In reality a computer does not really see a difference between a file in memory and on disk. Nothing prevents you from loading your application directly in memory from the encrypted file. It is harder to program but nothing all that hard using modern development tools like Microsoft .Net
You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
how to protect trading strategies for a colocated server
- Thread starter trend2009
- Start date
In reality a computer does not really see a difference between a file in memory and on disk. Nothing prevents you from loading your application directly in memory from the encrypted file. It is harder to program but nothing all that hard using modern development tools like Microsoft .Net
Quote from sjfan:
This is the mindset of the "retail trader". The retail guys keep thinking there's some magical "hedge fund strategy" that's uber secret and, if discovered, can make the secret tons of money.
The reality isn't quite so. There are few strategies that are original in any sense of the word. The key is implementation. For example, to implement a large a scale basket trading strategy, you need to get all your ducks from trading logic, to execution algorithms (for millions of dollars worth of blocks), to accounting and settlement, to risk control in a row. This isn't something that can be done in trade station or what have you.
Very few big boy strategies are truely secretive. Even back in the LTCM days, what they were actually were not exactly unknown around the street and most groups did some form of it. They were able to do it bigger, faster, and, eventually, fell harder.
Same applies today.
So if I understand your logic, all methods of trading: strategy, trading system, method of decision making have a maximum theoretical limit in terms of the ability to generate return? I get this from your focus on implementation as being the only way to get superior results. What do you think is the boundary for returns exclusive of implementation....the speed of light if you will that can't be broken with current technology or methodology?
For example if know what I want to do I can implement it with a pone call to my broker, a computer based trading app, collocated dedicated servers at every exchange. Assume itâs the same trade. My speed of execution will vary by 120 seconds perhaps but if I'm trading where the average trade lasts 100 times this length it will make only a minor difference.
Any thoughts :100%, 500%, 1000% return on total account capital?
Again, you are missing the picture. What exactly are you stealing? The core trading logic? I can assure you (I don't know about much James Simon, but i do know quite a bit about DE Shaw) that the logic itself will be overwhelmingly unimpressive. There will be nothing there that anyone who's been in any shop with quant operation can't tell you. There maybe a few clever tricks here or interesting insight there, but there's nothing close to a "holy grail" in it.
What will your mythical IT guy do exactly? he figures out the exact parameter trading parameter and rules. Great. Can he go code it up in tradestation/c#/c++ and go on making millions? Extremely unlikely. Even if he's got capital, it's unlikely he can duplicate the entire operation on his own.
But what happens if he's given a team of capable traders, quants, and developers to implement what he saw? Then he can possibly run a rival operation. But, anyone who's worked at a successful quant/arb oriented fund will know about as much as he does. And guess what, they start rival funds.
So, basically, our mythical IT guy has exactly zero value added (in fact, he's a net value subtraction since what he did was quite illegal and anyone who funds him is tainted).
What will your mythical IT guy do exactly? he figures out the exact parameter trading parameter and rules. Great. Can he go code it up in tradestation/c#/c++ and go on making millions? Extremely unlikely. Even if he's got capital, it's unlikely he can duplicate the entire operation on his own.
But what happens if he's given a team of capable traders, quants, and developers to implement what he saw? Then he can possibly run a rival operation. But, anyone who's worked at a successful quant/arb oriented fund will know about as much as he does. And guess what, they start rival funds.
So, basically, our mythical IT guy has exactly zero value added (in fact, he's a net value subtraction since what he did was quite illegal and anyone who funds him is tainted).
Quote from trend2009:
it could be generally true most of strategies are not innovative. But there are cases some strategies can make tons of money. for example, David shaw and James Simon. If the strategy makes tons of money, it will interest people to steal it because few people can make millions per year. what you mean is that the person who spends time stealing your strategy better invest his time writing his own strategy. it is not the case since those IT guys are still working for the brokers.
That's decidedly not what I'm saying.
Implementation isn't the only way to generate superior returns, but it's a necessity.
My point is that there are no super secret strategies that will make its holders millions just because it's such a clever idea. Almost all strategies used are variations on several well known themes.
Inevitably, the successful ones are the ones who can implement it better, more accurate, do more homework, and have better overall portfolio management strategies.
100% return on capital is not a statement that professionals make; Since returns are roughly proportional to risk[*], a 100% expected return on anything would require that your capital at risk is > 100%, which means you are either levered, or mistaken about your return assumptions.
[*] a LOT of people forget this point - even a lot of professionals (especially asset allocators) - and they paid for it in 2008.
Implementation isn't the only way to generate superior returns, but it's a necessity.
My point is that there are no super secret strategies that will make its holders millions just because it's such a clever idea. Almost all strategies used are variations on several well known themes.
Inevitably, the successful ones are the ones who can implement it better, more accurate, do more homework, and have better overall portfolio management strategies.
100% return on capital is not a statement that professionals make; Since returns are roughly proportional to risk[*], a 100% expected return on anything would require that your capital at risk is > 100%, which means you are either levered, or mistaken about your return assumptions.
[*] a LOT of people forget this point - even a lot of professionals (especially asset allocators) - and they paid for it in 2008.
Quote from Jerry030:
So if I understand your logic, all methods of trading: strategy, trading system, method of decision making have a maximum theoretical limit in terms of the ability to generate return? I get this from your focus on implementation as being the only way to get superior results. What do you think is the boundary for returns exclusive of implementation....the speed of light if you will that can't be broken with current technology or methodology?
For example if know what I want to do I can implement it with a pone call to my broker, a computer based trading app, collocated dedicated servers at every exchange. Assume itâs the same trade. My speed of execution will vary by 120 seconds perhaps but if I'm trading where the average trade lasts 100 times this length it will make only a minor difference.
Any thoughts :100%, 500%, 1000% return on total account capital?
I guess you might be right in terms of trading strategies for big shops since their code lines are millions. but if the code is just thousands of lines, and standalone, the situation may be different.
Quote from sjfan:
That's decidedly not what I'm saying.
Implementation isn't the only way to generate superior returns, but it's a necessity.
My point is that there are no super secret strategies that will make its holders millions just because it's such a clever idea. Almost all strategies used are variations on several well known themes.
Inevitably, the successful ones are the ones who can implement it better, more accurate, do more homework, and have better overall portfolio management strategies.
100% return on capital is not a statement that professionals make; Since returns are roughly proportional to risk[*], a 100% expected return on anything would require that your capital at risk is > 100%, which means you are either levered, or mistaken about your return assumptions.
[*] a LOT of people forget this point - even a lot of professionals (especially asset allocators) - and they paid for it in 2008.
Not a single institutional manager worth his salt will even bother letting you buy him a drink to show him the strategy. The reason is simple: 100% with a 3% draw-down over the last three years is a sign of a strategy with a lot of embedded "iceberg" like risk. The fact that you've only had 3% draw-down means you haven't seen what the actual risks are and (don't take offense to this) that has made you arrogant to it and possibly blind to it.
You don't have to believe in market efficiency (I don't). You just have to believe that not everyone in the market is a complete moron to rule out 100% return with 3% expected ex ante risk to be possible [*].
[*] unless you are a flow business, which you surely are not.
You don't have to believe in market efficiency (I don't). You just have to believe that not everyone in the market is a complete moron to rule out 100% return with 3% expected ex ante risk to be possible [*].
[*] unless you are a flow business, which you surely are not.
Quote from trend2009:
I guess you might be right in terms of trading strategies for big shops since their code lines are millions. but speaking from my own experience, that is, I am an individual trader, managing a capital of a few millions, and my return is more than 100% for the last three years with max drawdown less than 3%, I would say anyone could duplicate my return as long as he has my piece of code, which is only less than 10 thousands of lines of code, and it is a standalone software.
do I have idea that has never been discussed in the public forum? no. But no so many people compete with me because few people have done detailed research on the parameters. Suppose my piece of code can make tens of millions, and it hosted with broker, you think no one would have interests stealing it?
I achieved the return vs drawdown from high frequency trading. It may have risk, but not so large because when the signal comes I will take it, and in 85% of cases i will win, even if I lose, just 1 cent, and the trading is spreaded out to hundreds of stocks with 1000 shares per trade.
Quote from sjfan:
Not a single institutional manager worth his salt will even bother letting you buy him a drink to show him the strategy. The reason is simple: 100% with a 3% draw-down over the last three years is a sign of a strategy with a lot of embedded "iceberg" like risk. The fact that you've only had 3% draw-down means you haven't seen what the actual risks are and (don't take offense to this) that has made you arrogant to it and possibly blind to it.
You don't have to believe in market efficiency (I don't). You just have to believe that not everyone in the market is a complete moron to rule out 100% return with 3% expected ex ante risk to be possible [*].
[*] unless you are a flow business, which you surely are not.
I do not know what you mean flow business, I do trade using order flows in level II.
Quote from trend2009:
I achieved the return vs drawdown from high frequency trading. It may have risk, but not so large because when the signal comes I will take it, and in 85% of cases i will win, even if I lose, just 1 cent, and the trading is spreaded out to hundreds of stocks with 1000 shares per trade.
....annnnnnd I'm right. Iceberg risk profile. You 'think' your downside is 1 cent per trade. That might be true in 99% of the cases where you lose, but the 1% is what will kill you. Ie: system failure, regime shifts, extreme liquidity crunch (or the reverse), algorithmic trading screwing up the order book, etc...
Finally, this is probably not a strategy scalable to any significant amount of money, no? You'll be running into significant slippage at, say, 10x-50x your current account size?
Finally, this is probably not a strategy scalable to any significant amount of money, no? You'll be running into significant slippage at, say, 10x-50x your current account size?
Quote from trend2009:
I achieved the return vs drawdown from high frequency trading. It may have risk, but not so large because when the signal comes I will take it, and in 85% of cases i will win, even if I lose, just 1 cent, and the trading is spreaded out to hundreds of stocks with 1000 shares per trade.