Clamping Down on High-Speed Stock Trades: US, Canada, European regulators | Hacker News Reader

Clamping Down on High-Speed Stock Trades: US, Canada, European regulators

nytimes.com

48 pointsgtani46 comments

It's supposedly regulators driving this crackdown, but almost all the quotes in the article come from competitors to high-speed trading firms claiming — surprise — that high-speed trading firms need to be regulated.

Not that they don't have a point, of course.

Stocks should only be allowed to trade at the speed that value is created or lost. The stock market is a human institution, (supposedly) designed to measure value of human businesses.

Trading in microseconds is ridiculous.

Trades should be in human time - essentially, minutes and hours, not seconds and microseconds. Business value does not change that quickly. HFT is representative of a financial industry set on creating ever more abstract tools and products that are more akin to gambling than representing the value of the actual business. The stock market needs to return to it's roots. There's been enough chaos already.

Humans do make decisions in milliseconds.

Humans make decisions in milliseconds, but market factors do not change that quickly. Even major, unforecasted issues like the Japan Tsunami happened in human time.

That has a feel good populist sense to it, but do you have evidence that your idea would improve things? What exactly is your proposal?

Computers are better at processing information then people, eventually they will be able to determine, in microseconds, what the market effects of world events are and execute trades. I think the world will be better off for it, not worse.

Why would the world be better off with computers making trades?

I don't have a proposal, except that the financial industry needs to be reviewed. A good start would be to remove the products that are abstractions of the value of companies or markets rather than representative of real value. Eliminate futures and most forms of derivatives. Keep shares and managed/indexed funds. Keep everything physical (e.g. Gold, Oil, etc).

That would reduce a number of the swings in the market caused by relatively small real stock movement. Return the market to trading value, rather than trading the anticipation of what might happen

The stock market should be used for trading companies, not shifting money from owners of stocks to the financial industry by having people manipulate it.

Eliminate futures

This statement is almost identical to "Eliminate automobile insurance", for reasons you don't understand. General advice: Don't speak on that which your ignorance is total.

First off, business value can change quickly.

Sudden events happen. You can think of dramatic events such as terrorist attacks or death of an important executive; but there's other information, such as the public announcement of a new product (e.g. 'Apple are making a new type of device!', perhaps rapidly followed by 'Google have just said they'll support it!') that also result in fast changes.

But, most importantly, the market price provides information that influences further decisions. Maybe I'm not an expert on oil prices, but I believe if it goes below $70 a barrel, its a signal that the world economy is slowing down, and I want to dump my Google stock, because I'm only willing to accept a certain level of risk in my personal investments.

The fact that I can see the price of the oil provides me with a signal (maybe noisy), derived from aggregated intelligence of other investors. This is a useful function.

Its completely legitimate to want to sell my Google stock in response to other market signals, and as quickly as possible, in response to the new information they provide.

The fundamental value of holding the Google stock, to me, has changed, fast, because of other information that's become available.

So fundamental business value can change fast, for a variety of reasons, and it can add value to be able to respond to quick changes quickly.

You've made the arguments for seconds. Maybe tenths of a second. A tenth of a second is pretty fast in my subjective perception of things.

I can't see the same argument extending credibly to milliseconds.

Why should we arbitrarily limit the efficiency of the markets? This just feels like some kind of luddite like fear of technology.

The markets have breaks in place to stop rapid crashes. The flash crash was caused by human placing a bad trade, quickly followed by the automated trading systems leaving the market.

Someone really has to make a much better case then this then pension funds being upset someone has detected their buy order and is driving up their purchase price.

The same reason we arbitrarily limit the gain of amplifiers, so the noise generated by flaws in the system doesn't keep feeding back until it drowns out the signal. Traders in capital markets are producing social value if they cause better decisions about allocating our resources, not the same decisions a fraction of a second sooner because of a greater misinvestment in network hardware. Front-running a trade that's already been decided on is just scalping.

be careful about the idea of clamping down on legal tax-paying businesses based on how much "social value" they are producing. It is very hard to judge that and it is subjective. One could argue,I am not arguing it but one could, that lots and lots of web startups produce very little social value. Should we ban them too?

Perception of business value can change quickly. Actual business value itself does not tend to change so fast. Dumping your Google stock if oil drops below $70 per barrel is solely based on your own perception and your own trading rules. The value of Google's products remain the same. If a competitor releases a new device, that is based in human time.

Does it make a difference if you sell at 12.01pm after a competitor's product announcement rather than 1pm? Right now, yes. Do I think it should? No.

The reason prices swing so quickly is because trades happen so quickly. If trading was slowed down, prices would not swing as quickly. The current situation reinforces a vicious cycle where the fastest trader has an advantage in the market.

Public share prices is a measure of how much the company is worth to people that buy shares not of how much value the company creates by selling their products. PepsiCo should be valued on how well they sell soft drink, not millisecond-by-millisecond analysis of unrelated issues.

There is no such thing as 'actual business value'. The very definition of value is based on perception; something of value is something people want. The closest concept I can think of to what you mean is "book value", which comes with its own set of drawbacks.

But what is 'actual business value'? Lets say we are considering an oil company, which makes a profit from drilling oil. If oil rises in price, the company makes more profit. Surely, in a very real sense, the 'actual business value' of this company, depends on the price of oil?

Lets talk about PepsiCo. The weather service detects a new hurricane forming off the coast of Florida, sending orange juice futures up. At some discrete point in time, the weather forecast computer model spits out its results, and that information is on its way to becoming public, and there's going to be a fast change in the price of oranges, and hence the value of PepsiCo.

Here is an article, describing just such an event:

" Orange juice futures hit a three-year high on Tuesday as speculators eyed the development of a Caribbean storm that meteorologists said had a chance of moving towards Florida. Orange juice is a $20bn industry with prices affecting farmers in Florida and Brazil, processors, trading houses and bottlers such as Coca-Cola and PepsiCo. " http://www.ft.com/intl/cms/s/0/1831a65c-c5a8-11df-ab48-00144...

Now, lets look at Google. Lets say Google does better in a more buoyant economy, because they get more ads (maybe the opposite happens in real life; but its just an example). And the drop in the price of oil has given me information that others think the economy is tanking. Hasn't the 'actual business value' of Google changed, in a very real sense? And quickly?

You might argue, if you believed in a deterministic universe, that the value of Google hasn't changed - the economy was always going to tank, because the universe only plays out one way. But that's irrelevant, because we don't have access to such a model of the universe. What we have to make decisions on is the information we have access to, which can change suddenly, and which the markets do a good job of aggregating.

I'm not trying to argue for front running the trades in the market or anything like that. I'm just trying to argue that business value can change rapidly.

I think the reason prices change so quickly is because aggregated belief about the value of companies changes so quickly, and because our markets allow that change in aggregated belief to be quickly expressed.

If you slowed trading down, so that it happened hourly, as you might propose, what this would mean is that the price of PepsiCo is essentially out-of-date for an hour. I could no longer look at the price of PepsiCo on the market, and assume it was doing a reasonable job of aggregating the public information about the future prospects of PepsiCo - I'd have to imagine it was an hour out of date. This would probably make the market less useful.

You rightly argue that a rapid change in perceived business value is possible and furthermore that it is in the interest of any given individual investor in performing very quick trades.

But your examples support relative gain, not absolute gain. Even after HFT, there will be some investors who get rid of the hot potato before others (or the opposite acquire an attractive stock). It's just that the one's who are the first to trade will have to pay more resources to access powerful technology. Situations like these arise in many non-cooperative games, like an arms race, where an improved technology leads to increased loss to both sides but is unavoidable unless there is a scope for cooperation via a treaty.

Furthermore, this leads to a systematic bias in favour of large investors and hedge funds who are the ones with resources sufficient to win the arms race.

Note, that after HFT, the market value synchronizes faster by a short interval of time. Maybe there is an argument that this leads to a large absolute gain in some form by the way it affects investors decisions, but I dont see how it does.

There has been relative gain by a very important class of investors: individual retail investors. If you trade individual stocks for your own account, you've seen a steady decrease in the amount of spread that you have to pay to get your order executed and you certainly shouldn't be deluding yourself by thinking that you are competing with high frequency/low latency traders or that they will be somehow disadvantaging your orders. The people who are at the root of these complaints are the buy side institutions who don't want to invest in technology to improve their order placement and the older exchanges who are either also trying to avoid investment or who are struggling to protect obsolete business models. If the various proposals to limit the rate of trading are enacted, then spreads will widen out even for the large buy side players (perhaps even by an amount larger than what HFT is currently costing them), and the only winners will be the entities that want to remain fat, dumb, and happily protected by regulators.

I just realized that I had a wrong idea about how HFT works. Though I am still not sure about how the pros and cons add up, your comment convinced me that it is more complicated than I thought.

HFT is a disruptive technology that has been the target of lots of FUD from entrenched institutions and politicians looking for a "scary" issue to flog. Most of the practitioners are fairly small (relative to other investment firms) and technology-focused. Unlike complex, bespoke derivatives and products such as CDOs, this aspect of the markets is fairly well regulated, with a focus on bringing transparency to pricing for individual investors, which is in stark contrast to the regulatory environment in Europe and Asia.

subsecond liquidity is definitely overrated and only serves to enrich firms who can afford colocation with the exchange while effectively taxing all other traders.

some kind of system where all trades are matched once a second or every other second, would provide all of the needed liquidity and level the playing field for all traders.

1st second: trades are accepted 2nd second: trades are matched and results are published repeat

that would reduce liquidity drastically. Think about this. Suppose all the quotes show up once a second: if the values of the assets in question are momentarily mispriced relative to some other assets somewhere in the world, HFT algorithms will attack those quotes and make money from taking liquidity instead of providing it. That would also increase adverse selection and reduce the incentives to provide liquidity, resulting in less liquidity for small investors

it isn't even a crude/primitive AI, it is just algorithms that are let loose for a limited periods of time in a limited activity space - and humans are already feeling threatened(xenophobic) as this algorithmic trading feels completely unnatural. (The reason of fairness is complete BS because professional traders have always had informational and speed advantage, it is just that it was on the human scale - seconds vs. minutes, and now it crossed into non-human microseconds scale)

Well, welcome to the 21st century. The algorithmic trading will look like a child play when in the "drone arm race", the sides would be forced to delegate final shoot decision to the drone algorithms for exactly the same reason - human speed wouldn't be enough as the millisecond advantage makes the critical difference between wining and loosing. That would be a reason to feel slightly xenophobic i guess, though we'll possibly get used to it :)

Wrt. unfair advantage vs. investors not utilizing this advanced tool of algorithmic trading - whole human history since the time of apes is just a continuous sequence of "slow" humans being ruthlessly overtaken by other humans taking [unfair like any advantage] advantage through better tools and weapons. We're direct beneficiaries of such an unfair advantage and, as we aren't the ultimate goal, just an intermediate point in this history, we'll be a victims of it in due course of that history.

Here's your +1 to get you out of gray area, but WRT "Wrt. unfair advantage" you should really read up on HFT in detail. All of HFT is based on the premise of getting an access to the order flow a split second sooner than everyone else. To get such access HFT firms pay very hefty fee to Nasdaq and other exchanges. That's where the unfair advantage is - not in the use of computers, but in an exclusive access to the information that not even all banks can afford.

everyone can purchase fast access. If a brokerage or a bank wanted to do it, they certainly could. It's like saying that Kinko's has an unfair advantage at making copies because it bought expensive copiers and regular folks have not.

I'd say it's more like Kinko's discovering you want to make some copies...and quickly erecting a store right outside your house so you stop there instead of Staples.

and what's wrong with that? Or rather, would you ban that too?

Yes, because that key word...discovering...implies that our rhetorical Kinko's has somehow gleaned advance knowledge that you're about to leave home to purchase some copies.

I don't think HFT in general is ruining things. I think the ability of HFT firms to quickly enter orders and quickly widthdraw those same orders millions of times a minute enables them to gain some insight into the trading market that the market was never designed for. And I believe that is a perversion of the original open-call trading that the stock market was built upon. All buyers should be able to see the ask at the exact same time.

I think you still got it wrong. It is more like Kinkos discovering you need to make some copies and then buying up ALL the paper and ink at staples, and then setting up shop right outside your house, with a higher price :).

I don't see anything wrong with this, it simply forces the market (theoretically) to hold on to valuable securities longer, which in theory would make it impossible for HFT firms to do this. In practice, everything is irrational, but it sure beats the pants off overregulation.

A better solution would be to reduce the granularity of trades to, for example, 1 second. Nobody needs to trade faster than that. If all trades are executed in a random order every 1 second rather than the current 'fastest fingers first' nothing would really change much in the world, just the HFTs would go away. HFT is a leech on the face of the financial world.

And why do you think reducing frequency of all trades to 1 second would eliminate HFT ? the term high frequency would simply mean 1 second. But what it would do is drastically reduce liquidity in the market place, which would hurt small investors and not the HFT.

So small investors did significantly worse in the market prior to the advent of HFT? That should be easy to test.

yes because spreads were wider and stock trading was routed to specialists on NYSE who could "jump the queue" and peek ahead at who they wanted or did not want to trade with. Also stock commissions were a lot higher in those days. Also, it was unfair because to have "fast access" required presence on the exchange floor which is much harder than putting a computer in the colocation facility

It's like saying that Kinko's has an unfair advantage at making copies because it bought expensive copiers and regular folks have not.

Who's the one making copy machines artificially expensive so that Kinko's can afford them and ordinary folks can't?

Uh,

The stock market is a human institution that theoretically is not merely a battle field but competition which serves to allocate resources efficiently and for the benefit of people, ideally all people but certainly some people.

Thus it seems legitimate to have regulations aimed to focus investment activity on questions that are really related to said resource allocation rather what is ultimately just clever front-running.

I mean, the world's production, distribution and profits process might argued to change in second these days. But it will be a while before world-resources need to be reallocated by millisecond.

What's the harm in HFT? It certainly increases liquidity in the market, making it easier for other investors to enter or exit a position on short notice. That's a good thing.

Let's not fall pray to generalized anti-finance populism without good reason.

If you read article, you may find that the argument is that the speed of the trades creates the perception of an unfair advantage and has even supposedly impelled larger buyers and seller to move the "dark pools". Now all this might wrong but it seems like a bit more than "anti-finance populism".

And the thing with the liquidity created by HFT is... if that liquidity is going to be suddenly withdrawn all at once, you've lost the advantage of liquid market, which is that you always have a reasonable chance of quickly selling at a reasonable price.

Maybe requiring registering and stress testing of HFTs over a specific volume is a solution. Or as a cost of registering a HFT is a requirement to "make the market" whether the HFTer likes to or not , there should be plenty of opportunities to hedge or offset any unwanted positions.

Interesting question. What if HFT algorithms become so efficient they are able to easily out-compete human trading decisions, such that humans exit the system altogether? What if the stock exchange becomes exclusively HFT algorithms competing against eachother on behalf of large funds, such that being a 'trader' is all about which fund to put money into (which algorithm suite to back)?

In such a thought experiment, what if the algorithms discovered more money could be realized by trading on a vector different to 'the current/future worth of a company'? Say a genetic algorithm determined that whether the share price ended the week's trading on an odd number rather than an even number is more important for predicting the future direction of the price on Monday. In the world of genetic algorithms, stranger shit has happened. What impact would this have?

I don't know much about trading, particularly HFT, so I don't even know if this is possible or realistic. I do love thought experiments though, and believe that HFT algorithms are currently trying to predict human responses to market events and act accordingly. Just wondered what would happen if they dominate a market such that human influence is negligible. So yeah, interesting question.

My suspicion (which I can't prove--this is just a gut feeling) is that there is an economically correct price for any given thing, and that markets will move toward that price. This economically correct price only changes when factors in the "real world" change, or we have reason to believe factors in the real world are going to change (e.g., bad weather might reduce crop yield, or forecasts of bad weather make is expect a reduction in yield).

Actors in the markets get information not only from real world sources (e.g., news reports), but also infer information from the activity of other actors in the market, so there is a bit of a feedback look in the market.

My suspicion is that as you increase the frequency of trades, starting from very infrequent (e.g., people trading in person directly with the person they are buying or selling from) and then going to faster and faster methods (trades by mail, then telegraph, and so on), there are two effects. First, the market is able to more quickly converge to near the "correct" price for the item, and there is some increase in fluctuation around the correct price due to strengthening of the feedback loops.

I suspect that there is some optimum trading speed, which depends on the rate that the real world events (things like weather changes, fashion, wars, and so on) happen, and when you push trading speed past that you stop gaining on convergence speed, but continue increasing instability due to the feedback loops.

The result is a market that is less and less tied to the real world. In effect, the market ends up making up information from the trading activity itself. The trading essentially ends up being based on noise rather than signal. It's hard to see how that can be good.

trading activity is information because it is real people putting real money at risk and expressing their views of what the assets are worth. As in any competitive market, HFT traders cannot make money if they only trade with themselves - that is a mean zero proposition. But by trading with people who price the assets incorrectly at a point in time, they help move asset prices to their "correct " levels

I keep saying that one of the biggest problems with HFT is the false illussion of at will liquidity.

This is what happened at the "flash crash", many traders providing liquidity, which included most of HFT scene getting out of market, because they (and their algorithms) decided it was not worth the risk.

Old market makers did not operate like this, they were providing quotes even in a very volatile markets. Of course, many lost their shirts doing so, but were obliged by the exchanges to provide this service.

If none of the liquidity providers have an obligation, then we have no liquidity in volatile times, when we need it the most.

yes but if there is an obligation, you could end up with less total liquidity because if you oblige traders to trade when it is a losing proposition to do so, you will get less traders and less liquidity all the time, including volatile times. That's why exchanges actually pay traders to provide liquidity and charge liquidity takers. Imagine if you ordered convenience stores to sell twinkies to people even if they have no money. How would that work in practice ?

HFT = https://secure.wikimedia.org/wikipedia/en/wiki/Algorithmic_t... + https://secure.wikimedia.org/wikipedia/en/wiki/Front_running

Front running is illegal.

All I read was "Oh noes, the computer, we don't understand it".

I would suggest that the people involved take the time to understand and learn the new branch of their art, instead of calling it "unholy", as if was some evil device by the antagonist of $religion.

Gosh. You mean allowing billion-dollar-poker by professional gamblers who have a six-orders-of-magnitude speed advantage over regular investors, and are also indemnified against any and all losses, might be dangerous? I'm shocked. Truly shocked.

No fucking duh. If ANYONE with any power were serious about fixing society or the world economy, this shit would be banned. Yesterday.

Reply on news.ycombinator.com