I am not necesasrily saying small account, but below 2 million I would consider the cost prohibitive. You dont want to blow half a year or a year or two years of profit for that, or?
Hired Talent: What are the Costs and Risks?
- Thread starter achilles28
- Start date
I am not necesasrily saying small account, but below 2 million I would consider the cost prohibitive. You dont want to blow half a year or a year or two years of profit for that, or?
Quote from Runningbear:
How times have changed. I remember a time when an average hold time of 2 minutes was considered high frequency. Now its considered trend following
lol If the shoe fits... I'd rather automate the scalping and leave the swing trading for me.
just sent a pm...
Not a legal solution ..
One of the benefits of object oriented languages is encapsulation and data hiding. You should strive to separate your strategy logic into distinct functions. You can then distribute these functions to the programmers for coding. No programmer needs to see the entire system, just the functions that they are coding. Since no progammer sees the entire system, your secret is safe.
You then assemble these pieces into the strategy.
I don't know the specifics of your situation.
You have some legal and technical protections (ex. use several brokers, have programmers work on modules and not on the whole system etc.). We do all of them here.
But the most important thing is that your employee's benefits are aligned with yours. If they have an incentive to steal they will find a way.
Here people understand that strategies are not our most valuable assets. Even if someone can steal one it would proably not be worth it (financially speaking) to set it up elsewhere and trade it for himself. He wouldn't have a big enough account.
Ninna
That is very nice said - but it totally ignores the complexities. You need someone VERY knowledgeable making the planning and integration tests on something that size. We talk of a clsutered system here (uptime requirement). YOu need a good architecture.
Yes, you can hide the total for the parts, but the parts do not magically compose themselves according to the requirements.
AND you will PAY for this separation. Without automated integration tests etc. it will take a lot longer to develop every part.
If I were you I would hire a company not and individual (preferably a company as large as possible), so they have a bigger incentive to honour their nondisclosure and not compete contracts.
You mention that you manually scalp 2 instruments (~30 trades/hour/instrument). What makes you think your strategy is profitable when traded automatically/on other instruments. Are you sure you don't use any discretion/feel when you are putting on your trades? I guess you know what you are doing since you want to hire someone, but I would only do that if you are absolutely sure your strategy has objective positive expectation without human intervention. In my opinion, it would be more beneficial to you to learn programming on your own. Most people find that easier than finding a profitable strategy...
Yes, I tend to agree with Jelite. From your earlier post it seems you have at least some exposure to programming, regardless of the language you will still have experience structuring algorithms.
IMO, some of the discussion is overstated when considering infrastructure. I have worked with many of the large financial houses over the years, so do agree the sizeable running costs, but allot of this relates to the size of the operation.
Sure, by definition your risk and pricing (especially the more complex non-linear stuff) requires serious hardware, but these people are pricing and managing a significant quantities of products and volume. External pricing grids churning monte-carlo to complete within a given time-frame don't come cheap.
That said. You may be able to write a strategy with sufficient error handling and messaging running on an external vps or server-farm for a relatively small cost. Most ECN's or brokers with decent API's for strat running, will give you all the back-office you need and quant-libs are easy to come by for system integration.
IMO, some of the discussion is overstated when considering infrastructure. I have worked with many of the large financial houses over the years, so do agree the sizeable running costs, but allot of this relates to the size of the operation.
Sure, by definition your risk and pricing (especially the more complex non-linear stuff) requires serious hardware, but these people are pricing and managing a significant quantities of products and volume. External pricing grids churning monte-carlo to complete within a given time-frame don't come cheap.
That said. You may be able to write a strategy with sufficient error handling and messaging running on an external vps or server-farm for a relatively small cost. Most ECN's or brokers with decent API's for strat running, will give you all the back-office you need and quant-libs are easy to come by for system integration.
"This is a serious thread. If you've got nothing to contribute but hostility, please stay off it.
"Thanks for that detailed, albeit presumptuous reply. "
you should learn how not to personalize. maybe more important than your original request you should read a book on how to make friends if you expect people to cooperate with you.