Numerical Price Prediction Challenge

[expiated]

As it turns out, the project I was thinking would be over in February or March of this year is actually going to extend at least into September, so it probably will not be until then or the following month of October that I will at long last be able to return to full-time trading using this system finalized over the last three months.

 

[expiated]

NZDJPY started me off on the wrong foot in a major way, but other trades put me back on track.

As happened on Monday, relying on stop losses as opposed to managing positions got me started off on the wrong foot in a major way again yesterday, but thanks to the apparent validity of NPP's forecast models, subsequent trades made up the difference.

 

[expiated]

With the development of the Numerical Price Prediction (NPP) system having seemingly been fully completed sometime this or last month, until September/October rolls around, when I can begin trading the system in earnest, with diligence and on a full-time basis, my plan is to record random thoughts that might help me clarify how the strategy is best implemented.

My thought is that this might not only help me implement the system with maximum efficiency and effectiveness when the time comes, but could also perhaps enable me to do a better job of teaching it to others if at some point in the future I opt to invite local residents to join me inside a drop-in trading room café sort of setup where they can use what is happening on big-screen displays to execute their own personal trades at individual cubicles/stations.

So then, NPP is basically built on five moving averages or moving average clusters, categorized as follows:

1. A trio of day-to-day moving averages
2. An intraday moving average cluster
3. An hourly moving average
4. A tri-cluster set of short-term moving averages, and
5. An immediate (fluctuating) moving average (on one-minute charts only)

Position entries and exits are made based primarily on the tri-cluster set of short-term moving averages due to the fact that NPP forecast models suggest intraday day trading is the most profitable approach to buying and selling foreign currency pairs, and leads to the highest short-term and long-term success rates.

 

[expiated]

This is a representation of the final Numerical Price Prediction (NPP) set up, which has been simplified for the sake of explanation.

The immediate (fluctuating) moving average is not pictured, nor is the trio of day-to-day moving averages.

The tri-cluster of short-term trend lines is represented by the blue, purple, and black moving averages. The hourly trend line is represented by the crimson moving average, and the cluster of intraday trend lines is represented by the green moving average.

In terms of deciding when to enter or exit positions, the greatest amount of weight is assigned to the moving average tri-cluster, which consists of what are thought of as “actionable” trend lines. A trader should remain in long positions as long as candlesticks continue to form above these indicators and remain in short positions as long as candlesticks continue to form below them.

When it comes to reversals however (entries and exits), this three-fold indicator is prone to displaying false positives, so it is extremely important to consult the hourly (crimson) trend line before making a final decision, even though this indicator is a tad bit slow when it comes to recognizing intraday reversals in price direction—for it generally does not begin to hook in the opposite direction until and unless price has truly made up its mind to go the other way.

To judge the direction in which price is most likely to continue in the longer run, a trader should note on which side of the (green) intraday cluster candlesticks and trend lines are forming, as well as the cluster’s slope—which should not be overlooked!

However, the decision to go long or short is not simply a matter of noting whether these indicators are evidencing a bullish or bearish bias/sentiment. It is extremely important to take at least one other factor into consideration, and that is the amount of distance separating price from each of these “landmarks.”

Plotting corresponding envelopes on the chart (not pictured) greatly facilitates this process and is deemed by this system to be a better method of recognizing so-called overbought and oversold conditions than tools such as a stochastic oscillator or commodity channel index (CCI) because it enables a trader to gauge just how extreme—exactly to what degree—the asset is deviating from the norm, and to also make a valid assessment as to when support or resistance has actually turned price away (thanks to the information conveyed by the short-term tri-cluster, the hourly trend line, and the intraday cluster).

 

[expiated]

I wrote in yesterday’s post that the decision to go long or short is not simply a matter of noting whether the progress being tracked by a given set of trend lines is along a bullish or bearish trajectory, but that it is also extremely important to consider the amount of distance observed between price and those same trend lines.

For example, though the instrument pictured below is extremely bearish, I am watching and waiting for an opportunity to buy. This is because the asset is currently in a situation a lot of traders would call oversold.

Many traders use an indicator such as the stochastic oscillator or the commodity channel index (CCI) to assess these types of conditions—again, often referred to as overbought or oversold. But I prefer a more direct measurement in the form of percent deviation from the mean, which I feel does a better job of conveying the degree to which the asset is evidencing atypical behavior.

I’ve heard some traders belittle those who use the terms overbought and oversold, stating that such vocabulary is flawed, and that it would be more appropriate to speak in terms supply and demand.

I guess the fact that I fail to see a great distinction between the two just points to my ignorance. Either way, neither of these descriptions is able to help me conceptualize the situation in a way that seems logical to me in that, as best as I can tell, the most important factors given these scenarios appear to be how dramatic of an interval is formed and in how short a period of time this formation takes place.

If the asset were truly overbought or oversold, if supply or demand were truly exhausted, I would not expect to see an immediate resumption of progress along the same path after only a momentary break from the advance.

In thinking this over, my mind returned to when I used to trade equities. It is not uncommon for stock traders to refer to periods of profit taking when investors take advantage of their good fortune by exiting positions after having just experienced substantial gains.

Paper profits are not quite a reality until they are pocketed, so smart investors will take some of their earnings off the table from time to time to lock in their gains. I don’t know what market makers in the Forex domain are actually doing, nonetheless, if I imagine this going on, it gives me a framework for explaining in my own mind what might be happening.

It’s been my experience that an asset can continue in a given direction for what seems like almost an eternity, that it will continue along the same path for hours, days, or even weeks longer than one might expect—provided that the slope is gradual!

But as soon as price suddenly moves too much, too fast, it will inevitably snap back at some point. The fact that it will continue forever in moving slowly, but snap back if moving quickly, suggests to me that players not wishing to miss out on the opportunity to capitalize on a favorable development and make their gains a reality before they disappear will begin pocketing profits, thereby drying up supply or demand on a temporary basis, but only for a moment, with a resumption of former price action occurring once the players return to the table for more/seconds.

 

[expiated]

To judge the direction in which price is most likely to continue in the longer run, a trader should note on which side of the (green) intraday cluster candlesticks and trend lines are forming, as well as the cluster’s slope—which should not be overlooked!

Today I set about identifying the single fastest moving average void of "wiggle room" meaning that price might fluctuate on one side of the line or the other, but that once candlesticks cross over it, they will remain on that side for some time, especially if the moving average is sloping in that direction.

 

[expiated]

For example, though the instrument pictured below is extremely bearish, I am watching and waiting for an opportunity to buy. This is because the asset is currently in a situation a lot of traders would call oversold.

There exists in this world a Mathematics PhD from Berkeley who did seminal work in cryptography and AI and has to date made over $20B and counting trading in the markets. And he absolutely would not have the stones to insinuate 'powers of prediction'.

Compare this image to post #115...

Another contributor to this forum I believe would think me arrogant to imagine I could perhaps anticipate what price might do in advance, but I am simply attempting to abide by a principle laid down my the Messiah...

He told them, "You have a saying that goes, 'Red sky at night, sailor’s delight; red sky at morning, sailors take warning.' You find it easy enough to forecast the weather—why can’t you read the signs of the times?"

I'm just a simple guy, but even us simple folk can recognize repeating patterns. I've had a couple of people tell me that Jim Simmons does not try to make predictions, but I don't buy it. The meaning behind the words "he hired another mathematician, whom he’d met at the I.D.A., and they began to create models that predicted the direction of currency prices" are, to me, as plain as day.

At this point, trying to convince me that this cannot be done is likely to prove a hopeless endeavor.

 

[expiated]

Probably any question I might have in regard to how I can use Numerical Price Prediction to successfully scalp the Forex market via a guerrilla style of trading has been answered, so I am now going to try upping my game to another level by setting a personal challenge in the form of a 25 pip minimum profit requirement.

Numerical Price Prediction (NPP) is probably about as developed as I can get it. I can use it profitably to scalp, to pseudo swing trade, and even to trade five-minute binary option contracts. With all that in mind, the way I plan to use it going forward (both now and when I begin "serious" trading in September) is to have my one-hour profiles set up in my Tickmill demo account, which I will use to identify trade setups.

I will have a 60-, 5-, and 1-minute chart configuration for each pair I follow loaded in my OANDA live account. When a trade setup manifests, I will switch from the Tickmill to the OANDA account and follow the relevant pair, entering and exiting positions in accordance with the trend lines, cycles, and price ranges registering on the one-minute chart. So the object will neither be to scalp NOR to swing trade. Rather, it will simply be to remain in trades as long as the conditions for doing so remain favorable, and exit them as soon as conditions become unfavorable—whatever amount of time that turns out to be.

 

[expiated]

Numerical Price Prediction (NPP) is probably about as developed as I can get it. I can use it profitably to scalp, to pseudo swing trade, and even to trade five-minute binary option contracts. With all that in mind, I think it is probably best not to conceptualize it as being used to trade any specific or particular style, but to instead simply remain in trades as long as the conditions for doing so remain favorable, and exit them as soon as the situation changes—however long that turns out to be.

What I have observed during the last week has simply confirmed what I was thinking seven days ago. There are likely to be few, if any, major changes in how I configure my charts going forward.

For me, just moving averages and envelopes result in the best forecast models:

 

[trdes]

Well see overbought isn't as reliable as oversold. That's why people think it's not effective way to view the market. You cannot reliably calculate how many sellers are in a setup and markets are naturally bullish bias so if you're going to short or use overbought you need exact mathematical perimeters that you can measure and see / know are there before you short.


Where as when you're oversold it's different. There's multiple ways to calculate how many buyers are in a setup and therefore accurately predict how high a move up will go. Additionally with up moves you tend to get a lot more movement against your short, where as it isn't the same on moves down, price will often retrace faster and more consistent. Of course if your larger charts meets the sell / short sell perimeters than you'd have to be aware of that and expect more deviations down than normal. Which in turn means you could take shorts on smaller time frames now as it makes mathematical sense.