I’m thinking about doing a series of posts where I address what I feel are the most common misconceptions among new (and even established) traders. I see these again and again when reading books and blogs. I feel kinda funny using that angle, since I’m not a wise (or filthy rich) trading guru.
But, I do have strong opinions, and a website… so… here’s the first article :-)
If I’ve talked with you recently about a topic, and then it shows up in this series, don’t think I’m singling you out! I’m only going to talk about things I’ve noticed all over the place. And, it’s just my opinion anyway… I’m not an investment advisor, or anything…
Misconception 1: Assuming a Moving Average Is Support
How many blog posts have you seen out there that matter-of-factly mention “and it had support from the 8MA” (or 5MA, or whatever)? A lot, that’s how many. But, when I go look at the charts, at least half the time all I see is a candle near a moving average. Since when is nearby equivalent to providing support? In fact, a lot of the posts that see fit to mention MA support are talking about failed trades, where the price fell right through their “support” level.
Math Doesn’t Hold Up a Stock. People Do.
First of all, think hard about what a moving average is; it’s just a calculation, usually only on the closing prices of the candles! It’s a lagging approximation of the price, which is purely an abstract concept. As such, it is incapable of providing support to anything. People support stocks (ok, and increasingly, machines too). The only reason moving averages work at all is because enough people are using them as guidelines.
What should that tell you? What it tells me, is that I don’t believe a moving average is support unless I see traders buying it up every time the price gets close to it. That’s a pattern that takes time to develop, and I think that usually means it’s a strategy more applicable to daily and weekly charts. Most stocks are just too fickle to trend along a single intraday moving average for long.
Here’s an example from the last two days of XOM. I might look for support near the 21EMA towards the end of the day on Tuesday. That touch around 2PM would represent the third bounce from the 21EMA area. But, I wouldn’t use the 21EMA for anything on Wednesday… the stock is totally disregarding the MA line…
People support stocks. That also tells me that there’s no reason to tie yourself down to any “magic” moving average period. If a stock seems to be following the 24MA, then use it, because apparently the people holding up the stock are, too. Only using the 20MA is like only drawing trendlines angled at 15 degrees…
That’s not to say that being near a given moving average is useless information. A stock that pops up and falls to its 5MA has slowed down until it’s able to touch its smoothed price from 2.5 candles ago. That tells you that it’s not running at top speed anymore. A stock that keeps tapping its 8MA is oscillating around its price from 4 candles ago. And so on… (the same logic goes for EMAs too, which also lag by 1/2 their period).
It’s A Support Area, Not a Support Point
A variation on the “magic” moving average mindset, is the assumption that a stock will stop dead and turn around right at the average price. The fact is, when you look at enough stock charts, it’s apparent that the price relationship to a moving average is usually rather sloppy even when it’s working. And yet, a lot of traders will place undue importance on a single point, like 15.73, if that’s what their average du jour happens to be right now. It’s no surprise these traders are often shaken out of their trades, even though they often turn out to be right about the eventual move.
A lot of people try to address this by making rules like “the candle has to close below the average before I’ll take action.” Maybe that’s an okay idea… I don’t personally go for it. I generally try to avoid putting special importance on the closing value of an intraday candle. After all, it’s not like trading stops between that candle and the next one. If you shifted the clock by five minutes, all of your 15min candles would close at different prices. Would all those candles suddenly mean something different? I would say no, aside from the influence of all the traders out there that do put importance on the candle shapes.
If the stock has been following the MA for a long time, you can try to gauge how sloppy the relationship is. Then, you could use that information to tune your entry, or at least make sure your stop is wide enough to keep you in the trade. More and more, though, I’m sticking to the same principle: people support stocks. So, I prefer to watch the T&S information, and try to identify apparent steady bids somewhere around the level I’m expecting support to materialize. Of course I can be fooled by the games they play, and am fooled all the time. It’s not an easy game, no matter how you try to play it!
Consider It a Signal Processing Problem
Maybe at some point I’ll do a post about all the interesting ideas floating around from the signal processing realm. Essentially, they describe methods to decompose stock movement into:
- a trend component, which is the overal direction of a stock, and is usually assumed to be either a straight line, or a simple curve
- a noise component, which is semi-random oscillation
From this decomposition, they can make observations about the signal-to-noise ratio of the stock, and attempt to jump onto an uptrend while at the bottom of the noise range. It reminds me that, because of the way many trading signals are built, a lot of us get in at the top of local noise, which puts us immediately into the red until the trend we’re on pulls us out of it (if we’re lucky). Maybe there are better ways?
It’s neat stuff, but obviously not foolproof. If it were, they’d all be filthy rich. But, reading about it really clarified the reality about indicators like moving averages for me.
interesting post…btw, since when are u not filthy rich…the ad revenue alone that your slaves (me, zoomie, prospectus, phileo) are bringing in should have put you on forbes list
i absolutely agree that certain moving averages may work better on certain stocks at a specific time
it seems to me that when there is a clean trend, moving average support works better…the same can be said for TA in general…when its working it appears to be magic
overall, i agree with your assessment that “people” are either supporting a stock or not…in that sense, moving averages and other forms of TA are just a guide
I didn’t know Forbes maintained a list of starving bloggers. It must be huge by now.
I think it’s very important for traders to understand that a stock never moves up because of an NR7, or because it touched some fibonacci retracement. It moves up because people created demand for the stock, possibly because _they_ were looking at those things. Of course everyone would say they understand that, but then when they are trading they disregard it.
LOL
i agree…i find myself guilty of that – especially when its a new tool
i use a 1 period MA, it always shows me the perfect support/resistance levels. price always bounces off it, except when it doesn’t.
[...] to me are roads with lots of potholes, as long as you are prepared they can be smooth sometimes and bumpy others. Reward/Risk everytime [...]
B2C: You should try the -1 period ma. It nails the close every time. Freaky…
B2C and Prospectus: i will add that tomorrow…i think its the missing ingredient…i am also going to incorporate the 1 tick chart with the 1 period ma
john, if you have immense success with the new blend of ingredients — as no doubt you will — then you owe it to us to bring back the Tale.
[...] I traded Cephalon, Inc. (Public, NASDAQ:CEPH) today, as an OR breakout. Due to my poor performance the last couple of weeks, I approached this trade with caution, meaning a later entry, a smaller R-size and conservative profit taking. There weren’t a lot of setups out there today, but later I saw CEPH break above the OR high, along with the magic 5-ema: [...]
I applaud you for recognizing that a shift in time periods changes the shape of candles. Another certain highly viewed traders site, which relies heavily on the shape of candles, refuses to acknowledge that candle shape is can show something completely different if the candles just rolled over at a different fifteen? minute time period.
@yo: thanks. I’m much more comfortable interpreting shapes of daily candles, because at least there is an actual trading break between them. Even then, I think it pays to drill down into them, a bit, to see what really happened.
I do think there is a small effect, where surges of volume do flow into stocks near 30min boundaries. And of course, since everyone nowadays learns the standard candle shapes, you can see the effect of the believers acting on the shapes they see. But, I think these are minor compared to, say, resistance at round-number prices.
[...] at MoveTheMarkets posted an insightful analysis arguing against “magic moving averages,” and the entire post is absolutely worth a read for deeper study. I have included a few [...]
[...] pretty awesome, eh? Nudged right up against 19. Has “support” from the 5EMA (heh…). Anyway, I scalped it for 3R, but it looks like a decent dummy [...]
Nothing guarantees success in stock trading; however, there is a tendency for stocks that uptrend to pullback to certain moving averages before moving up again. There are no guarantees, but a person can increase their probability of success through technical analysis. I personally practice risk management on every single trade.
Not sure who you are talking to here, Rick. The post didn’t say “technical analysis doesn’t work.” My whole business is built on technical analysis. The point of my post was that people assume a-priori that certain MAs will be spport, whereas I believe the better approach is to wait for the MAs to establish that they are support (via evidence that people are taking action near them). Regardless, good luck.