I take this completely to heart and hence why I am here trying to get real live input before splashing my money around somewhere. I've been saying that my results MUST be "too good to be true," and at best - they probably are - but could still make SOME money. As for TradeStation - I found that one and liked it - but then I heard that the broker side of the company tends to "delay data" and cause problems with the trades that can screw up a traders price fulfilliment or ability to monitor the market if the data "freezes" up for minutes at a time. Any experience with that or would you say that this software AND broker are reputable? Thanks.
yes a classic question, thank you for asking since some older ETers still don't get that part (guys like Mike805) well what if they trade bigger size and have faster access their bot can screw your bot, there is more but I gott go
for the data - opentick (free) or esignal (not free, thats what i use) then to create bots : if you have time to dedicate to this, in esignal you can follow easy tutorials that will give you enough javascript knowledge to be able to code indicators and scripts to do your backtesting for you automatically. Basic indicators can be coded easily (its just a mathematical formula that will be plotted on the chart) but coding scripts that will do backtesting for you (in a script you assemble the conditions for the entry and exit of the trades you want it to take) can be more complicated and tedious. Once you have a script that generates money, its already 3/4 of a working bot, because it takes and counts the trades and the P&L correctly, you just need to add the commands for the order entry(Buy conditions met, send buy command, etc) which is pretty easy. I use ninja trader for that purpose. The bot runs in esignal, sends commands to the ninja trader platform, which execute the trades through my broker. if you dont have time, you can hire people to do it for you, but youd be handing them your secrets. If your thing works and it generates money, you don't want to do that
Many people (I mean A LOT) go through the system design process and find good results right away. They then trade said system live and soon find that it fails. The reason is usually a simple lack of experience - there are just a lot of things that you don't know yet. It's okay, just remember it takes time to learn what does and doesn't work in the markets. Backtesting can be deceiving, you need to be careful at identifying what exactly you are trying to exploit price-wise. If you're using computing power (i.e. optimizing parameters) you're likely barking up the wrong tree. If you're going the high-frequency route (1 trade per day or more), your competition stiffens even further. I can go on an on, the pitfalls are virtually endless. But, good systems can be found, just make sure to educate yourself as much as possible before using real money. TradeStation is OK. Not that good and not that bad. Their quotes will sometimes lag when the market is moving (post fed meeting), but in general, they offer good service, great historcial data and a good/intuitive programming interface. I use them entirely for quick and dirty backtesting. I've tried OpenQuant, NinjaTrader, MultiCharts... I keep coming back to TradeStation because it is just really easy. Note that I don't use TS for my main trading activity, but, when I have, they did pretty well (I trade large blocks and TS WILL NOT budge much on comissions). Also, their trading forums are great - if you intend to seriously pursue designing and trading a system, log onto TS forums and do a search for a poster named "Nemo" and "Mathemagician". When you read through their posts on systemic trading you will then know what you need to know. Again, software = good, broker = fair, and trading forum = excellent.
Cold, I know what you are talking about and a trader like the OP doesn't have to worry about that for years to come. Also, moving a market is a cost of doing buisness. I know how the bots work (I've designed some of them) and their edge exists in "impulse" order flow, i.e. aggresive taking of liquidity - something this poster is no where near. When one gets to this level the execution services available minimize this cost substantially, i.e. the edge of the bot doesn't beat the spread in most cases. Personally, I'm okay with paying the spread. In terms of brokers doing what you suggest, well, be careful what you write here. The SEC does monitor these boards and you've essentially incriminated yourself/your broker. Mike
first of all my apologies, I thought you were slow and second well I lie a lot, I never really worked for a broker not that I know of anyway
cold, like I said, I don't doubt this shit goes on but it's not something any trader should be worried about. There are bigger obstacles to finding your way than worrying about the broker stealing your method. Heck, if I owned a brokerage firm I'd fade every noob that walked in the door and win 95% of the time. It's more likely the broker will take the other side rather than steal his method. And, studying one's past results won't make you money. There is so much judgment involved in trading that you could never black box a system just by getting somone's triggers. I don't mean any disrespect but this guy has bigger walls to climb before he needs to consider covering his tracks. I don't hide a thing from my broker. If they follow along then more power to them. I don't care.
You have to understand things like impatience, impulsiveness, situations of fear, and so on. Once you begin trading with real money you will know what I'm talking about. Once the market wacks you you'll remember it and it will weigh on you. You will start equating a current setup to the one you got wacked on when you shouldn't. This is a destructive path and something you have to be aware of. EVERYTHING changes when go in with real money and the biggest chance will come from within. You have to get a grip on that and understand and learn how to think and act like a trader. This is step 1.