Stock Trading Language [S.T.L.] Rui Hu Thomas Lippincott Rekha Duthulur Matt (Yu-Ming) Chang Nikhil Jhawar Birth of S.T.L.       Wanted to develop a language for non-techies who do not know how to program. English language considered to be the easiest way to introduce programming to one and all, but… English with its vast grammar and millions of different permutations and combination of words in its domain would require a great deal of time and effort for any language to parse and make sense of it. Hence, we decided to restrict our easy-to-program language to a particular domain to be able to capture the essence of our idea and still generate a language capable of dealing with nonprogrammers. The domain was chosen to be as the one belonging to the financial industry, more specifically Stock Trading Intended users – day to day retail traders looking to bring some automation into their dealing with the stock market. Language Definition  STL is an interpreted computer language designed to facilitate the rapid prototyping of stock trading strategies. The language is mainly aimed at individuals in the financial industry, with little or no formal background in computer science. Language Highlights     Goal: The goal of the Stock Trading Language is to enable a stock trader to write code to trade in the stock market to automate his buying/selling requirements. Ease of use: The language is user friendly in the sense that traders can write simple buy/sell codes using simple words which they use in their financial dialect. Portability: The language is also very portable as the user written code is converted into java code thus enabling it to be compatible with any machine running the Java Virtual Machine. Automation: The language brings about automation for a trader who would have to otherwise sit in front of the computer for the whole time of the trading session. Implementation – Responsibilities      Tom – Lead Tester, GUI, JFL Interfacing Nikhil – Project Management, Testing, Documentation Rekha – Interpreter, walker, AST Rui – Lexer, Parser Matt – Interpreter, walker, AST Implementation - Architecture The G.U.I. - Input Source Code Lexer Parser Exception Walker Symbol Table Interpreter Stock Source The G.U.I. Output Java Financial Library 1.6.1 Implementation - Schedule Date Completed Task 31 January,2007 First Team Project Meeting 02 February,2007 Stock Trading Language Evolves 07 February,2007 White Paper completed 11 February,2007 Code conventions formed 14 February,2007 Lexer started 18 February,2007 Parser started 21 February,2007 Walker started 01 March,2007 Interpreter over Compiler chosen 05 March,2007 Language Reference Manual completed 20 March,2007 Graphical User Interface started, Lexer/Parser Completed. 25 March,2007 Testing of Java and STL code started April 5,2007 Bugzilla (for error reporting) implemented April 19,2007 Simple demo with TA April 26,2007 Walker finalized April 29,2007 Graphical User Interface completed May 5,2007 Testing of STL functionality completed May 6,2007 Final project completed May 7,2007 Project presented Chose interpreter over compiler Implemented grammar early Examples – Conditional buy $i=money(10000); $j=value("MSFT"); $max=400; if($j<$max){ buy(50,"MSFT"); $k=stocks("MSFT"); print("You currently own "); print($k); print(" stocks of Microsoft"); print("\n"); } else{ print("Stock too costly..."); } $l=stocks("MSFT"); sell(20,"MSFT"); $j=stocks("MSFT"); print("After selling You currently own "); print($j); print(" stocks of Microsoft"); //Set the amount of money you have //Assign currently traded value of Microsoft to a variable //Set acceptable price limit on a stock //If the price is favorable buy 50 stocks of Microsoft else error out //assign the number of stocks of Microsoft currently owned to a variable //sell 20 shares of Microsoft from your current portfolio //reassign the no. of stocks of Microsoft currently owned to a variable Examples - Delta /* Program which allows traders to buy stocks based on their volatility */ $i=money(1000);//Set money for transactions //Choose your portfolio /*Go through the list and buy only stocks which have variance that is acceptable*/ foreach %read(@portfolio){ @portfolio=["IBM","GOOG","YHOO","CECO","MSFT"]; print("Delta Values\n"); print("-------------------------\n"); if(delta(%read)<0.5){ buy(10,%read); print(%read); print(" bought\n"); } //Prints out the delta values of each stock in your portfolio foreach %iterate(@portfolio){ print(%iterate); print(": "); print(delta(%iterate)); print("\n"); } $k = 0.5;//Set acceptable variance print("Stocks bought...\n"); print("-----------------\n"); else{ //Else do not buy print(%read); print(" too volatile to buy\n"); } } Example - Recursion function $gcd($x,$y){ if($y<=0){ return($x); } else{ while($x >= $y){ $x = $x - $y; } return($gcd($y,$x)); } } print($gcd(33,9)); print("\n"); The GUI Clear input screen Run all tests Open code from a file Money you have Portfolio Manager Lessons learnt - general      Starting early does not hurt Set realistic goals in terms of envisioning the project and allocating time for the various components that go into building a project Team mates should get to know each other well for optimal results Good communications an absolute MUST Work together at least once every week if not more Lessons learnt - technical      Divide project components properly based on strength and weaknesses of team members A fixed LRM and Grammar go and long way in stabilizing the project development cycle Realize, what will be more appropriate for the project – an interpreter or a compiler Perform tests on every module as it is developed Scoping is difficult to implement, so allocate enough time to get it working DEMO TIME!