Shadow Trader wrote:Thanks gwhitlock64, entries being the bigger leak fits what usually happens when market orders chase a move that is already running.
Shadow Trader wrote:Logging each unfilled limit with what the plan would have produced settles it within a month or two.
Fair point, and tbh I've only been going off gut feel about the "missed moves", which is dodgy maths when I only ever remember the winners that got away. I'll add an "unfilled" tag in the sheet this weekend, and I reckon I'll need to put an expiry on the limits too (a few candles, say), or I'll end up with stale fills that muddy the numbers. Will report back, probably grumpily
Shadow Trader wrote:Thanks gwhitlock64, entries being the bigger leak fits what usually happens when market orders chase a move that is already running.
Shadow Trader wrote:Logging each unfilled limit with what the plan would have produced settles it within a month or two.
Fair point, and tbh I've only been going off gut feel about the "missed moves", which is dodgy maths when I only ever remember the winners that got away. I'll add an "unfilled" tag in the sheet this weekend, and I reckon I'll need to put an expiry on the limits too (a few candles, say), or I'll end up with stale fills that muddy the numbers. Will report back, probably grumpily
Yeah and do you use exponential money management or linear?
What are your favorite trading pairs?
And do you invest into stocks or indeces as well?
Preserve your own money. Scale with the market's money. Exponential growth is the ultimate key.
The language question comes up constantly. The answer depends less on taste than on what the code needs to do.
MQL4 and MQL5 run inside MetaTrader. They're tightly integrated, with direct access to charts, orders and history. Huge libraries of examples exist, and many brokers support them. The language is quirky and limited, but for a strategy living entirely inside the platform, it's the shortest path.
C# in cTrader's automation environment offers a modern language with strong tooling, which suits developers who value structure.
Python is the favourite for research: analysis, backtesting, machine learning, data handling. For live trading, it connects to brokers through APIs or bridges, which adds infrastructure but also flexibility.
A common pattern: research in Python, then port the finished logic to the execution platform's native language, testing that both give the same results on the same data.
Whatever the choice, the hardest part is rarely syntax. It's handling errors, state and the quirks of live markets.
One more consideration is community and support. MQL has the largest pool of freelancers, forums and ready-made code, which lowers the cost of getting help. C# has fewer but often more professional developers. Python has an enormous general community, though fewer people who know the quirks of broker connections.
A practical test is building the same simple strategy, say a moving average cross with a fixed stop and target, in two languages. The effort needed, not the elegance of the result, usually settles the choice.
Which language has been used for trading code, and what drove the decision?
gwhitlock64, both effects you describe are real and they work against each other, which is why the three columns are worth keeping. Partials trade some average R for a smoother curve. Whether that is a good trade depends on what the smoothness buys you: if a calmer curve stops you from abandoning the system in a bad month, giving up some R can be cheap. A useful next column is the maximum drawdown of each version, so you are comparing cost against benefit rather than R alone.