Dr. Bandy, Could you please shed some light into the following curiosities: => A search for your claim of founding chair of the computer science program at the University of Nebraska came out empty handed. A general google search did not produce any evidence. A name, alumni, and content search on the homepage of the University of Nebraska did not produce any results at all. Nor did the same search on the Computer Science Department website of said university. Could you please point us to any reference of your claim, preferably a reference as part of the university web. Maybe a year in which you were chairing the computer science department at the university would help. => Same lack of results when checking your claim at the University of Wisconsin (Milwaukee). Nothing, not a single entry. => Regarding Pacific Lutheran University, I found the following: http://issuu.com/plu-archives/docs/1988-1989_catalog_of_pacific_luther/126 I understand that publications before the internet arose may be lacking or insufficient, but I think we can all agree on the fact that the alumni and faculty databases of most universities (certainly of the ones you mentioned) are updated in the way that they reflect attendance and teaching and research of those who did so even before the internet had gone mainstream. At least a quick look at the alumni and faculty databases of mentioned universities showed faculty members going all the way back. Just your name is curiously missing. Could you please point us to references that support your claim?
Hi Volpunter -- The personal attack is unwarranted and I am disengaging from anything further related to it. The book I have written and published is a mathematics book. Read the book. If there is some specific content that you do not understand, we can discuss that. Best,
It is a mathematics book? This is the absolutely most hilarious statement I have ever read about an author commenting on his own book. Your book's content is copied material available in public domain that feeds off the lack of knowledge of beginners wanting to get into the "algorithmic trading space". You should be ashamed of yourself. You are still refusing to clear up your personal record and CV, nowhere is your teaching engagement at 2 universities to be found. I am also done, I have invested enough time to warn newbies to stay far away from snake oil salesmen such as yourself. The fact that you still apply such cheap sales tactics at your honorable age clearly points to a lack of your claimed abilities, people your age are enjoying retirement rather than feeding off newbies' lack of awareness. You have been holding conferences for years, you seem to do contract work, you sell books, and if you still have to do so now I only must wonder... Enough said.
don't mind volpunter, he is our resident angry nutjob, thinks he knows everyone and eveything in the world, even when proved wrong. A total internet nutcase, and I am glad we have him here for our entertainment pleasure
volpunter- I'm curious if you've ever measured the difference in order processing time between the API and FIX. An IB rep did tell me once that FIX orders are processed more quickly, but he didn't put a number on it. I'd imagine it's not an enormous difference but you seem like the type to quantify things when possible, so figured I'd ask.
Any order that is not FIX needs to be translated to FIX at some point. Since all of the execution engines that run the markets speak only FIX (with very few exceptions). By delivering an order already in FIX, it requires one less step to process.
It highly depends on a number factors. Generally FIX is probably faster but only because there are a number APIs there that are based on REST which is extremely slow, so averaging everything out gives FIX the advantage. But I have seen APIs that run circles around some of the common FIX engines out there. But probably the fastest of the bunch I have seen is a highly optimized FIX engine, highly optimized meaning optimized in terms of message parsing. But keep in mind most often one does not have a choice between the two. Most professional ECNs, exchanges, and brokers do not offer APIs but FIX connectivity, only.
that is not true. Most broker APIs have nothing to do with FIX. They ship the message over tcp in serialized binary packets. I can show you how certain deserialization protocols (generally used when APIs are in place) are significantly faster than processing any sort of FIX message. Hence an incoming order from a client can technically be processed faster via API than FIX if done right. The order is then pre-trade risk checked on IBs side before being forwarded in whatever required format to an exchange or other order processors. I would claim nobody on this site knows how, for example, IB processes such messages. Technically a FIX message can be left intact while risk checking the order and subsequently forwarded to the execution venue. That could indeed shave off sufficient time to make a FIX message reach the exchange target faster than an order originating from an API gateway.
The execution engine at the market runs on FIX mate. It doesn't matter if the brokers API is binary, it must translate to FIX in order to execute in the market.