Amibroker Growing Pains
Data, Databases, and Your Trading Brain
This week, I almost introduced a PC to low Earth orbit.
The plan was simple: take an opening range breakout idea, wire it into Amibroker, and march toward a fully automated intraday system. The reality was symbol lists, data feeds, database crashes, and the realization that my RAM is the real bottleneck in my trading.
Here’s what actually mattered.
1. Cheap data is not cheap
I started with Polygon because it was “only” thirty bucks versus roughly a hundred for IQFeed. After burning a week fighting imports, symbol formats, and massive minute-bar files, it hit me: I had already spent years of IQFeed fees in wasted time.
Lesson one: if you are serious about systematic trading, pay for the cleanest data you can afford up front. A hundred bucks a month is nothing compared to the cost of debugging mystery candles at 11 p.m.
[Graphic: simple checklist of Amibroker setup pitfalls with green check marks: data feed, symbol list, session times, RAM]
2. Your symbol list is not “just a list”
There is no single “US equities list.” Do you want ETFs, ADRs, warrants, OTC, options? IQFeed ships a gigantic master file; Dave has a script to slice it down into something Amibroker can digest without choking.
The key is to let your data feed define the universe, then filter. If your backtest keeps dying, triple check:
You are using the feed’s own symbol list
Database bar count matches your intended history
Session start and end match your actual time zone (my one hour offset literally crashed backtests)
3. RAM is a trading resource
On a small one year database, everything felt fine. Push to five years of one minute data with custom columns and suddenly the program “freezes” or dies. That last phase where Amibroker builds custom columns is a RAM hog.
Right now my practical reality is:
One year: smooth
More years: stair-step it up and see where the machine taps out
Long term: profits from this intraday system are earmarked for a ridiculous, noisy, high RAM tower under my desk
4. Treat it like a game, not a personality test
Discretionary trading often turns into “why am I like this.” System building is different. I have a boss battle right in front of me:
Get clean data in
Make a stable one year database
Replicate a simple strategy
Scale the database as hardware allows
Each step has a clear pass or fail. No vibes needed. Quant beats vibes.
If you want to skip straight to trading with tested edges while you grind through your own Amibroker journey, that is what we built StatsEdgeTrading and StatsEdge Pro for: fully backtested swing, momentum, mean reversion, and day trading systems with 25 plus years of history behind them.
Come grab the stats, the bots, and the Discord alerts at www.statsedgetrading.com.