This is to say THANK YOU to all those wonderful people, including our great support channel, who have replied to many of my questions during the year! Your patience and helping hand has saved my countless hours of work. It allowed me to do some afl programming I could not have managed without you.
AmiBroker has come a long way and as it evolves we are provided with new trading opportunities in system innovations with each new release. From what I see trades are getting shorter. Real-time market scanning, portfolio trading, and auto-trading (or as close as one can get to that) is almost a necessity. We can do all that now with AmiBroker however I hope that data providers and AmiBroker will soon start to offer more specialized/innovative tools in this area. For example, when Scanning 1500 tickers to select the daily tickers to trade, we don't require real-time, quote by quote, backfilled data histories. It would be quite acceptable to scan/explore only the last n-quotes in a longer time-frame, say 3, 10, or 15 minutes. I know this can be done now however the process requires a basic-timeframe backfill, this is terribly slow. Being able to Scan (sample) in larger time-frames, without the need to backfill the basic time-frame, would reduce bandwidth and execution time. I recognize this is not just an AB issue, data providers must also get involved - it would save them lots of bandwidth. Another area where we may see improvements is multi-database environments. Almost all of us use separate development and trading DBs. Many of us need to access EOD as well as Real-time data when trading. Many of us would like to access more than one DB without a reload. Multi-DB setups is impossible because many features are DB-specific. Take for example Watchlists, each time we create a new trading DB we have to manually create many Watchlists. Watchlists (lists in general) have far greater use than just for the traditional categories; they are very convenient as Status-bins. For example, when trading large portfolios it can get quite messy to maintain persistent ticker and order status from day to day. Often it is easier to categorize Ticker/Order status and create a Watchlist for each status condition - after all we can create an unlimited number of WLs. Membership to one of those Watchlists tells you order and ticker status, and does so in persistent form! A new function that would help here is an IsInWatchlist() that can be used in an indicator. Another nice function to have is a new category (and related functions) that enable you to work with lists (files) that are accessible from any DB, i.e., that are non DB-Specific, and that can be Backtested, Scanned or Explored directly from the AA. I wish you good health and enjoyable trading for the coming year, Herman
