It's a zero sum game, but one might be smarter than others. Some companies doing algorithmic trading consistently skim off their share of the market's inefficiencies.
What I'm wondering is how much of an investment it would be to get into high-frequency trading for yourself. Any idea?
I love the idea (for the guys who run it AND the guys who it opens up opportunities for).
Might be something you could play with.
Depending on what sort of HF strategy you want to run, I imagine this can be done relatively cheaply. (300K-800k? Pretty wide range I know... that's all I got though).
Obviously your strategies are limited, you're not competing with some of the super low latency firms for the fastest of opportunities because of hardware limitations or price prohibitive vendors/deals with exchanges or whatnot.
IAMA thread looks very interesting, thanks (although you never know when it's someone pretending to be something he's not).
I'm a bit skeptical of algodeal. If they are seriously doing this (actually licensing strategies and following them), wouldn't they be setting themselves up to get ripped off? It seems that if you know what strategy someone is going to use, you can use it against them (effectively changing the market between historical tests and a live run).
I actually think AlgoDeal sounds great, but can understand your skepticism.
I haven't really looked into it, but what it sounds like to me is they provide you a platform to build strategies and test them, play with fake money. They may optionally take your strategy and run it in the market. If it is successful, they split returns with you.
When you say they're setting themselves up to get ripped off, I think you're suggesting that Evil Person X may come up with a strategy A that does well enough in backtests, but is paired with a strategy B designed to rip off that strategy A. Then after AlgoDeal applies strategy A, I put my secret millions to work and make better returns for myself.
They don't HAVE to run your strategy. I'm assuming they're smart guys and have some selection criteria other than "performed well in backtests", so they'd try to choose strategies that were less likely to get ripped off by the evil person's counter-strategy. Also, I think of this as being the VC who seems massive dealflow. Half of it is the fun of seeing all the good ideas and picking the top ones. Some will still fail, but choosing from a much larger base of ideas is great. Plus they utilize their own expertise to tweak strategies for themselves.
Its basically YC for algo strategies, except they don't have to pay you, they can steal your startup and change it a bit, and they take a much bigger percentage (i'd guess 50-80%, based on what I think are something like industry standards).
Its quite common place in finance to have people give you strategies and give them a cut of the profits. This takes it to the next level: crowdsources strategies and removes the barrier to entry for a lot of smart people (trading costs) in exchange for the right to tap those ideas.
Regarding the IAMA thread: Small data point, but I've actually chatted with him a few times, and he seems legit. Can't verify his earnings or size he trades, but he seems to know his stuff.
When you buy shares they represent a share in a company. This company may rise in value due to good management, successful products, expanding into new markets, etc. If you sell your share at a higher price, no-one is going to lose the money you will earn, the company is just worth more than when you bought it.
Other securities may make you money in other ways, e.g. if you buy T-bills the FED is paying you interest on the money you effectively loan them.
Relative to the average market performance , making choices as to what to invest in, seems to be a zero-sum game (by definition, as the net total performance of the investments will be average), but I'm not a finance guru, so feel free to correct me.
What I think you are implying is that if you invest in companies part of some market index¹ and end up above the average performance of this index then someone else, who also invested only in companies in this index, will end up below the average.
This however is not zero sum game or has anything to do with markets or investments, it is just how averages work. You could apply the same logic to grades in a school class.
¹ There are lots of market indexes and they only cover a fraction of the investment possibilities and are often trade specific.
I think what you are saying is that if everyone invested very intelligently, then everyone would be better off, as more money would pour into the more viable ventures (is this what you mean?) I can agree with that, FWIW, but from the short-term algorithmic/high-frequency trader 's perspective (edit), this does not appear relevant (some nth order effect).
P.S. Traders often say how their work is useful to everyone, because it makes the markets more efficient. I wonder if it's true. The analogy I'm thinking about is frequent lane swervers on the freeway ("lane arbitrage"). Are they making the freeway more efficient for everyone? I don't think so.
All I was saying is that it is not a zero sum game.
As for invested very intelligently I don’t think you can define that, nor what it means for everyone [to] be better off and more trading of a stock does not mean more money go into the company the stock represent (assuming the stock is not sold directly by the company).
This is all very long-winded to get into here, but as a quick example of how this “intelligently” and “better off” is ambiguous I think we would be better off with more alternative energy, so people should invest in that. But a lot of these investments will likely turn out to be non-profitable (even cause a loss) but on the bright side, a lot of R&D in alternative energy will be done, which is good, and those companies with good projects may end out making a difference in the world, etc.
Why aren’t the prices going to rise higher and higher? There is absolutely nothing that says they shouldn’t (or they should for that matter).
And how does someone lose money if the prices stagnate? The prices have to actually drop for someone to lose money.
Fundamentally this stuff is pretty simple:
I have an idea for a business, I need $1,000,000 in startup capital. I sell 1,000 shares costing $1,000 each.
After a year my business is profitable and valued at $2,000,000. The stock is now worth $2,000. That means everyone who bought my stock now earned $1,000.
Who lost on this? you can say that the consumers who paid my company (to make it worth more) lost, but a) these are not part of the stock market, and b) my product may actually have saved them more than they spent.
Well in the long run, it IS a zero sum game. A stock is valued higher only if someone wants your stock. If no one is interested in buying your stock and you try to sell it, its value will be zero. Also, if everyone holding a certain stock starts selling it, in the end the value will drop down back to zero. So everything earned by someone is lost by someone else. The only real earnings which you get by owning the stock is the dividend which the company pays out.
It is true that if no-one wants the stock then the value of it will drop. This however should (under normal circumstances) only happen if the company is not profitable and is in debt (to the point where its physical assets amount to the same or less than its debt).
It is not a psychological game, i.e. if someone somehow managed to make the Microsoft stock drop to zero then I would effectively be able to get all of Microsoft for nothing, that’s a pretty sweet deal, and why that stock won’t drop to zero as long as they are profitable.
The “psychology” that affects the stock prices are in speculation about the future worth of a company, and that is why you see a disconnect between a company’s net worth and market cap — generally the market cap should be above its net worth, otherwise someone should buy all the shares and liquefy the company :)
Market players who didn't invest in your idea (you could call it "opportunity cost").
If the baseline strategy is "invest randomly", or "invest a little bit in everything", then those who deviated from it by way of not investing in your idea, lost.
It only makes sense to evaluate a market player's performance relative to some baseline strategy. I was trying to make this point elsewhere in this thread.
And how do you define "loss" and "gains" for market players?
PS: Your definition appears to be "you lost iff you left with less money than when you came in". OTOH, my definition is "you lost iff you made less money than if you would have if you'd invested it in an index fund".
If you are talking about the intrinsic value of your company's stock increasing, then it's definitely not a zero sum game. But they do pay out part of their profits (intrinsic value gains) as dividend. If we consider just the stock price fluctuations due to demand/supply (and no changes in the intrinsic value), then it is a zero sum game.
The way I like to think about this is, imagine a hypothetical company, let's call them ROCK. They have a wild business idea of mining asteroids in space. You spend all your money on this stock during their initial public offering and end up owning 1% of it. 40 years in the future, ROCK has had hundreds of successful mining missions and tremendous wealth was derived from the metals they have salvaged.
Through their efforts there are now thriving mining colonies living permanently in space, and a whole industry was created around their operations. Without the initial funds they received from their IPO, none of this would have been possible. As the years pass, they expand further and further into space, with no limits in sight. Another 10 years later you decide to cash in your 1% share and find yourself the proud owner of your own small moon.
Probably 10 - 50 million. You will need 5 - 10 hard core programmers, a network engineer, a trader or three, a bunch of the fastest servers around, etc. etc. This is an insane idea, I would not recommend it! (The margins are being pushed to zero as more and more companies get into this.)
I guess I meant excluding human labor (imagine that a bunch of programmers, etc. with the necessary skills wanted to do it). Would they need to rent a building next to the NYSE and put the machines there?
I don't do HFT, but with what our company does, I've spoken to a few people who do. So take all of this with a pinch of salt, but as I understand it, your execution and autonomous trading systems live in the same rack as your exchange/dark-pool endpoints. Your slower strategies can be offsite, as is the Big Red Button to close down your positions. Still, faster's invariably going to be better, all else being equal...
>The capital-gains tax is a reason not to get into short-term trading
In the same way taxes are a reason to not get into money making ventures.
Short term trading encompasses a lot of ideas. Some are good some aren't. Many "short term" trading strategies have significantly higher sharpe ratio's than Fool style investing. Fool investing is great. I learned a LOT from that site. But I also want returns in shorter time spans, so I'm willing to pay that capital gains tax.
Motley fool is a GREAT source for long term investing knowledge for anyone interested in that. A lot of other great classic investment texts are getting thrown around in this comment thread (Intelligent Investor, for one).
Its sort of in the same vein, but slightly shorter term, backed up by a lot of statistics. Basic principle: Choose the best 20 companies every year. Define "best" as top PE or some other really simple metric. You can tweak it a little as well based on your personal risk profile, desired industry exposure. Consistently outperforms the market.
The motleyfool is a site that looks amazingly like spam and reminds you of all the "earn 50 bazillion percent every hour using options!" emails you get, but is actually chock full of great information.
Greenblatt's book sounds shady as hell, but its a quick, easy read and quite informative, backed by just enough logic to make you want to try it.
it's not a zero sum game. The definition of zero sum games is "A situation in which one's gains result only from another's equivalent losses.". There are many aspects of a stock market that make it NOT sum to zero: the fact that things can go up or down in value with no one buying/selling (aka GAPS) itself makes it a positive/negative sum game. Not to mention taxes, broker costs, IPOs and all other 'game moves' that inject and/or take money out of the 'game'.
An interesting note on the zero sum discussion, while the stock market clearly is not a zero sum game options on the other hand ARE zero sum.
On the exercise date the payoff for the buyer of an option that's in the money will exactly match the loss of the seller. Vice versa if it's out of the money. Kind of neat.
It's a zero sum game, but one might be smarter than others. Some companies doing algorithmic trading consistently skim off their share of the market's inefficiencies.
What I'm wondering is how much of an investment it would be to get into high-frequency trading for yourself. Any idea?