A tool for comparing your levels to random levels.

AdamAdam Grimes's blogs, General Comments, Trading Theory2 Comments

Squeezing in a quick blog post here in a break from the writing the book, but you guys are going to have to do some of the work!

I have had many conversations in person, via email, twitter, chat, etc, about this video I posted on 12/28/10. Two traders told me it changed their approach to trading completely, but I think this is probably the beginning of a long process for both of these guys. In this video, I put completely random levels on a chart of BIDU, and we then went through the chart examining those levels as if they were actual real levels. You can repeat this experiment yourself with other stocks, and, in almost all cases, you will be stunned by how good these random levels appear to be. So, the question I then ask is how do you know the levels you are using are any better than made-up, random levels?

For me, trading one of the things that has always kept me on task with trading are the epistemological questions. If you don’t already know, epistemology is the branch of philosophy that deals with meta-knowledge, or questions about knowledge. Some of the most important questions are: what do we really know? How do we know we know it? How do we learn new things? These are profound questions, and every trader must grapple with them eventually. Asking these questions about our trading strategies is very hard and challenging, and it means questioning many things that we accept as truth, but the end result is that you end up at the end of the journey with a set of tools you really know and trust. Perhaps more importantly, you understand the limitations of your knowledge and your trading systems and you build a degree of rooted, humble confidence that is necessary for success in the market.

I want to share a couple things with you today. First, after I recorded the video I was going to close the workspace, but then I decided that it might be interesting to see how these random levels travelled forward. So I made the decision to save the workspace, but, before I closed it, I dropped a couple lines above BIDU’s price chart because I kind of guessed it was going to head higher. This was not recorded on the video, so you may not choose to believe I actually did this—perhaps I drew really good levels in yesterday just to make my point? So, I colored the new levels blue so that you can easily tell the difference between the levels you randomly saw me draw on the video and that ones that you cannot verify were drawn in December. I am attaching a .rar file that has charts for every day the market engaged those levels. Watch the video again, and then carry forward by looking at the charts.

The second thing I want to share is an outgrowth of one of the tools I developed for the research in my upcoming book. Many traders asked me how they could test their own levels, and I didn’t really have a good answer. (It’s not a trivial question. You are asking if levels show distinctly non-random tendencies in the market, and even the “real” statistical tests dealing with randomness have some thorny issues.) I came up with a solution that you can find in the attached spreadsheet, but it is still a subjective test. Here it is, but you have to really do some work to get any value from it (so I’m realistic about how many reading this will actually do it):

  • We all have journals and notes of levels we were watching on previous trading days. Whether these are floor trader pivots, Camarilla levels, major chart points, or however you generate those levels, go enter the level you were watching into the first yellow cell in the spreadsheet.
  • The sheet needs to know how far away to place your random levels, so enter the ATR of the stock into the second yellow cell.
  • Press F9 to generate a new set of random levels. If you don’t like how close/far it is placing them, then you can change the ATR input. (The sheet uses a mixture of normals with standard deviations equal to 1 and 3 times the ATR you enter.)
  • I would suggest you go through about 10 charts, and from far enough back that you don’t exactly remember the details of the trading day. Put your actual level on, and also the random levels, save the charts, and then maybe come back to them in a day or two. It’s important that you do not know, even subconsciously, which levels are real and which are not. One of the guys I gave this to went to another degree of paranoia: he did 50 charts of 50 individual days, on timeframes he was not used to looking at (he uses 5 min charts, he put them on 7 min charts). He also had someone else actually put the levels on for him. (Good if you have a long suffering friend who will do this, but I think he made his kid do it.) And he did not stop there: he hid the price labels, the ticker, and the grids (no lines at whole numbers on his charts) so he would have no idea what he was looking at. This is bad enough, but the person doing this for you will have to catalog each file so you know what it is you’re actually looking at when you check your work!
  • If you are a swing trader, you can figure out how to accomplish the same idea with your levels, but you do not want to see the “left” of the chart so you can see where the levels came from.  We are only interested in how well the level worked after you identified it, since that is how we get paid (or pay lol).
  • Then go back through your charts and score each level based on how well price action respected them. (It goes without saying that you must look at the price action following the point you identified the level, not before.) If your levels are good, they should score better than the random levels. If not, well that’s something else to think about.

I have been thinking about this for about 2 months now, trying to figure out how to answer all the people who said I left big questions hanging with that video. I do not have answers for many of the questions people ask me, but this little random generator is a tool that may help you ask the right questions of your own trading.

Don’t be afraid to ask hard questions of your trading. Don’t be afraid to examine what you are doing and figure out what is really working and what is not. Don’t be afraid to knock down things that aren’t actually contributing to your bottom line, no matter how secure they make you feel having them. As Emerson said, “Heartily know when half-gods go, the gods arrive.” Now, go look at your trading and see if there are some idols you can knock down!


2 Comments on “A tool for comparing your levels to random levels.”

  1. This is deep. Dealing with epistemological issues in trading is particularly thorny considering the cognitive biases we encounter when executing under stress.

    Also reminded me of a thesis from a book by an Australian trader, Brent Penfold (The Universal Principles of Successful Trading). Penfold says he sometimes wonders if every system successful traders employ is actually a kind of placebo that gives us the confidence to stick to our trading plans, which succeed via proper risk management. (It’s just a thought experiment, since he also believes you need a system with a definable edge–a positive expectancy.) Thought for food.

Leave a Reply