Technical Analysis – Any use after all?

[Originally published June 2007. Obviously, ha ha, oil’s been a LOT more interesting since then. However, the same principles continue to apply.]

So how good is my trader-fu?
In Mar 2006, I showed a couple of charts and explained a little about Technical Analysis (TA).

At the time, the charts looked like this, and I predicted (and I quote):

Oil has been dropping, a little, but it looks like it’s about to explode in one direction, and probably upwards

I got to wondering about this the other day – was I right? Well, here’s what the charts look like now:

note that the previous charted ended at the start of the coloured rectangle

I’ve put the same indicators on there as last time. To re-summarise:

Top chart, red wiggly lines: Bollinger Bands – when they cramp together, the market is humming and harring to itself – they expand out when it takes off (in either direction)

Bottom chart, histogram+lines at bottom: MACD – magenta is general trend (below zero=down, above=up), yellow shows change in that trend, green dotted is a smoothed version of the change in trend

(that’s the Cliff notes version, anyway).

Ok, so, let’s see how we went. The left hand of the coloured rectangle is where I did the analysis – right at the end of March, 2006.

In the next month, oil went racing up to (almost) all time highs, hitting 94.60 on the 21st April. Excepting the one point where the market flicked to around 97, in Aug 2005, this is the highest the market has EVER been. So, that was a pretty good estimate – and bang on time too.

Of course, things got a LOT more interesting after that.

The next big peak
See that extra peak, just to the right, in July 2006 (rightish side of the box)? Wow, I bet a few people dropped a stack of cash on that one. That’s almost the highest oil has EVER been, 95.49 (again, only beaten by Aug 2005 – which if you remember was when hurricane Katrina took out most of the major oil refineries for the entire USA).

The really interesting thing was the day after that peak, namely 17 Jul 2006. The market high was 94.54. Hang on a minute – within 6/100 of a dollar of the peak we predicted in April? What are the odds of that? Well, that’s the thing with traders, they’re damn superstitious. So, when the high the previous day wasn’t sustained, people would be looking to see if the market stayed above the previous (April) high. It didn’t – which often bodes very, very bad things. It means that, essentially, people just don’t have that much faith in the market – not enough to put their money where their mouth is, at least.

This pattern (where the market creates a top, then goes up to it and nudges it again) is called a double top. Usually the market will drop away as much as it climbed up to the left of the first top (ie, at least down to 80, maybe 75), in about the same amount of time.

So, anyway a lot of times, you get a massive, massive drop. Maybe not overnight, but soon, real soon. People start to panic.

See how in August (two bars from right hand edge of the box), it climbed up again, but this time it couldn’t even muster the energy to nudge the previous high? Nope, people are getting nervous. Time to bring your emergency pants onto the trading floor.

End of the rectangle – how many pairs of pants will I need?
So, at the end of the rectangle is a really good time to look at the indicators again. The market has dropped, but not too far – still in the previous trading range (ie, above 85). Is it going to plummet, or still keep kicking around? What do the indicators tell us?

The bollinger bands (red, at the top) have been cramped in tight for some time, about three months. This means where-ever the market is going to go, it’s going to explode, and a LOT. See how usually they kinda pick a direction and go there, up, down, up, down, always somewhere? Well, for three months they’ve been in going horizontal. This is also called “range trading” (coz, you know, they’re trading in a range. Traders aren’t a super imaginative lot. You may have noticed).

Ok, so bollinger says something is well overdue. What does MACD tell us?

Well, here’s where you have to be careful. Pink was above, a little, but has just crossed down. Does this mean a drop? Well, maybe – but notice the same rough pattern happened in June with no major market drop. So, it could be a false alarm.

How about the yellow line? Well, this is a bit different. Remember this tells us how fast the trend is changing. Usually you can compare it to the green line for further sensitivity. See where pinko dropped in June, how ol’ yella nudged against green dotty, but didn’t cross it in any major way? Well, have a look at it now. Yellow is below green, and dropping fast, REALLY fast. You can also see that in a bar or two, the yellow line crosses the zero – further indicating how fast the market is moving. If you look at the picture, everytime this sort of thing happened, there was a massive move. The last time was in Oct 2005, and the market dropped 15 big points.

Yeah, yeah, yeah, but what does this all MEAN?
Well, first let’s talk about 15 big points. What does that mean? Well, normally when people talk “points”, at least in this market, they’re actually talking about movements of 0.01. Why? Because each 0.01 is worth $10 per contract. It’s pretty usual to trade anywhere from 5 to 50 contracts at a time. Since 3-400,000 contracts a day get traded, you know there are people in there trading very, very big numbers. A thousand contracts, several thousand, big money.

So, each point move of 0.01 is worth $10. Which means that a full decimal point, a big point (ie, 100 ‘points’ = 1 ‘big point’) is worth $1000 per contract.

Now think back to Oct 2005. That means if you held just a single contract, you would have made, or lost (if you were long) $15,000, in two months. That’s US Dollars.

Now have another look at the drop that started in July last year. It went from 95 down to a low of around 60 in Jan 2007. That’s 35 big points. $35,000 a contract. Even if you got in at the right hand side of the rectangle, in August, when the market was around 86, if you’d held waiting for a definite cross in the MACD, which happened in about Feb (at 66), that would be $20,000 in profit. Per contract.

And even that little move I predicted last year? From around 84->94? That’s a cool $10k per contract, my friends – just for paying a little attention on livejournal. Who ever thought this place would pay off in numbers like that?

So you wanna trade futures, eh kid?
Of course, before you rush off and get a futures account, have another look at the MACD, and the Bollinger Bands. These are just two indicators. TA has thousands of the damn things. They jump around like madmen and give crazy signals left right and centre. This is not a game for people with weak stomachs. 90% of people trading lose money (the other 10% do very, very well, of course). Personally, I’ve watched $50k of my own money disappear in about half an hour. This is not a story that raises any eyebrows when I tell other traders – it’s pretty typical. It’s also pretty typical that any trader will biff away USD200k or more in “education” before they start to play a steady game. It’s definitely not for everyone. It’s not even for well adjusted people, it could quite reasonably be argued.

I heard a story, years ago, that if you wanted to be a futures trader, you need to do the following: Wait for a really windy day. Gale force was best. Go down to the bank, and pull out $10,000 in high denomination notes. Walk to the centre of the town square, and throw the whole lot in the air. Watch it all blow off down the street, then go home and sleep well, really well that night. Think to yourself “Ok, that was stupid. I won’t do that again”, and nothing more. If you can do that? Then this might be the game for you.

I used to think it was just a nutty story. Then I traded a whole bunch. Now I think the number is probably a bit low. Try at least $20k, that’s what I’d recommend.

If, however, you think you can handle that kind of pain, and sleep well the following night, then hey, come on in! The water’s lovely!


  • No Related Posts

June 30th, 2008 | Technical-Analysis |