Numerical Price Prediction Challenge

I have transposed all of the relevant knowledge I've gained via "dynamic probability trading" and using a "triple-zone strategy" down to the one-minute timeframe in preparation for an ongoing attempt (which I will begin next week) to establish a precise set of commands that might enable an automated trading program to trade the Numerical Price Prediction system flawlessly.

The idea will be to enter positions at key levels, exit them when warranted by price action, and reenter them when circumstances make it favorable to do so. It is hoped that this will virtually guaranty accumulated profits and essentially eliminate any drawdown.

 

My chart setups have been essentially the same for several weeks now, except that I have replaced a number of “standard issue” moving averages with derivations of what I call “instantaneous” moving averages.



I made an initial error when coming up with the latest set of guidelines for applying the use of this refined configuration, but things have been going relatively smoothly since then, so I plan to continue operating in compliance with the guidelines and setups as is...

 

Present configuration compared to October 2018 (on the right)...

NOTE TO SELF:

I’m looking at a 15-minute GBPJPY chart with the latest configuration (not pictured) and the white moving average is bearish, so I’m regarding the pair’s overall intraday bias is being the same.

The short-term moving average has just turned south (within the last hour) and candlesticks have begun forming below the moving average cluster (within the last 30 minutes). The purple moving average is the only indicator that has not yet conveyed even a hint of the pair initiating a new leg downward.

However, I have other things to do and cannot wait around for this last indicator to concur, so I’m going to go ahead and enter a short position (Sunday night at 9:30 p.m. PST) and come back later to check whether or not the trade is successful—all in an effort to determine whether I’m assigning the correct amount of weight/significance to each of the various moving averages plotted on the graphs I'm working with.

Most recent progress thus far...



UPDATE:
Take-profit target was already hit at 9:38 p.m. (I shouldn't have been so conservative—I could have almost doubled my gains.) Daily success rate is finally above 90% again. I am only entering positions following pullbacks and only when practically everything is aligned just about perfectly, so I need to start writing down as text what I'm doing based on what I "feel" I'm looking at so that I have it all in a clear, objective, unambiguous format.

 

My initial self-directions for trading on a 15-minute chart, is to (1) enter trades in the direction of the white moving average whenever (2) the short-term moving average is coming out of a pullback and (3) candlesticks are beginning to form on the “inside” of the moving average cluster (4) in the same general area that the white moving average is located.

Fortunately I checked in on my last trade and ended up bailing out early, having lost some, but not all, of the potential profit that had been available. So I should probably put into words what I dislike about what I see developing...



First, something about the short-term moving average is bothering me. So what is it? It's the cycle! The wave is now completely horizontal. So then, it is okay for the short-term moving average to cycle up and down so long as a straight line drawn through the center of its amplitude is angled in the "right" direction. If it is horizontal, it's time to stand on the sidelines.

Also, the moving average cluster is also progressing horizontally. That's not negative in and of itself, but the candlesticks need to be sliding along and then breaking into a resumption of the previous trajectory.

Yeah, there are several things going on here: (1) The candlesticks are not sliding along, they are bouncing up and down. (2) The short-term moving average needs to slide along as well, but it is instead crossing over/above. (3) The moving average cluster is beginning to hook north a second time! (Though ever so slightly.)

So, to explicitly define or recap what I need to see to approve a trade:

1. A straight line drawn through the center of the amplitude of the short-term moving average's wave cycle needs to be angled in the "right" direction.
2. The candlesticks AND the short-term moving average may not counter the prevailing trend beyond simply sliding along the horizontal progression of the moving average cluster.
3. The moving average cluster may not initiate more than one hook that counters the prevailing trend unless intervening progress was made in the direction of the prevailing trend, led by candlesticks and the short-term moving average after their sliding off of the moving average cluster.

 

What changed?



One: The candlesticks pulled back behind a down-sloping white moving average but did not continue north, so there was an extremely high probability that the pullback was almost certainly over. Consequently, a stop loss could be placed right above the local high because the forecast model says that under such conditions price ain't goin' back, so if it does, everything is invalidated and you most definitely need to exit the trade immediately.

Two: The candlesticks looked to be making a clean break below ALL the moving averages.

Three: Price appeared to be attempting to break support (the previous local low).

 

Before retiring for the night I entered two position that ultimately proved to be successful trades, but I only need to evaluate one of them in writing to continue my effort to translate the things I see that make me “feel” good about a given trade into a clear, objective, unambiguous format.



My main question is: “Why would entering a long position in the first instance where it appeared that price might’ve been reversing to the north (the arrow on the left) have resulted in a losing trade, whereas doing so in the second instance (the arrow on the right) led to my realizing a profit?”

The first thing I notice is that in the first instance, the trip south consisted of only about a 40-pip drop, whereas that leg of the journey in the second instance was about a 115-pip drop.

I would also note that the swings in the short-term moving average (AND the candlesticks) in the first instance were noticeably more “violent” than in the second. Moreover, the moving average cluster clearly crossed to the other/"wrong" side of the white moving average. That didn’t happen in the second instance, nor did it happen in the image attached to Post #95 (or if it did, it wasn’t by much, and it definitely wasn’t all that clear).

Perhaps the key difference, maybe the most significant of all, is that there was never an actual pullback in the purple moving average in the second instance. Of course, the image attached to Post #95 does have an actual pullback in the purple moving average. But the difference between that situation and the failed attempt at a reversal pictured above is that the pullback of the purple moving average was followed by a clean breakout in the direction of the prevailing trend in the previous post, whereas that never materialized once the purple moving average pulled back in the present example.

 

SUMMARY OF ANECDOTAL NOTES:
Numerical Price Prediction Challenge Post #97

1. Always enter trades in the direction of the white moving average.
2. It is best to enter trades when the short-term moving average is coming out of a pullback.
3. However, before doing so, wait for the candlesticks to begin forming on the “inside” of the moving average cluster.
4. Also, try to find trade opportunities where the exchange rate is still located in the same general vicinity as the white moving average. Perhaps the most ideal situation is when the candlesticks pull back behind a clearly sloping white moving average, but then fail to follow through on the attempted reversal.
5. It is technically okay to enter trades when the short-term moving average is cycling up and down, but NOT if the amplitude of the waves it is displaying are relatively pronounced. (If the waves are very mild [low amplitude/relatively long frequency] or the candlesticks are simply sliding along the longer-term moving averages, there is much less need for concern.)
6. Moreover, entering trades when the short-term moving average is cycling up and down is much more acceptable in situations where a straight line drawn through the center of the associated wave’s amplitude ends up being angled in the same direction as the prevailing trend.
7. Consider it a warning sign if the moving average cluster crosses to the “wrong” side of the white moving average.
8. Likewise if the moving average cluster begins to display more than a single hook in opposition to the prevailing trend (unless intervening progress is made in the direction desired led by the short-term moving average and a progression of candlesticks immediately following their having slid off the moving average cluster).
9. The above warnings are more or less nullified if and when the candlesticks look to be making a clean break beyond ALL the moving averages—especially if they will be establishing new local lows/highs at the same time.
10. If a position is being entered when it appears that an asset is attempting to initiated a major reversal in price direction, it is likely to be more successful if the previous leg of the journey was more substantial (i.e., 100 pips or more, as opposed to only 30 or 40).
11. Whether looking at a possible major reversal, or a continuation of the prevailing trend, it is not a good sign if there is an actual pullback in the purple moving average. In such situations, a trade should probably not be executed until and unless there is a clean breakout in the direction of the prevailing trend.

 

As pleased as I was with the performance of the chart setup I started using yesterday, I discovered late this morning that it was unable to account for all the price action I observed when reviewing charts after the final trade I made during the last 24-hour market cycle.

I’m hoping the inclusion of one additional moving average (the red one plotted on the above setup) will do the trick. Technically, this is not something new. It is a moving average I was using regularly perhaps a year or so ago.

However, the guidelines based on the previous setup called for strict adherence to limiting trades to positions coinciding with the trajectory of the white moving average. This will no longer be so cut and dry. I’m now going to have to evaluate the relationship between the white moving average and red moving average, figure out how they interact, and then how I should react to the dynamics between them.

Moreover, though the red moving average provides a more honest assessment of the direction in which price is probably ultimately headed, it evidences a significant amount of lag, to such a degree in fact that it seriously compromises the dependability of the tool’s application! So how do I compensate for the invalid communications that result from so much delay in measurement?

At this point, I have no real answer for that question, so I’m just going to have to hope I can keep my success rate up while I look for one.

 

I have no idea whether anything I am about to write has any credence whatsoever. I’m just recording my initial thoughts and observations in the hope that I will eventually reach a point where I am gaining insights that are actually true and useful.

First, I have eliminated the inner moving average envelopes and bumped the envelopes that remain up to a higher period. Given that Ichimoku Kinko Hyo already has dibs on the term “cloud,” I am going to refer to these bands as the “shell,” since I find them reminiscent of orbiting electrons.

My first thought is that it appears to me, at least at first glance, that if the purple moving average begins to reverse direction after having reached the atmosphere of the shell, it ain’t turning back until after it has once again made contact with the red moving average. Hence, I should not execute trades that would require price to head back toward the shell once the purple moving average has begun retreating in this manner.

Second… If the purple moving average reaches the atmosphere of the shell, until and unless the short-term moving average (or the moving average cluster?) pull back to such a degree that it (they?) appears inside the space between the purple and red moving averages, I should expect the candlesticks to eventually resume progress in the direction that matches the slope of the shell’s trajectory.

 

This might not be necessary. The third "rule" listed in post #97 is: "Wait for the candlesticks to begin forming on the inside of the moving average cluster before entering a given position." If this rule is followed, whether or not the white moving average accurately reflects the direction in which price is ultimately headed might not be an issue after all.