This is also good. I went with Amibroker because you own the software. WealthLab went to a subscription model a few years ago. Both have free trials and community of developers that can help with problems. https://www.wealth-lab.com/
I use WealthLab. However I have an older PC based version with plug-ins to pull financials, etc. that I am still using on a desktop in the corner; primarily I screen for trade candidates based on EOD data using a combination technical and fundamental analysis. Note that I do not use WealthLab for creating an automated system for trading stocks. The best systems I have seen which are profitable for stocks are usually based on a quick deep dip in the stock price and going long on the recovery. There is a user "Glitch" on the WealthLab site who has outlined multiple variations of this type of system over time which tend to be profitable; noting however that most generate a limited number of trades and while profitable they are not going to make you wealthy in terms of the size of wins, etc. In the old days, WealthLab allowed people to publish their ChartScripts for others to review on their website and even provided information on tested trading results (at times even holding contests). Most of these automated trading systems were not profitable when slippage, etc. are added in. I do utilize WealthLab to backtest strategies that others propose (most of them vendors pushing their schemes in articles). Usually I find that their proposed automated trading schemes lose money in back-tests over long durations despite their claims of generating endless wealth. Now most of these vendors are pushing magical AI trading -- where you cannot actually discern an actual strategy that can be coded up in ChartScript for backtesting. Shortly I plan to subscribe again to Wealthlab's latest version; I have trialed it in the past. The Premium Annual Plan ($399.95) will most match my needs because I will need the extensions for fundamental data, etc. Now that I am retired... I will have some more cycles to devote to stock trades. Typically my trades are multi-day time-frames (1-8 days) rather than intra-day.
i have to state every program i try is crap for accurate backtesting. they all paint a better picture than can be realistically obtained. for that reason, i always take the parameters in the 75% range and hope it holds up in real life. in fact over the years i have just resorted to using my eye for judgment and trading with real money in a micro account to prove concepts.
1. Overfitting Overfitting occurs when a model is excessively tailored to historical data, capturing noise rather than the underlying trend. This results in a strategy that performs exceptionally well on past data but fails in real-time trading. 2. Look-Ahead Bias Look-ahead bias happens when future data is inadvertently used in the backtest, giving the illusion of better performance. This can occur if the model uses information that wouldn't have been available at the time of trading. 3. Survivorship Bias Survivorship bias arises when only successful entities (e.g., stocks, funds) are included in the backtest, ignoring those that have failed. This can lead to overly optimistic results. 4. Data Snooping Data snooping involves repeatedly testing a strategy on different datasets until a favorable result is found. This increases the likelihood of finding a strategy that works well by chance rather than due to a genuine edge. 5. Ignoring Transaction Costs Many backtests overlook transaction costs such as slippage, commissions, and taxes. Ignoring these costs can inflate the perceived profitability of a strategy. 6. Small Sample Sizes Using a small sample size can lead to unreliable results. A strategy might perform well over a short period but fail over a longer timeframe. 7. Multiple Testing Bias Testing multiple strategies or parameters increases the chance of finding a strategy that appears successful purely by chance. This is known as multiple testing bias. 8. Neglecting Market Impact Backtests often assume that trades can be executed at the desired price without affecting the market. In reality, large trades can move the market, impacting execution prices. 9. Ignoring Psychological Factors Backtesting does not account for the psychological pressures of real-time trading, such as fear, greed, and stress, which can significantly impact trading decisions. 10. Curve Fitting Curve fitting involves adjusting a model to fit historical data too closely, resulting in a lack of robustness when applied to new data. 11. Unrealistic Assumptions Backtests often make unrealistic assumptions about market conditions, liquidity, and execution. These assumptions can lead to overly optimistic results that don't hold up in real markets. 12. Lack of a Written Plan Without a clear, written plan, traders may deviate from their strategy during live trading, leading to inconsistent results. 13. Execution Timing Errors Assuming perfect trade execution can lead to inaccuracies. In real markets, execution delays and slippage can impact performance. 14. Ignoring Currency Risk For international trading, ignoring currency risk can lead to significant discrepancies between backtest results and actual performance. 15. No Walk-Forward Analysis Walk-forward analysis involves testing a strategy on a rolling window of data to ensure it remains robust over time. Without this, a strategy might not adapt well to changing market conditions. 16. Limit Order Execution Issues Platforms often assume that limit orders are executed at the limit price as soon as the market reaches that price. In reality, not all limit orders get filled immediately, especially if the market only touches the price briefly or if there's insufficient liquidity. This can result in an overestimation of the strategy's performance. 17. Ignoring Market Depth Backtesting platforms might not account for market depth, meaning they don’t consider the order book's volume and the available liquidity at each price level. This can lead to inaccurate execution prices, especially for large orders that would impact the market price in real trading. 18. Bar-Based Execution Many platforms execute trades at the open, high, low, or close prices of a bar (candlestick). This doesn’t reflect the true intra-bar price action, which can lead to unrealistic trade executions. For example, a stop-loss might be triggered within the bar, but if the platform only checks prices at the end of the bar, it could miss this event. 19. Simultaneous Orders In backtesting, multiple orders might be executed simultaneously without considering the sequence in which they would occur in real-time trading. This can lead to unrealistic execution prices and order fills. 20. Bid-Ask Spread Backtests often ignore the bid-ask spread, executing trades at the last trade price or at the bar’s open/close price. In reality, you buy at the ask price and sell at the bid price, which can significantly affect the strategy’s profitability. 21. Slippage Platforms might not accurately simulate slippage, the difference between the expected execution price and the actual price. In volatile markets or for large orders, slippage can have a substantial impact on performance. 22. Order Queue and Priority Real markets have order queues where orders are executed based on priority (e.g., first-in, first-out). Backtesting platforms might not simulate this accurately, leading to unrealistic assumptions about order fills. 23. Partial Fills In real trading, large orders might be partially filled at different prices. Backtesting platforms often assume full order execution at a single price, which can distort performance metrics. 24. Order Types Some advanced order types, such as trailing stops, bracket orders, or conditional orders, might not be accurately simulated in backtesting. This can lead to differences between backtested results and real trading performance. 25. Execution Delays Backtesting typically assumes instant execution of orders, but in real trading, there might be delays due to latency, broker processing times, or market conditions. This can affect the timing and price of trade executions. 26. Incorrect Stop-Loss and Take-Profit Triggers Backtests might not accurately trigger stop-loss or take-profit orders if they rely on bar data. In real-time trading, these triggers depend on tick data (actual trades), which can lead to discrepancies. 27. Corporate Actions Platforms might not properly account for corporate actions such as dividends, stock splits, or mergers. These events can impact stock prices and should be factored into backtesting to get accurate results. 28. Commission and Fee Structures Simplified or outdated commission and fee structures can lead to inaccurate performance assessments. It's important to use up-to-date and realistic costs in the backtest.
yes some of it is and some of it's on you the user. why should i spoon feed you the numbers of the items that are directly the fault of the program? you can read. besides you should know already. 23, 24, 26, 21, 20, 16, 13 and probably more all platform issues every platform has these issues and no provisions to correct it.
Which one do you prefer for strategy development? Does quantshare offer any advantages over amibroker in your experience?