Thursday, July 16, 2009

The Trend is a Lie.

You all saw it coming with my last post... I think that the trend is a lie, or at least not helpful to most traders. So whats the big problem with the trend?

First, the trend is BY DEFINITION LAGGING INFORMATION. For there to be a trend it has to have already occurred before you can define it yes? Then you are chasing lagging information, and typically doing so using more lagging information (i.e. moving averages, momentum oscillators, etc). This compounds an already huge problem even more. Whats the other issue?

The trend requires YOU to define it. You can give 100 traders an identical chart and each one of them will define the trend on the chart differently. Some will use market structure, others MACD on X setting, some using MACD on Y setting, and almost all using some combination of X, Y, and Z moving averages. So now not only are you trying to define lagging information, but you're trying to do it in a manner where your opinion is supposed to be the one the market respects. Guess what? The market could give a shit about you. Cold hard truth I suppose but nevertheless true information.

Would you agree that most traders focus on the 20/50/100/200 ema's because they figure most other traders watch these same moving averages? Which means what they are really trying to do is fix the problem of having them define the trend for themselves (because they recognize the futility of this) and figure out what the market (or at least the greatest majority of players in the market) is using to define the trend. But this isn't the biggest problem.

Then you have those trend traders who do great on the runaway trends but get NAILED as soon as sideways consolidation happens. Most cop out by running to another indicator in an attempt to filter bad trades. Others run to the "define the type of day" defense as if its possible to figure out what type of day (trend or range bound) will follow after the first X amount of time passes in the market. Seriously? I think most of us who have traded a while understand that at any point in time ANYTHING is possible. The worst chop in the morning can turn into a massive trend day, and the best trend day can stop and coil sideways for hours. To suggest that we can accurately define something that BY DEFINITION has to have already occurred to be defined ahead of time is bullshit.

Can anyone tell me what the rarest type of day is in the S&P 500? You win an e-high five if you answered "Straight trend day". Most days are range bound most of the time. We all know in the back of our minds that a straight trend day is an exception rather than the rule. As traders we should really try to identify the rule rather than the exception... but most don't.

But the problems continue when we start to look at the actual mechanics of entering a trade based on lagging information. If the trend must have already occurred (at least partially) to get confirmation to enter the trade (MA crossover, stochastic flip, etc) then price is typically already a pretty far distance away from where it was when the actual trend started. This leaves us entering into the market at an area where price is usually already at another decision point in the market. The difference being that by entering now we are going into battle with a much larger amount of risk (we have to put our stops down where the trend started if we don't want to get stopped out) but whats worse is we are going into battle at a turning point in the market with open risk on our position. The smart traders are able to get in where the trend actually started and go into this same turning point battle with their stops at par. Which would you prefer?

Then there is the simple logistics of what you are asking the market to do for you. Lets say you're a smart and savvy trader that knows they need to achieve 2X their initial risk or better on a trade to make their account grow. Good idea. Problem is that the larger your initial risk, the larger the move you need the market to achieve. So if you are entering with your larger initial risk based off your lagging trend information you have to ask the market to go much further in favor to get a needed outcome. By contrast if you were able to get in where the trend started you could use so much less risk that the market has to achieve relatively small goals to get you out of your trade with 2X the risk or greater. And the truth is it is much more probable for a market to move a smaller distance than a large distance.

And most traders can't figure out why they lose money? Most traders look to identify the trend and get trapped into this vicious catch 22 where the odds literally are stacked against them. The trend may or may not exist but the point traders need to realize is that whether it does or does not exist, it isn't profitable to try and define it.

Still want a moving average on your chart?


E-Mini Player said...

As far as definition of "trend" goes, I've always thought of it to be "a series of higher highs, higher lows or lower highs, lower lows". MACD, Stochastics, etc do not play a role in defining the trend, in my opinion. Of course the indicators can also trend, but the same definition (series of higher highs, higher lows or lower highs, lower lows) would apply to the indicator itself.

I view moving averages as a barometer of strength/weakness in the market; not trend definition. For example, if price is in an up-trend, and is bouncing off the 20 period moving average, that would be an indication of strength and would tell me that the bulls are still in control. Now if price is in an up-trend but breaks below the 20-period moving average; that would indicate waning momentum and possibly a reversal in trend. I find moving averages to be helpful visual aids, but I view them in the context I stated above. I'm not blindly buying/selling above/below a moving average, but the MA provides a good reference point to judge which side is currently in control.

Being aware of the trend is definitely helpful in my trading. At the same time, I realize that the market is very complex and just identifying the trend is not enough to be profitable.

Maybe I'll write a post to counter your "trend is a lie" argument...maybe ;)

Btw, classes for Summer Session 1 end tonight so I'll be looking at your stuff tomorrow or over the weekend. Classes for Summer Session 2 begin on Monday (sigh) :-/

TST said...

my 2 cents regarding MA:

1) without moving AVERAGE, the trader may have problem to define an entry with MINIMUM risk and an exit target with MAXIMUM return.

2) the trend and ranging markets are still there, even without a moving average .

Omer said...

Your post reminds me the precise reason technical analysis works in financial markets, and i.e. everyone has his or her own view of different tools. I trade things statistically and defining a trend using a particular indicator like MA gives me an additional probability.

For example, u go in bar n u saw a man who has just had 2 fights with 2 different guyz who tried to hit on his girlfriend. Now, the "trend" of that person's mood is highly "angry" and if u try to hit on his girlfriend too, derez more probability of u (or anyone who tries dat) of gettin punched than not. Markets are traded by people who in turn trades using emotions/psychology etc. So, the same rules apply in the markets.

I dont say MAs or MACD are wonderful trend predictors..I just say that having a reasonably high probability trend detection tool can give u additional probability while trading..

Just my two cents. :)

Matt said...

Great responses gents... The post was meant purely to stimulate some conversation... and you guys have made some excellent points.