I encountered a similar situation with Fidelity yesterday. All of a sudden my cash available to purchase securities was $0. I called and was told that I traded several positions in cash exceeding funds that settled three times, so I violated the "good faith" rule, which I never heard of. Anyway, the bottom line is that I can't trade with unsettled cash for 90 days, but I can trade on margin. Here's a question for the group. If I open a trade on margin and decide to carry the position for several days or longer, is it possible to "pay off" the margin with available cash? When I trade on margin am I automatically "borrowing" even if I have cash, unsettled or not? I don't intend to day trade either, but I close a position when it starts to go against me and sometimes that's immediately. I hate the idea of paying interest on a margin balance when I have plenty of cash sitting there.
Well I wanted to cut them a little slack... My bias against Java is the amount of logic all used to get it going, and that it runs differently (messages) on different platforms and with different releases. Also in general it creates a lot more overhead than more native approaches. The Visual J++ from Microsoft was a great implementation, but Sun is more political than technical and thus killed it off. Unless it's changed dramatically since I worked there, my expectation of great things from Sun is pretty low. Thus basing the software on this can eventually lead to not quite a wall, but a steep hill at least. I agree that a fresh new company can create good things with the right people and processes. And good/bad software can be created with any language. More a function of the people than the languages/techniques used. One of the best software presentations I witnessed was one of the original Mac designers who wrote his code in "object-oriented assembler". It was great! What I'd really like IB to do is host their existing API from their own servers using TCP/IP. Then any programmer using their API doesn't have to deal with all the Java/release nonsense and they get to keep all the good things of IB w/o the pain of their software decisions. -Jim
No. You dont pay interest in that situation. Watch your leverage value. If lev <= 1 then no margin interest.
I think you are underestimating the complexity of the FIX protocol used to communicate from TWS to the IB servers -- all of which is hidden from us by the current design of the TWS API. Most other brokers that offer a pure FIX connection a) charge an arm and a leg and b) require complex and expensive certification tests whenever your software or their software is updated.
I'm not talking about FIX, yes that looks complicated. What is currrently a localhost connection to TWS is that which would like to see going to IB servers that are either running TWS or other backend. These in turn will talk FIX to the existing IB servers. So the model stays the same, but the TWS is hosted on the IB side and then talks FIX to their servers. They would have to augment their existing API only to provide login details. This change involves only a minor addition to their API but retains everything already programmed for third party software. And if they did this, they could QA TWS PRETTY GOOD trying to run lots of copies locally! But my hunch is they would have to write a true server app that just looked like TWS to socket connections. To me the big upside is then programmers could write all sorts of apps wrapped around just the Socket stream, COM, ActiveX or DLL which talks Sockets. Then TWS is used by people that like it, but is not in the way of other solutions. IB already had a contest already to use the API, this lets them do the same thing where people could write VB, C++ or other apps as front ends to the user. Right now my code has to programmatically manage TWSStart, which then programmatically manages TWS and then sockets connect to TWS and have to be managed. The balancing act changes whenever something changes in TWS/Java/TWSStart. It's a dread when it doesn't work. With proposed solution my logic just manages the socket connection which "appears" to go to TWS. Always too many words to express intent - hope this clarifies. -Jim