Recently, I’ve been reviewing some traders’ backtesting records and found that many have fallen into pitfalls when adjusting MACD parameters. Actually, the MACD indicator itself is fine; the problem often lies in parameter settings. Using the wrong parameters is like having a good tool but using it in the wrong direction.

Let’s start with the standard MACD parameters, the 12-26-9 combination that everyone uses most often. The fast EMA (12) captures short-term momentum, the slow EMA (26) looks at long-term trends, and the signal line EMA (9) filters out noise. This set of parameters is quite stable and is the default on most trading platforms, making it easy for beginners to get started. But this doesn’t mean it’s the optimal choice for everyone, especially in highly volatile markets like cryptocurrencies, where responses can sometimes be a bit sluggish.

I’ve tested several different MACD parameter adjustments myself. For example, the 5-35-5 setting has much higher sensitivity, allowing for quicker detection of turning points, but it also introduces more noise. I once backtested Bitcoin’s daily data from the first half of 2025. Using 12-26-9, there were 7 clear signals—2 of which resulted in successful upward moves after golden crosses, but 5 failed. When switching to 5-35-5, the number of signals increased to 13, seeming to improve success rate, but many of these were just small fluctuations, and the profits were actually eaten up.

There are also other settings like 8-17-9, which is suitable for short-term forex trading; 19-39-9, used to filter medium- and long-term noise; and 24-52-18 for long-term investors. Sensitivity and stability are always inversely related—you want more signals, but you have to accept more false signals. That’s unavoidable.

A common mistake many make is overfitting—adjusting parameters repeatedly to make backtest results look good, fitting too closely to past data. It’s like looking at the answer key while taking a test; a perfect backtest doesn’t mean much if you blow the live trading. The key to adjusting MACD parameters is to find a combination that fits your trading style and stick with it long-term. Don’t keep changing it every time.

My advice is: if you’re still exploring, start with the default 12-26-9 for a month or two. If you find the response too slow, then consider trying 5-35-5 or 8-17-9. The crucial part is to backtest with historical data and see how this set performs within your trading logic. Make sure there’s no overfitting before applying it to live trading.

Honestly, there’s no absolutely optimal MACD parameter—only the one that best suits your current trading habits. Instead of obsessing over the parameters themselves, focus on understanding market cycles and your trading logic. MACD is just an auxiliary tool. If you want to dive deeper into how different parameters perform, Gate has complete candlestick data and backtesting tools for you to experiment with.