Monthly Archives: April 2015

Shared time elements

There are two main areas that I wanted to look at in regards to just time: How time relates from Complete wave to wave, and within a single wave. Single waves are harder to do because there are a lot of short waves (2, 3, 4). Doing it from one complete wave to the next is easier, and gives a “set” that the trader should be on the lookout for.


As long as the wave length of the current wave length developing is 4 or greater, there is a high probability that at least one of the wave points will share the same hour as the one before it. Not within a span of 1 hour above or below, but the direct hour. Interesting. Average probability is 70%.


Edit: More in depth look at the numbers



AJK Wave move filters

The first thing I thought, before I even started looking at the data, was how often the sequence is TST – that is, the current move, retracement, continuation. Simple! and with the general “laws” of the market, this is very expected and therefore not very useful. So I started to break the wave components down a bit.

The most common continuation wave type is TST (Trend, Spring, Trend). The most common reversal wave type is TE (Trend, Expand). There is a second short continuation wave type: TEE (Trend, Expand, Expand).


So TST and TEE wave types make up 66% of all continuation wave strings, while TE wave types make up 42% of all reversals. wow!

Naturally, I went to investigate how these numbers change over the course of a wave move developing


Kinda cool to note, although not nearly good enough as it’s own edge because the numbers float around 50%. I think the probability changes more as a result of price rather than as a outcome of time (length) of time not taking out a price level, although this does still lend support to investigating what is happening in these waves. Do these two elements work together? Or does one simply explain the other? In the case that they are relatively independent, wave lengths and time can sometimes offer spots to target certain levels.


Currently price is on wave 6 (TSFTS F/T) and floating around the 50% mark. Hmmm.



Intro to retracements w/ AJK concepts

I think this, along with time, are the two biggest concepts to explore. They should be done together, but sometimes I wonder if I even have enough data to make a statistical inference with how many possibilities there are when you use as many categories as there are possible to separate possibilities! Hmm.

My first test here was to look at the probability for the next wave type to be a continuation or reversal wave. The numbers here look a bit better (stronger) I think, and this may be due to the results inherent “HTF” seeking categorizations, although with how it blends waves together. Whether this is good or not, I don’t know! We’ll see as I progress.

net ret

When we look at the numbers this way, the probability for the next wave set to be a continuation wave is 66%. quite high! Trend trading trend trading trend trading. It’s suggesting that the biggest issue to over come is getting out of positions that will ultimately fail (33%). Since the base line is 67/33, all probabilities must be compared to that. Any numbers above 67 are favoring continuation moves, while any number lower than 67 suggests increased probabilities for the move to fail. It looks like in general, as long as the retracement is no larger than 66%, the move is favored to continue. Interesting stopping point, somewhat close to 61.8.

AJK concepts, Intro to Timing

Finished my book! I think.. Easily the most interesting part of the book was the suggestion that time plays a big role in successful (swing) trading. Okay. Fair..

Naturally, when we think of this statement, we think that we should logically be trading specific periods of times during the day. Yes, past studies show that extremes are more commonly made during certain hours of the day, but I think that area of research is quickly ended. That is, I find that that logic flow doesn’t allow itself to be expanded upon and tested through more filters and additions. What is being suggested now is the repetition of time, combined with price action. Now that is a tall order for testing.

It leads itself to the question of “how do I know if my swings are correct?” To that I believe that we must simply do the best we can do. “perfect swings are different for every trader because the criteria we need varies. In the end, I think that there is a minimum threshold that a wave indicator or system must meet, rather than requiring a very specific kind (the “best”)

I decided to take a stab at repetition to see if it was worth pursuing more.

I think that this idea will eventually be dumped if I can’t find a way to drastically improve it, but it IS quite interesting to look at:


The numbers along the top, 0-23, are times (hr)
Each category below are my wave types
The first row in each category is the actual result, while the second row is the expectation of randomness

Lets take the “All” as an example (that is, all strings as they appeared in history). The probability that a swing extreme occurred at 0:00 was about 3%. The probability that a swing extreme occurred at 12:00 was 7%. Therefore, assuming randomness, if a swing extreme occurred at 0:00, the probability that the next swing extreme occurred exactly 24 hours after (0:00) should be about 3%. In reality, the probability that 2 consecutive swings occur at 0:00 is.. zero. Never. Hm. Now, it should be noted that the average time length from one swing to the next is somewhere between 15 and 16 hours. So really, maybe it’s not too crazy to see that two swings don’t occur right after each other 24 hours apart. But if that’s the case, then why does 17:00 have a whooping 7% chance of 2 swings occurring at that specific time?

Additionally, maybe there’s something to look into when we take specific swings into account. The probably that two trending waves occur at 6:00, 7:00, or 8:00 are all over 10%! If I take this 1 step further, if I had an extreme occur at 6:00, 7:00, or 8:00, the probability that the next trending swing occurs at 6:00, 7:00, or 8:00 (using a wider range now), is over 25% or any of those three hours. HMM.

Will definitely need to look into this more.

Also, i think this area is generally not talked about as much as price, and it might be fun to try a little. Maybe it will be of more interest to the readers in this way 🙂

Edit: Perfect example


Wave 2.2 Sequences

Completed wave sequences! These are “real”, or taking actual price into account. The difference between the two is that one is looking to forecast the next 3 leg wave, and the second (this one) is looking to forecast where price is going.


This picture is only showing part of it because throughout the years, there are a LOT of different wave types, especially if the first wave is large and price retreats into a range.

Only counting single waves


Calculating doubles


Hmm. Kind of interesting and reaffirming a common theory. Price that is in a particular direction tends to stay in that direction. That edge remains 60%, which is what I’ve been looking at in previous studies. Secondly, price that does not make a breakout (continuation) will make a reversal. This also seems to be true from the above stats. The second set includes all trend strings. If the string is TEFSETE, the first will record just that, the second one will record TEFSETE, as well as TE. This means that the second method has a tendency to pick up additional strings in areas that are, in reality, flats within the major trend swing. What this translates to is more time, since every letter is equal to x time length, and more letters=more time. Thus, more time spent not breaking out=more potential for reversing! Neat.

I think now that I need to run some additional studies on the movements themselves, and then lower time frame analysis will definitely be required as well because calculating swings this way leaves a much higher chance of “truely” seeing the higher time frame.

Perhaps this more “complicated” way also provides more clarity and precision by also being more forgiving.

Wave Modeling 2.2

Hmm. Been a while since I touched my data sets. The book has been kind of nice in giving opinions that I can then translate back into my work to test to see if the theories hold. Will have to look into more later.

There are a couple of things that are really crucial for  my current work to function well. No wave theory is perfect, but some are better than others. Currently one of the issues are consecutive tops and bottoms, which occur with about 20% probability. I think I’m missing some form of “active S/R”, and this is the beginning for that. Nice to see some skew on heavier retracements

max retracement



Some support for the 30% min? =O


Very robust finding…The next steps will be something I haven’t quite attempted yet.

Tracking the Newbies

This was an idea that I, as well as many other people, have wondered for a while. If retail traders always lose, why not trade against them? There’s no way to accurately backtest such a strategy, so it really can’t be done until someone actually does it. I don’t read forums much anymore to know if someone is doing it, so I’ve decided to add it to the list of things that I do now!

With such a theory, there are so many ways to decide how exactly to trade against the retail traders. I won’t reveal the exact formula atm (though I have no issues doing so if it does well) but it is taking data (manually) from Oanda, ForexFactory, and Myfxbook. Not sure which pairs to trade, but I’m starting with 4.

Lets see how it goes!

Edit: discontinued.. didn’t have the time to keep up with it properly