I suspect this is going to be an embarrassing question/s: First, why when I look at a shorter term chart such as just below. The current price of SPY is above the 3, 5, and 10 day SMA. But, when I switch to 5 year chart it appears to only be above the 10 day average? Second, doing a visual inspection of charts, it looks like using a a combination of 3, 5, 10 sma averages and only being long or short when current spy price is above or below all three is a winning approach. I assume I must be naively missing something; what is that? 5 YEAR>
Daily charts and moving averages react quicker cause of them closing each day, whereas monthly don't change but once every 30 days, so if current month has huge increase in one direction, it is going to take monthly charts perhaps a few months to catch up but daily react much quicker. And it would be the same comparing daily with one minute charts. There is much more to trading than just getting in, but does show strong trend when price is above longer moving averages like 21 on daily. There is risk management more important than getting in, what to do after you get in than the getting in, most experienced traders spend 95% of their time learning about risk. You could as easily flip a coin heads buy and tails sell, and do well if you understand risk management and how to manage the after. It is like knowing all the answers before a question comes up.
well, I guess to clarify my second question, am I imagining things that going with trend of the three moving indicators pointing to the same thing is generally a good trade and if so, how do I backtest it if I don't do programming?
On the interactive charts at Yahoo Finance, when you select "3m" the chart is rendered using daily data points, whereas when you select "5y" the chart is rendered using weekly data points. Handle's points stand, otherwise.
ah, that makes sense; thanks! So does that suggest I can't really come up with a trading idea based on moving averages visually by looking at bigger situation? Or only if I slowly go through a chart that shows the daily details? If so, any ideas which charting service [ideally free, but willing to pay if one is clearly better at this] would work for this? And better yet, is there a way to back test moving average ideas if you don't program? thanks for any help!
I am an experienced programmer, although not such an experienced trader! One of the first things I did was code a moving average system, which I backtested. Anyone who has done this will know the results I got. Simple / naive moving average systems in the absence of other signals will not give you consistent positive expectancy. Using them as a filter or in combintion with other things is another matter. From personal experience and from what others have said over and over, focussing primarily on moving averages will not be the best use of your time.
thanks Jtrades. That makes sense, as too simple. Does that apply to using moving averages in combination?
I think the keyword is naive. I wasn't able to make any moving average system work, regardless of how I combined them and applied them to multiple timeframes. It may be that your approach would be more sophisticated and bring you success.
Another way to look at moving averages is from a digital signals processing (DSP) standpoint. MAs are lowpass filters. John Ehlers is the man when it comes to DSP in trading. Look up his two and three pole supersmoothers, and his roofing filter.
Newbies think you can load up a chart, add a moving average(s) and that price gonna react upon it/them or offer working crossovers just like magic. Well, it does not work that way, as most of you already found out. Some even load multiples, not for the crossover, but to lie to themselves, if its not this one, the other, or the other, or the other. Please ! Do yourself a favor, learn to distinguish when price is listening to any particular MA by examining its history and when it's not, and even when it is, those are no guarantees but do listen to the outcome, particularly if its a reaction that ends up failing as those tend to take trend-traders to school; yes the trend can be your friend, BUT! not only to follow it, but to see where trend-traders got fucked. Last but not least use them to determine not where price will rebound but when traders got trapped. I could expand but a chart will help more. For instance here's a Weekly Chart of APPLE COMPUTER with a classic 20 EMA. As you look at its history you can see price has rebounded several times off its upslope, this does not mean 20 ema is special in any way, it just means in the AAPL weekly chart, 20 ema has good history, in other words, it matters, we listen to it. Never listen to a MA when price is paying no attention to it. In other words, when it has no good history. Area 2 and 3 are easy to see, a classic retrace resumption, this means price has been paying attention to the weekly 20 ema. However, by 4 we can notice that eventhough buyers were found it was unable to make a higher high, or even if it did make a higher high, not a pronounced up-swing, this is your first warning. Now check 5, buyers are shaken (stops taken) but then those stops are absorbed and price makes an attempt to continue business as usual, except it did not, your second warning. Typically two things can happen here after such price action, price can range for an unknown period of time as it dances up and down the 20 EMA instead of uptrend resumption while killing any breakout trader looking for a breakdown or a breakout or simply the trend can reverse forcing those with a more loose stop to spit their positions. It isnt hard to see when something is no longer working as it was, that's when the predator, you!, come in to make others pain, your gains. Just a simple example, nothing special, but gives you a start of what to look for and how to think.