Wednesday, January 31, 2018

Unmasking the VooDoo - Michael's Rules

This post is a little different than the others where I usually take a deep dive into one bit of technical analysis or another to expose what's really going on beneath the hood. In this post, I'll simply list my rules for dealing with the analysis in the real world and not in some idealized trading lab.

Sorry AI. This is your daddy, RI (real intelligence). This is what you are trying to be when you grow up.

Follow these rules. You may not make big bucks trading but you will up your win rate. Plus you will definitely avoid misinterpreting the market's message and taking bonehead trades.

I've got some comments below the list, as well.

Michael’s Rules

  1. If you cannot see trends and patterns almost instantly when you look at a chart then they are not there. The longer you stare, the more your brain will try to apply order where there is none
  2. If you cannot figure out if something is bullish or bearish after three indicators then move on. The more studies you apply to any chart the more likely one of them will say “something.” That something is probably not correct.
  3. You can torture a chart to say anything you want. Don’t do it.
  4. Be sure you check out one time frame larger than the one in which you are operating (a weekly chart for a swing trader, a monthly chart for a position trader)
  5. Look at both bars (or candles) and close-only line charts to see if they agree.
  6. Patterns must be in proportion to the trends they are attempting to correct or reverse. I like the trend to be at least three times as long as the pattern.
  7. Patterns should have symmetry. A triangle should look like a triangle and not a mile high and an inch wide (or vice versa). A head and shoulders should look like a central peak with two smaller but equal peaks around it.
  8. Price rules but it is better when volume, momentum and structure (patterns) agree. Sentiment is a luxury.
  9. Always confirm one type of analysis with another type. For example, confirm RSI not with MACD but with on-balance volume or relative performance.
  10. Don’t get hung up if all your indicators do not agree. They never will all agree and you will end up missing every opportunity. Therefore, pretend you are a trial lawyer gathering a preponderance of evidence, not guilt beyond a shadow of a doubt.


on point 5 - In other words, get different charting points of view. They don't have to be in lock step but they cannot tell you radically different things. If they do, find another thing to trade.
on point 6 - This is a pet peeve. You cannot count the trend as part of the pattern. The pattern must be a separate chart entity.  Also, cup-with-handle patterns are continuations, bottoming reversals. Call the latter something else, please!
on point 9 - RSI and MACD are essentially the same thing. Look at something that covers volume. Look at something that has some sort of sentiment component. Look at something with a time component.

Friday, November 17, 2017

Unmasking the VooDoo: Trends

When discussing technical analysis and trends, the erudite snoots among us, which unfortunately can include me, like to talk about physics. Objects, or trends, in motion will stay in motion unless acted upon by an outside force. That's inertia and yes, markets have it. How else do we explain momentum stocks, let alone bubbles?

As the old Faberge Organics shampoo (with wheat germ oil and honey) commercial used to say, "If you tell two friends, then they'll tell two friends, and so on and so on and so on." The same goes for the markets. It's the ultimate expression of keeping up withe the Joneses. Everyone covets everyone else's profit-making assets

Trends exist because investor B wants in on the action of investor A. But why?

My reasoning is that trends exist because information flows around the market imperfectly. Some people get it sooner and others later. Some people act aggressively and some not at all. Eventually, everyone assimilates the information but guess what, the market is already way up by then. But that's another topic.

If not for this fact, stocks would only trade the bid-ask spread until the next news or earnings report. They would then jump or fall, in one fell swoop, to the next price level where they would trade around the new bid-ask. It is the imperfect flow of information, either lagging or leading when someone figures out the news before it is releases, that allows smooth trending.

Nowhere but in the financial and commodity markets do people want to buy more as the price goes higher. My economics 1a professor, Barney "guns and butter" Schwalberg would not approve. Yet, in the markets, the more people that want it, the higher the demand and higher the price, and (drum roll) the more other people want it.

Talk about FOMO!

With that said, there is still a bit of sanity left here. If prices rise too fast - and that is clearly subjective - then some people wake up and realize they should probably cash some of this windfall in. Supply increases and sure enough price action heads a bit lower.

This could happen a few times and the market establishes a sustainable trend with a velocity that makes sense for that market at that particular time in history.

Right now, Bitcoin merits an insanely steep trend. It's new and, as far as a concept can be, really shiny. Maybe next year, when $30 million simply vanishes as it did just this month in competing crypto-currency ethereum, we'll see the market's invisible hand show itself. 
Here is the headline "A coding error led to $30 million in ethereum being stolen." That is absolutely frightening, but I digress, as I always do.
The real question for most markets most of the time is what do you do with trends. The answer is, "ride them."

The next question is how you tell when they end. The answer is, "the market will tell you."

Huh? So they will ring a bell at the top? or the bottom?

Well, not exactly. It is durn near impossible (yeah, I speak hillbilly) to know when the peak or bottom in any market is happening as it happens. Chances are you will have to give up a bit before the signal becomes somewhat clear. But that's the whole point of stops. They let you ride the trend until it changes.

Now, you may have other constraints such as scaling out of a position after X% and Y% moves. Or maybe you see better opportunities in other markets in trends that are not quite so old. Or your funnymentals change. Whatever your reason, you decide to get out. But that does not mean the trend is over. Only the market knows that.

The old saw, "the trend is your friend" is absolutely, positively true. However, you cannot forget about the second part of that, "until the end."  How may people go broke NOT realizing the trend changed and they continue to BFTD? Can you say Internet bust?

Bonus thought: We usually like to lambast the little guy for being in the wrong place at the wrong time. It is the crux of COT (commitments of traders) analysis saying that when the little speculator gets massively long the bullish trend is about to end.

The thing is that the little guy is right during the meat of the trend. Elliott Wavers cleverly call that "wave 3." They may be slow at the start of the trend and overly enthused at the end but the middle of the trend needs the masses to really get going with size and power.

So, you still have to know a little bit about the trend you are in. Identifying that is more than half the battle when you make your buy, sell or hold decisions. Everything else is gravy.