Recently, while reviewing trading records, I noticed an interesting phenomenon: many people get stuck on the same issue when using the MACD indicator—the question of how to set its parameters.

Speaking of MACD, this indicator is really common. It consists of three core components: the fast line, the slow line, and the histogram. Theoretically, it can help us capture trend momentum and also judge market reversals. But I found that most people are using the default 12-26-9 parameters, then start doubting whether they are using it incorrectly.

Actually, the standard setting for MACD parameters is 12-26-9. This combination means the fast EMA (12) reacts to short-term changes, the slow EMA (26) reflects long-term trends, and the signal line EMA (9) is used to generate trading signals. Why is this set so widely used? Because it’s quite stable. EMA (12) looks at recent two weeks’ changes, EMA (26) considers the past month’s momentum, and the difference between the two helps us judge mid-term trends. The EMA (9) filters out short-term noise. More importantly, this set is the default on major trading platforms, creating a kind of market consensus. When key signals appear, they attract a lot of investor attention, which in turn enhances the reference value of the signals.

However, for those of us trading cryptocurrencies, especially short-term traders, 12-26-9 might be too smooth. Highly volatile markets require more sensitive MACD settings to effectively reflect short-cycle trends.

I later studied several common parameter combinations. The 5-35-5 set reacts the fastest, capturing short-term trends quickly, but also produces the most noise, making it suitable for short-term traders or highly volatile markets. The 8-17-9 reacts faster but with more noise, suitable for markets with larger fluctuations like the 1-hour forex chart. Then there’s the standard 12-26-9, which is the most stable, suitable for stock daily charts or 4-hour forex charts. There are also longer cycle options like 19-39-9, which effectively filter out most noise. Lastly, the 24-52-18 reacts the slowest but shows the clearest trend, ideal for long-term investors looking at weekly or monthly charts.

The core logic is that the more sensitive the MACD parameters, the faster they can catch trends, but they also generate more noise, and signals may fail immediately. Conversely, less sensitive settings are more reliable for trend detection, with less noise, but signals occur less frequently.

I conducted an experiment myself, comparing the daily data of Bitcoin in the first half of 2025 using the 12-26-9 and 5-35-5 MACD parameters. The 12-26-9 produced 7 clear signals over half a year, with 2 successful golden crosses leading to subsequent rises, and 5 failing. The 5-35-5 generated 13 signals, with 5 followed by noticeable price increases or drops, and the rest failing. It’s clear that the more sensitive 5-35-5 produced more signals, but with a higher proportion of small gains and losses. Interestingly, at the breakout point on April 10, both sets accurately caught the move, with the difference that the 5-35-5’s death cross appeared earlier, resulting in slightly worse profits.

Many people, after adjusting MACD parameters and seeing some good results, become obsessed with finding the “optimal” parameters. Honestly, this is a misconception. Different markets and timeframes vary greatly, and a single set of parameters is hard to perform perfectly across all markets. Moreover, backtesting can easily lead to overfitting—tweaking parameters to fit past data, which might look great on paper but could perform poorly in real trading, like answering a test with the answer key.

My advice is to pick one set of MACD parameters for long-term observation and avoid changing it frequently. Beginners should start with the default 12-26-9. For short-term trading, try 5-35-5 or 8-17-9, but always backtest with historical data first to confirm it fits your trading strategy before going live. If you notice a particular set performing poorly recently, consider adjusting it. Some traders also use two sets of MACD simultaneously to filter noise, which can be helpful, but it increases signals and tests your decision-making skills.

Ultimately, MACD is just a tool. No matter how well you tune its parameters, there’s no absolute best setting. The key is to find what suits your trading style and stick with it, while maintaining regular review and backtesting to avoid overfitting. Only then can you truly integrate MACD into your trading system.