Hacker News Reader

I find this idea that all the investment managers in the world exist simply by skimming off peoples retirement savings mind boggling.

There are two main kinds of retirement savings, those where the retiree is in control of where their money goes - in which case they can choose to invest it as they wish. In an index, a company or anything else, and those where a company is investing on their behalf but with a legal obligation to pay out.

Now, if the company goes bust then that is not the fault of HFTs or managers skimming. Either they failed to make adequate contributions or they simply went bust because the business failed.

If you invest your money in ACTIVE MANAGERS then you'd best know who they are, what they do, and why they think they can win in a negative sum game. By definition very few actually can.

HFT is not the cause of pensioners going without - that is a combination of lack of contributions (based on unrealistic expectations) and poor investment decisions. They don't prevent you from spending all your money on Apple shares and making a killing, and nor are they to blame if you invested in Northern Rock and lost everything.

It is a fact that more money is spent on portfolio transaction costs for large pension schemes than any other cost. It is also true that much of it is unnecessary. But these are functions of investment choices based on active management, along with poor implementation and usually high management turnover.

What a lot of people are missing is "the market" can't grow faster than "the economy" over the long term. There has been a huge influx of money into the stock market as an increasing percentage of people hoping for historical returns. Causing people to chase after ever lower returns. It's gotten so distorted that the 'smart money' practically ignores growth in favor of other games.

HFT is the perfect example of this as they don't chase growth. There goal is to tax cash flows and market inefficiency. So, they leave low churn stocks like Berkshire Hathaway alone in favor of cheap stocks with a lot of turnover. If you keep your stock for an average of 10 years then HFT is meaningless to you. But, with hedge-funds often doing quite a bit of trading they can extract money without you realizing your trading.

In the end it's yet another reason 401k's are growing a lot slower than many people predict. Sort of like how people expect the economy to 'recover' when it was what there remembering is an unsustainable bubble.

PS: And of course Dividends are something of a special case in the above analysis.

I disagree that the market can't grow faster than the economy. Derivative markets for example.

Note: over the long term

Compound interest is an exponential equation. Even 1% growth means doubling every 70 years and 1,000x growth every 700 years and 1 million x growth in 1400 years etc. Let's say US GDP is X$ and stocks are 'worth' a 100,000 X. At some point the market get's so far from the 'fundamentals' you get a crash, but it's more like a return to rational behavior.

PS: That's not to say tax breaks like 401k's cant shift things for decades. But, there is only so far you can inflate any bubble before it pops.

In terms of overall capital - money - this should be self evident. Money is a proxy for resources. Resources including labor are finite. There are only so many resources to chase. Even if the market extends control to all the world's resources there is a limit on its growth. Any numerical growth after that means nothing but inflation.

There are lots of ways to extract rent from the system.

I've read some stories where a pension / 401k plan was placing a large trade and HFT was able to detect it in progress and make money off the transaction. Look up front running, placing bids and withdrawing them, etc.

Even if you trust your investment manager, there are shenanigans going on in the system. The stock market isn't what it used to be.

> This can all be proven by something called the efficient market hypothesis.

Of course, the efficient market hypothesis is basically an unproven assumption.

The general gist of it is obviously somewhat reasonable, but you have to be extremely careful what kind of statements you really believe.

Just to make two obvious examples: the strongest form of EMH claims that asset prices reflect even hidden information. That cannot possibly be true, because the circle of people with access to that hidden information most likely does not have enough money / market power to affect a shift of sufficient size.

The other example: slightly weaker forms of EMH still claim that prices adjust immediately to new information. That cannot possibly be true either, because traders simply aren't that fast. HFTs typically cannot afford to put so much money into affecting big price shifts, and other traders are slower. Besides, it's clearly falsified empirically: if this form of the EMH were true, stock prices would look like step functions, and they don't.

Another argument against EMH is complexity theoretic: one can most likely show that if EMH were true, the market would solve NP-hard problems efficiently (I remember seeing a line of work that attempted to show such a thing, though obviously the modeling is very messy).

Now you might argue that EMH still somehow approximates the real world. Maybe. But small errors can add up, so you have to be really careful in your thinking.

> That cannot possibly be true either, because traders simply aren't that fast.

Forget traders...our universe has a speed limit.

> Contrary to what the article states, computers that compete against each other will tend to eliminate all short term unjustified price swings and leave only accessible prices based on real company fundamentals.

This is assuming that computers only try to extract money by going long on every stock they purchase, an incredibly ignorant view of finance.

It seems to me to be naive to believe that a bunch of computers running software built by humans will compete perfectly to create a perfect market, the hypothetical "efficient market".

Instead what you would probably arrive at is an arms-race of retail investor-accessible HFT systems competing for investor dollars. That arms-race would just lead to a market even more arbitrary than the one we have now where the HFT systems are using fundamentals and at the same time learning about the trading patterns of the other systems indirectly to exploit them. Investors would be investing in the trading system instead of the real market, it would be reduced to gambling on the race instead of investing in the market.

My personal opinion is the HFT systems of today add zero value, as argued in the article. The liquidity provided by them is an illusion as they skim a cut off between the original seller and their buyer on the two ends of the trades they perform.

While ultimately I do agree with you, especially about humans trying to beat HFT in the short term, I don't think HFT and algorithmic trading are _currently_ the panacea you are making them out to be.

http://en.wikipedia.org/wiki/2010_Flash_Crash

I agree that the system is not perfect because, lets face it, computers are not always that reliable. However, this is a problem that solves itself. If an algorithm goes haywire and drives the market down temporarily, that computer stands to lose a whole lot of money so there is tons on incentives to fix or retire that particular machine.

Otherwise it would be easy to make a very profitable algorithm on the back of the machines that create flash crashes by detecting those crashes and buying tons of shares when they are discounted in order to sell them back when prices returns to rational levels. By buying on the crash this algorithm would push the price up thus reducing the price swing.

Google the term "bear attack," you aren't accounting for the possibility that the algos can (try to) kill a company.

Tried it. Only found ursine mauling.

Tried duckduckgo too. Got tired of scrolling.

I see algorithmic trading as not a symptom of, but one of the solutions to the problems in investment banking This is incredibly naive. Amazingly so.

"This can all be proven by something called the efficient market hypothesis" really? This is even worse.

Front running is a betrayal of the fundamental behavioural "assumptions" underpinning economic theory. By definition, without these assumptions, there is no EMH. So this whole thing falls apart.

That's why this is such a problem. If you really believe in markets, the first thing you do is make sure their underpinnings are not corrupted.

Its not clear that OWS believes in markets. The critique I'm outlining above is a seperate position alltogether.

I'm just going to handle this one point: "Nowadays every time a human trader tries to beat the market by predicting short terms swings in prices, there is a computer on the other side of his internet connection that is relying on much more information and even models of the human's own personality and biases to trade against him. This means that unless the human has inside information that is not accessible to the computer, (and I'm sure those who operate these computers try to feed them or make them infer the greatest amount of direct or indirect insider knowledge as possible ) it will tend to be impossible to beat the market."

This isn't accurate. Computational trading uses models and algorithms. Even if the computer has all the information necessary, sometimes the models themselves have flaws. We're not at the point yet where the computers are making these perfect, predictive models. The humans who make them will make incorrect assumptions. We saw this issue big time in 2007, and we'll see it again.

Assuming that in the future most trading will be done by computers, this means that in a large part the market will be deterministic, or at least predictable, but not rational or contributing to a better economy. The whole idea of stock markets is to have people apply their sensibility so that the economy moves towards what will be "a better future". Since only people know what is good for them, and 'good' is not possible to define as a specific set of optimization problems, i don't see what is the value of an automatically-trading market.