Well, that depends on which time frame you want to observe, doesn't it? That's exactly why you have the signal line, and the histogram. Read again the wikipedia article, and try to uderstand the terms "velocity" and "acceleration".
Stockcharts does a pretty good job explaining it: http://stockcharts.com/school/doku.php?id=chart_school:technical_indicators:macd-histogram Generally speaking, if MACD crosses UP through its signal line, this would be a bullish signal. If it crosses DOWN through the signal line this would be a bearish signal. Histogram being positive is Bullish, Histogram being negative is Bearish (some people use crossing above or below zero as a signal)... As also noted at Stockcharts, people also look for divergences between the price and MACD movement.
Consider the price action of an index to be the trail on which you ride. The horse you ride is named MACD. It does not matter if you measure the stride of each leg, the number of snorts it makes or how if flips it's tail.........it knows nothing of the trail over the hill..........
Choose two points on the MACD-line and draw a line between them. The angle of this line will tell you wether it is angling up or down. However, this method is prone to bias in how you select which points. So it can be both angling up and down, depending which two or more points you choose! If they are far apart, you maybe filtering out too much information or be too late to the party. Too near eachother, you maybe following noise. Instead: Look at price and how it moves, both short-term, long-term and maybe even medium-term as well. Are you confident it may continue that way undisturbed and that the timing is good (ie. you're not fighting the market direction), you can try paper trading it and learn money- and risk management to be able to handle being wrong. Don't risk money you can't lose.
Xela, this might sound pretty funny but believe it or not the MACD does have a powerfull magical predictive properties, lets say accuerate instead of magical, iam more than happy to show you and prove to you its resilincy, the key with that however is that u cant expect it to work everywhere the key is to know where (which instruments) it works on which constantly changes, although u can change its settings, its easier to change the instruments your trading and maintain original settings because changing its settings can have infinite possible combinations let aside changing the time frames, unless u have a software thats backtesting different macd settings (iam searching to see if there is) MACD signals picked up the dot com bubble and its top and bottom as well as september eleventh and 2008 top as well as the notorious silver bubble in 2011 And before u jump and point out examples of where it DIDNT work i can tell u that in the times it didnt work it didnt hurt u much overall assuming u trading it systematically, not even in brexit which is the worse drop on the charts for several years(see my post above)
You're arguing with things I never said. I didn't say that it doesn't work. I said that the problems Ross has with it stem from the fact that he's trying to use an indicator with absolutely no idea how it works, or what he's trying to display with it, or why (let alone what settings he should be using on it), and that that's because he has an underlying belief that this indicator is going to have some sort of magic predictive properties (i.e. that it will "just work" on standard settings, whenever and wherever he tries to use it - and you and I both know that that isn't true).
So did any number of other oscillator/ma type indicators. They also gave just as many false signals. Play with the parameters and you can find what you believe to be true....
The vast majority of retail "traders" use indicators like the MACD as if they were primitive sculptors, and had just come across a Phillips-head screwdriver. "It works great," they exclaim, "to create bas-relief channels by just giving it a little twist" and they bang out a raised little ridge caused by two of the tip's four facets gouging out parallel grooves, with the other two facets slicing uselessly through the air. "See? Isn't that neat??" Clueless. ...Like Rufus Mrehin, who has no idea about why a MACD calculation might be different between MACD on GM, and MACD on TSLA, or for that matter, MACD on GM on 5min set-up versus 1week set-up. Dude, before you cost *others* their fortune, too, LEARN. PROMISE yourself that you ascribe *real*meaning* to every number a MACD study puts out. (You'll know you're ready to speak out loud, when you figure out that "26,12,3" or whatever, is an absolutely arbitrary set-up, and is only *accidentally* commonplace.)
@Xela, yes i agree, and in that sense we know its isn't true MACD works whereever and whenever, and I think Ross is after something different any way with the angle usage versusu the cross over,,, none the less let us be supportive and productive to both newbies and long time traders as well,,,