Obviously, there cannot be a single best indicator for every trader on the planet, since everyone has a different tolerance for risk. One person may be perfectly happy with a 2-to-1 win/loss ratio and only 40% winning trades while another person needs a majority of trades to be winners, whatever the win/loss ratio over time. Those who need more winning trades than losing trades are almost always male, by the way.
Another problem with the question is that if you put ten traders in a room with one indicator and one security, you will get ten different outcomes at the end of a day. By the end of a week, the outcomes will diverge even further. This is because indicators only indicate. Traders apply a great deal of judgment in the exact timing of indicator-based trading decisions, and at least some percentage of the ten participants will decline to take some trades the indicator indicates. In other words, they will cherry-pick which signals to take.
Third, depending on a single indicator is foolish. No indicator is right every time. The best way to use indicators is in multiples with the confirmation approach. When you have a reversal candlestick plus a break of support/resistance plus a buy/sell signal from an arithmetic-based indicator — now you are cooking with gas. Any one of them could be wrong, but the probability of all of them being wrong simultaneously is pretty low.
Fourth, we always assert that Forex is fractal, meaning that if a chart is not labeled, you cannot tell whether it is a 10-minute chart or a daily chart. The patterns looks the same and therefore, indicators should work the same. This is largely but not completely true. For example, the close of the day is always the most important number on a chart, and the close of a 10-minute period or a one-hour period or even the popular four-hour period cannot compete with it. Therefore, indicators that lean heavily on a component involving the close, say the relationship of the high or low to the close, will, in fact, work better on a daily bar than a 15-minute bar. Everybody knows the close of the day but opinion will differ as to the “close” of a one-hour bar.
And finally, arithmetic-based indicators are “objective” in the sense that everyone will obtain the same indicator value from the same formula, but what the indicator is measuring is not a physical object. The engineer measuring water flow through a pipe of a certain size can be exact as to water units per unit of time, but securities prices are not water flowing through a pipe. To borrow measurements from science does not make technical analysis “scientific.” What is being measured is human sentiment, and sentiment does not behave like water or other physical objects. Note that many of the top indicators were designed by engineers — Welles Wilder, Tushar Chande, George Lane — and one of the reasons they persevered in developing their indicators is precisely because prices refused to behave like objects.
Having given all those qualifications, we can still name the top indicators. Everyone will order their list differently, but any list of top technical indicators for Forex would have to include:
When you see charts in books and on websites, chance are that at the bottom of the price chart will be at least one of these, and usually two. (If we were speaking of trading equities, we would add volume indicators, which can be tremendously useful, but as noted elsewhere, we do not have data on volume in Forex.)
Note that the popularity of indicators waxes and wanes. ADX could not be in favor in 2014, probably because it takes some time to interpret, while MACD was hot in 1990 and the stochastic oscillator was the only thing anyone could talk about in 2000. RSI has never really gone out of style, and Ichimoku Kinkō Hyō, a special kind of candlestick chart with moving averages shifted forward and backward, started being the hot new thing around 2008 and is still going strong.
We have found over the past 35 years that on the daily chart basis, the most consistently reliable Forex indicator is MACD. But MACD, despite being a momentum indicator and therefore “leading,” is also wrong sometimes, or indeterminate, and in any case, may not be best for you in your timeframe.
Right away, you can see that if you resolve to use all the top indicators, you will have a very messy chart and also conflicting signals. You have to choose only a handful of these. If you are going to use all of them and also add moving averages and channels, and bands, you cannot get everything on one chart and have to switch to multiple charts. If you are also using multiple timeframes for confirmation, the job just got exponentially bigger. You could easily end up with 12 charts (four charts with two indicators each times three timeframes), and that is not counting candlestick and pattern analysis, including (possibly) a point-and-figure or Renko chart format.
You will always see promotions, online and in your mailbox, for some indicator or another that cannot fail, has generated some absurd amount of profit in a very short time, and is the only indicator you will ever need. It is always fun to follow up on these to find out what the indicator can possibly be. It is usually RSI or MACD, by the way.
The best indicator is therefore personal. The only way to find your best indicator is to back-test a few indicators, separately and together, under different market conditions over some minimum period of time. Statisticians will tell you that you need a minimum of 30 observations to start making any deductions. You do not need software to conduct a reasonable backtest (and in fact, owning backtesting software can become a trap in which you become enmeshed, spending far too many hours finding the ideal parameter for an indicator). All you need is a piece of paper and a chart with the indicator on it. Write down the buy/sell outcomes, noting when the indicator gives a false signal. An easy way to check your work is to draw a vertical line at every decision point. An honest accounting will reveal whether this indicator is for you.
To be really thorough, if a hand-done backtest is not hard enough, you need to add your stops and targets to the indicator. Otherwise you are using a stop-and-reverse concept in which the indicator signal change dictates the exit from one position and at the same time, entry into the opposite position. But you do not have to take every trade your indicator generates if you can see that your stop will easily get hit or the target is not worth seeking.
Previous lessonTopic 12 - Leading vs. Lagging Indicators
Advanced Chart Analysis