I have a system, now what

[maler]

Quote from sle:

Maybe we are coming from difference perspectives (my background is BD and IB prop desks)? My expererience shows that usually IP is pretty much worthless without the provider of the "I". Once someone is gone, the strategies that he used to run are either dropped right away or slowly decays, unless there is a junior that is ready to step into his shoes (that I've seen numerous times).

The main feature of a good systematic trader is that he knows how to tweak existing strategies and build new ones, not in the invention of the holy grail. The only real time when intellectual property is worth protecting is when you are talking with someone who is doing exactly what you are doing (e.g. when I am talking about volatility strategies with other volatility traders) - you don't want to reveal the underwear of the strategy.

I agree that on an IB prop desk, due to virtually zero trading costs,
cheap financing, access to order flow etc. one can implement "public"
knowlege strategies and make money. You don't have to fear other capital
competing your edge away because for most sell side firms prop trading is
a secondary focus, due to lower risk reward compared to the customer side of the business,
and buy side will never be able to cost compete with you.
I was referring to strategies that can be run on the buy side,
where the only barrier to entry is awareness that there is some money to
be picked up where noone is looking for it yet.

 

[sle]

Quote from maler:
I agree that on an IB prop desk, due to virtually zero trading costs,
cheap financing, access to order flow etc. one can implement "public"
knowlege strategies an make money.

All of that would be nice, but in real life i found that there is virtually no difference between trading at a fund and trading on a prop desk, at least in my product space (I have done/doing both). The only advantage of working for a bank was stability of capital vs allocation/redemption cycle at a fund, while there were numerous disadvantages (politics, to name one).

Maybe, if you are running something super-complex, your detailed IP might be worth something - e.g. if you are doing statistical arbitrage and found a few extra factors, that might make a big difference. Even there, the half-life of these advantages is very short unless it requires massive technological investment - there are so many smart people playing this game.

 

[hilojack]

I used to feel the same way about my IP, went so far as to lawyer up, had the code copyrighted, threatened enforcement of BS NDAs, had a guy fired, had a guy black-balled.

To what end? Cost me a lot of time and aggravation, not to mention money (never let a lawyer get into email wars with those who have wronged you at $450 an hour).

The truth of the matter is that anyone who is going to steal your stuff is probably too untalented to make it work properly anyway, and if your edge is that slight that what you're doing is no longer going to work if someone else were to piggyback, then you have bigger problems than someone stealing it in the first place.

That said, there are many different components to a strategy most importantly risk management namely when to get out. If you have a robust system, then multiple parameters should work (just with varying degrees of success).

In addition, if your program is making money, why would they get rid of you? Having 50-80% of something that works is a much better trade for them than having 100% of something that may not.

Lastly, you can take steps to protect yourself. Obviously never discuss with anyone the specifics of your code, and you can also add into your code a random order entry code that will generate "false positives", that is trades unrelated to the strategy. Exit those positions manually. If someone steals your code, they will have problems trying to decipher the good from the bad.

Do yourself a favor and get into the game, don't sit on the sidelines worrying about what might never be. Hopefully in a few years you'll look back on your own paranoia and laugh, rather than wishing you had taken a shot.

 

[heech]

I fully agree with what hilojack said above.... it's been largely my experience as well (minus expensive investment in attorney fees). For a rational investor, paying you 30% of gains is very worthwhile.... versus blindly reverse engineering your system.

But can I say, I love the idea of false entries? Brilliant! I also have some randomized parameters in my system for a similar reason.

 

[SeventhCereal]

in my experience code is damn near worthless unless you are going to take it to another company. You need like an entire team of developers just to understand it and the technology costs to run it could easily bankrupt a hollywood celebrity in a few years.

 

[Longbow]

Quote from sle:

Actually, most people in the industry know what RenTech is doing, plenty of people came and gone. I will note that there is a big difference between "public" and "published" - plenty of stuff out there is public but have not been published.

Everyone is doing more or less the same thing - the strategies are fairly apparent and it's the details that make a difference.

sle, what would be these strategies? I would be very grateful if you could point me to some relevant infos (particularly with respect to what RenTec is actually doing) if there are any?

thanks

 

[Gammahedge1]

This may be a silly question but have you ever place a live trade with your system? Why not take some vacation time and trade it live for a week or two just to get a feel for it? I'm not sure if your strategy is based on options or equities but handling real time slippage and getting your face ripped off from exploding vol can be quite a sobering experience. I primarily trade indexes (mostly RUT) but every once in awhile on huge vol spikes I'll put on a way OTM strangle in AAPL/NFLX/PCLN and all I can say is when these fuckers MOVE you better not be getting your chai macchiato because getting OUT when spreads get wider than a jumbo jet your capital will evaporate faster than a gunfight in a phone booth

 

[gmst]

Quote from Longbow:

sle, what would be these strategies? I would be very grateful if you could point me to some relevant infos (particularly with respect to what RenTec is actually doing) if there are any?

thanks

+1 to this respect.

Sle, I was actually very surprised to hear this because I never heard anyone say before that anyone outside knows what RenTech is doing. This is primarily because most of employees' money is invested in Medallion and so the attrition rate is chillingly low there.

At least none of my batch mates have a clue as to what RenTech is doing. Only one of them interviewed with them and went as far as to see their offices and just came back with the stories of how they store thousands of servers in their basement!!

I also think there was a thread on this at nuclearfinance but that thread also did not generate any useful information. Basically, no one knew anything

So, if you will be kind enough to post some specifics of what they actually do, I will be thankful. PM will be great - if you don't want to write publicly. Thanks.

 

[DT-waw]

400k drawdown with 700k avg annual return?
Not really the best stats, but i'd suggest keep your current job and work on something much better, so you could trade it with personal money.

Unless you have at least $200k to trade, i don't think trading solely for yourself makes financial sense.

prop shops won't rather have technical and intellectual capability to autotrade your strategy...
and even if they do, be prepared to make at least 10 trades per instrument per day- or they will fire you. and better make money with 4-5 pips spread per trade.

OPM. without 3-year live track record dont even bother.
also, be prepared to manage <1M during your first year.

 

[CT10Gov]

Quote from DT-waw:

400k drawdown with 700k avg annual return?
Not really the best stats....

If a drawdown/return < 1 is not good enough for you, exactly what is?