new information and a lot of FUD.

If you want an introduction to HFT, I'll shamelessly plug my blog posts which describe the basic mechanics of it:

https://www.chrisstucchio.com/blog/2012/hft_apology.html https://www.chrisstucchio.com/blog/2012/hft_apology2.html https://www.chrisstucchio.com/blog/2012/hft_whats_broken.htm...

I also wrote a few posts later on about controversial HFT topics: https://www.chrisstucchio.com/blog/2012/flash_crash_flash_in... https://www.chrisstucchio.com/blog/2014/fervent_defense_of_f... https://www.chrisstucchio.com/blog/2014/quote_stuffing_is_a_...

My posts probably read more like a "how to program python" tutorial than a compelling Michael Lewis narrative, however.

The book Trading and Exchanges ( http://amzn.to/2f0qtJb ) goes into a lot more detail than my blog, but it costs about $50-$90.

I appreciate your response. I'm familiar with your blog but I haven't set a time aside to review your posts on HFT. I will certainly do so now. Thanks for linking me the relevant posts.
I was thinking more of outside support for the proposal. Pretty obviously, the inside-the-industry support is just supporting burdens imposed selectively on competing business models, which is common independent of nerd/jock associations.