Under the usual code revision, I sorted ut with a small modification of the training set that seems to have a large impact on the stability of the neural networks in the model. And also in our trading behaviors. I was able to find some interesting patterns in the data, that show that it is possible to reduce a bit of profitability for more reliable information This is why we need to make sure that the training set has enough accuracy.
As I wrote, the difference is tiny, just 1 bar. All training positions have been adjusted. We now have a more reliable set of predictions and the most affected indicator is probably the rV.Target. It was estimating the maximum extension of the move. It now represents the level that when cleared gives a start to the close position procedure.
There is a small (averaged) loss in presumed profitability, compensated by what should be a more reliable set of previsions.
As you know, I do not backtest anything, so let’s see how it preforms in the real world.
r.Virgeel’s model is quite stable, since a couple of years. It took a lot of time to set it up and verify it, avoiding ex-post testing that induces a lot of uncertainty.
I’m talking about the core code that builds up the model to be fed to the neural networks for getting back the reading of the market conditions and also a forecast of its future behavior. Depending on the approach, this code can easily push the hardware to its limits and crash the whole system, so it is crucial to optimize it carefully.
A few week ago I begun testing a larger model, more accurate in the inputs, even if the change seemed not to reflect in the outputs: the neural networks are strange bits of code, very error-tolerant. I almost doubled the dimension of the model, passing from about 12 million neurons to more than 24. Even if the outputs di not changed radically, and that’s a good thing demonstrating that the model is already well-shaped, I suppose that the calculations are now more “focused” having much more parameters to correlate.
At the moment, I’m watching and checking the day by day readings and forecast and it seams in good accord with the market evolution.
Lately, I came across a weird and disruptive idea: some scientists that study the human mind have determined that we do not see what “enters” into our eyes, but instead we “project” through our eyes what our brain is expecting to see. Some scientists call this holographic reality and this process can easily explain some difficult-to-be-accepted aspects of our common experience, as we often do not see something unexpected, or, on the opposite, see just the details we already know (in a person, in a picture, in a text or in an event).
No need to say that if you apply this process to the trading and investing activity it is somehow explosive: yes, we all project our knowledge of the markets into our expectations and then we fund on our opinions to trigger decisions. And we all make mistakes, as our decisions are what we want to see in the market, what our minds project on the reality and we think that this is reality. Delusion is just behind the corner.
Is there a solution? Yes, more than one. You must train your mind to be as free as possible and to accept every possible aspect of reality. Accepting does not mean embrace, but recognize. You must study your field of interest and collect any point of view, slowly building your own. Always take note of the opinions you do not share. In the field of trading, you may use some agnostic techniques as technical analysis – that is not foolproof – to reduce some macroscopic errors. And then you can take profit from artificial intelligence tools, that, if well designed, offer an unbiased view of reality through pattern recognition. It frees you from the necessity to always have an opinion on everything (such a burden!) and so your mind is much less solicited, calmer, more open to recognizing reality as it is.
Today, like every first of the month, the monthly forecast has been brewed by r.Virgeel. The monthly forecast looks into the future for 24 bars, two years from now! and has demonstrated in the past to have a particularly happy touch. Obviously, I cannot make the forecast public, for respect of the paying subscribers, but I want to share a couple of interesting numbers.
First, beware the area 3780/820: if the S&P 500 crosses these levels making a monthly close below, we might be in a long term reversal. Very dangerous and not much probable.
Then, I want to share a valuable number: just to put under pressure r.Virgeel’s skills, it is forecasting a yearly close at 4350. Forecasts is always a risky business, so I take my risk.
Take note. It’s nine months in advance. As the model is adaptive and responsive, this end-of-the-year target may change in the future and I will keep you informed.
spxbot.com has a new blog: new address https://blog.spxbot.com, new graphic layout and new content: a better selection of the old posts that will be updated whenever necessary. Navigation should be easier than before, readability has been improved and I hope you will like it.
The whole spxbot website is under refurbishment, with the aim to point to just the focus of its existence: let you access quickly and easily to the forecasts of the S&P 500 index by r.Virgeel.
I have to thank this man, again. For two reasons, mainly: first, because his writings were the eye openers to understand the markets and how they work, even after many years of investing activity. Second, because without the “it’s all interconnected” message, I would have not tried to experiment with the artificial intelligence model that powers r.Virgeel. It’s always a pleasure to hear him, and here he explains so clearly why the euro is going to fail and why we have to stop thinking linearly if we want to succeed as traders/investors.
If you occasionally read this blog, you may have noted that I often say that r.Virgeel is not a trading system. I want to warn the newcomers that they will not find any automatic fount of wealth here.
Any serious trader has matured its activity in a deeply constructed attitude: call it a plan, or a set of rules, or a set of indicators or all these things together. Traders hate losses like cats hate water: sometimes you need some, but better avoid. So, they build their framework, each one different, each one influenced by the history of each trader itself.
When I first designed the model that will become r.Virgeel, my trading experience was at the “trade-with-technical-analysis-the-fundamentally-chosen-stocks” stage. Many experiments started with sets of t.a. oscillators and were largely unsatisfactory. It took time to understand that I might ask to the model “impossible” answers.
What I understood is that I must ask the model “impossible” answers, otherwise better use the t.a. methods an go on with the traditional analysis. With “impossible” I mean not available using traditional programming techniques. All t.a. is computer-based nowadays, so code is the key to performance. But traditional code cannot perform certain tasks in a manageable reasonable way, like forecasting next bars as the weather channel forecasts the temperatures for the next days. The model, neural networks based, can. Now, consider that saying “tomorrow low will be at XXXX.XX” involves a certain responsibility, bearable, in my case, by the long term performance of the model.
I’ve developed various indicators, during the never-ending development of r.Virgeel, and many have disappeared, leaving an affordable set of unique information. To best fit into everyone’s set of weapons, you have t.a. mimicking indicators, as rV.Target or rV.Stop, and some “impossible” indicators, as rV.Future Bars or rV.ExpectedTurn (which evaluates how many bars to the next turn) or rV.ColorBars (which evaluates at which stage inside the ideal position is the market and shows it as an easy to read colour code).
At a certain point, I started to develop the rV.Position indicators: they are many, all derived by the same learning process that reads the S&P 500 index flow into positions, long and short, picking always the best market position. Now, you may say, this IS a trading system and yes, in some way it is. In the middle, there is the behaviour of the model, that sees and detects a huge amount of patterns and correlates dozens of different inputs. Experience has taught to my faithful subscribers and me that r.Virgeel is very responsive, sometimes too much (it is still young, you know), rarely it gets blind for one bar (this behaviour has been drastically reduced by latest improvements), and it is correct most of the time. I do not consider rV.Position indicators as a trading system, so I do not follow blindly its entry/exit signals; instead, I search for correlations between all the indicators to confirm any trading activity. Of great importance is the Weekly analysis, that has demonstrated to be well integrated with the Daily analysis.
Since few months, I have introduced an automatic summary: r.Virgeel writes a brief summary evaluation on the daily report and this is intriguing: it is a small text and it is working well, opening the road to new possibilities in the model output presentation.
If you think that the trader’s activity is similar to a game, I would like to first say it’s not. For many reasons. What makes to someone the trading activity similar to a game is a misconception of the bet. The novice tends to see a buy of a stock share as a bet, and, well, it somehow is a bet if you know nothing of what you are doing. Because it’s not a game.
But staying with the game analogy, I would prefer, to much adrenaline-driven action games, the more time related Virtual Regatta Offshore, a provider of oceanic regattas simulators in real-time with real conditions, here in the current RORC 2019 where I participate as minushabens. In this case, it is a 16/17 days ride from the Canarias to St. George’s Island. We are about 4 days from the arrival and I’m placed quite well: I was lucky, indeed!
Let me explain: the player have to decide its route considering the present condition of the wind and the future evolution of the wind, based on some known performance of the boat we are racing with (known as the polars of the boat, simply the speed of the boat at different wind speed at all angles). So, the player relies on the weather forecasts to evaluate its alternatives. While in real-world regatta you may apply a wide sort of tactics against your opponents, here you just race against the wind. No need to know the real sailing (it helps), we race against a worldwide huge number of participants. Here, at the RORC 2019, we are 37.000 and counting.
Now, I can guarantee you that my choices are not bets. Sometimes, as it is now, the choice has been more favorable, others are not so exciting, but if you want to appear in the first third of the ranking you cannot trust the simple bet. You must change your mind: you are challenging a force of nature, much bigger than you and good forecasts for 5 or 8 days ahead may easily change in 2 or 3 days, not much certainty around. Is anything sounding familiar? Now, if you substitute the word wind with the word stock market (the S&P 500, in this case), is it clear the analogy of the behavior of the player of the regatta and the position trader on the index?
If you see the game as a be-there-at-that-time-to-be-elsewhere-at-another-time dynamic, it’s not difficult to place yourself in the first 700, possibly between 300 and 500, where I often find myself at the arrival. A bit of luck today is appreciated. Obviously, you bet, there is a number of route calculators available as apps and websites. Every service is based on one of the available forecasts, then provides different best paths, but the game platform has often slightly different conditions and there you are gaming. You must be well placed in the space and time of the game to fully take advantage of the suggestions from the routing software. Here, as well, a bit of luck is largely appreciated. To remain once more in the game analogy, you are requested to dominate space and time, always need to consider the two as a whole. This is a common analogy with adrenaline games.
End of August. r.Virgeel is already positive for two days and gives a projected rv.Target around +3%. Great time to jibe!
The target was then reached at around 3000 in 10 days.
PS. Update – The RORC 2019 finished with me at 787th / 39101
In various previous posts, I’ve stated that the monthly forecast is to take into account as a sort of path tracer, an indication, a generic route for future market action. I tend to consider it in this way, having a much more precise weekly and daily forecasts.
Today, I decided to get back to the far away in the past previsions and checkout their validity: I continuously check the forecast, obviously not always correct, but usually, I tend to be a bit superficial with the monthly forecasts. So, let’s see how it performed.
The first chart was the forecast published on the 1st of December 2017: almost 24 months ago! You may note that the forecast for the present November 2019 (the 24th bar in the future) is in the range 3100/3300.
In the second chart, brewed seven months later on the 1st of July 2018, the monthly forecast for the current month of November 2019 (17 bars into the future) was in the range 3050/3150.
Now, I leave to you any consideration about the quality of the forecasts by r.Virgeel, I just want to say that this is what I mean when I say that artificial intelligence is different and that you may “know in advance”.
Most of us have a general knowledge of what statistic does: it extracts from data some relevant information about the data itself. What a neural network does is to add a layer of correlation between the data and relate it to the desired output. You must instruct a neural network, before using it. You must associate one or more values to each of your info sets (records of a table in a database, usually). This is a process of knowledge transfer. It works well with classifications to produce diagnosis systems.
When well taught, and it is a long and delicate task indeed, the neural network has the ability to recognize patterns in never seen before data and so output the less improbable evaluation from its data bank of knowledge.
What are patterns? A pattern is an event that repeats in different shapes, but always in the same manner, staging always similar processes. The concept of pattern was introduced by the architect Christopher Alexander in its A Pattern Language (Oxford University Press, 1977). It was aimed for architects, and you may consider the approach with this sample 125 STAIR SEATS.
Then the concept of pattern has rapidly spread into the world of software programming, producing the revolution of object-oriented languages (OOPL). Leveraging on the human relational attitudes, patterns rise in every context, not excluded trading and investing, as traders know well. If you use technical indicators, you know what I’m talking about: your eyes are trained to recognize patterns in the charts. That’s why they call it artificial intelligence.
The r.Virgeel’s model is stable since enough to let me say it is now entering its early maturity. I’m working in the refinement field to gain speed, affordability, stability. Less, and this is my fault, I work in the field of communicating how to read and use the model’s output. This is obviously useful to newbies (to the website and also to trading), but also to experienced traders and investor, and even to long term subscribers, that already know the mechanics of the forecasts.
The global goal of r.Virgeel is to detect the best S&P 500 market position available: better stay long or short? Then it provides indicators to detect the incoming market waves projecting targets, stops, with the support of confirmation indicators. r.Virgeel was not conceived to be a trading system, but it sorted out to be a sort of. I do not push it, I always suggest to get the suggestions and verify inside your framework.
Let’s take the Long Position and its dynamic: r.Virgeel detects a deteriorating downward move and warns the short end is near. Signals that mark the bottom are fired. r.Virgeel confirms the long position. Occasions to add to Position are detected, to let you enter/add with margins of safety. Warnings are fired when the position is detected weakening. Next inversion is detected/confirmed and so on.
The previous is the ideal path that resides on the learning side of the model, then when applied to the real day by day market, some signal may be missing, some turn may be a bit bumpy, some uncertainty arises here and there, usually sniffing incoming turbulence.
Please, note that these warnings are to be used inside your framework: your trading and investing style must not be rebooted, just may integrate the r.Virgeel’s suggestions. It may need some time to adapt, but the utility of a totally unbiased a.i. robo-advisor is out of doubt.
Here a visual representation of the Long Position as detected by r.Virgeel. For the opposite Short Position, just reverse the rules.
We have the three main indicators rV.Stop, rV.Target and rV.Position that defines the global situation. Some confirmation indicators are needed because sometimes the main indicators diverge and generate uncertainty. (see bifurcation to approach the topic)
In the real world, here you may see the Alpha Chart generated on 28th of August 2019: it summarizes .Virgeel’s opinion for the incoming wave, well confirmed by recent market action.
Am planning to take monthly subscription. Before subscribe please let me know below queries.
Which platform needed to do trading ? What exactly I get forecasting daily? Any additional analysis i need to know? Before take trade? Will i get entry and exit positions? Will I get text messages to mobile? How much capital needed?
Here my reply:
Which platform needed to do trading ?
Whichever you want. Trading activity is totally up to you.
What exactly I get forecasting daily?
The daily forecast is based on the daily timeframe and I enclose a few samples from of the Alpha Chart for your evaluation. It is composed of many indicators that outline the expected behaviour of the market in the immediate future.
Any additional analysis I need to know? Before take trade?
Investing and trading is a high-risk activity: consider that statistics about the new traders entering the market says that they are destinated to go out of money in about 18 months at a rate of 80/85%. Every trader has its own attitude and this is first what I suggest to you: find the market and the time frame and capital that gives you confidence. Some technical analysis may help, but the simpler the better. I love Elliott’s Wave theory and I’m sure every trader has its own preferred tools.
Will I get entry and exit positions?
r.Virgeel (the brain of the a.i.) was not designed to be a trading system, but it gives signals and a global evaluation of the best-suggested position relative to the market. It produces a chart and a word report to support your vision of the market.
Will I get text messages to mobile?
No
How much capital needed?
It’s a hard question to reply: it depends on your expectations, but basically on how much money you are ready to lose, before getting into a panic. Loosing is an active component of the trading activity. You need to learn (if you do not already know) how to reduce losses and you need to learn (if you do not already know) to trust your tools to maximize your profits.
If you are starting right now, I would suggest you try with a very small account, not much more than your pocket money. In this way, you have a sort of hard stop loss if things go wrong. Consider that every trader has learned the lesson going broke at least once, me as well.
When you get the patterns into the wave observation, you will see the cycles. Cycles that expand and contract, that generates trends, cycles of any dimension. Cycles that change continuously. Unlike technical analysis that tries to avoid the so called”noise”, neural networks love noise, you feed them with noise, the rawest data as possible seems always better, and then you have to train them.
I know at least three methods to extract cycles from an historical sequence: the Fourier’s Transforms and the method of Armstrong. Well, in effect I know almost nothing about Fourier’s Transforms and not enough of the Armtrong method. But I know about another way: the neural networks. What is interesting is that all the three methods are totally different, use different tools and apply different logics.
Training a neural network means trasferring knowledge to the data. You associate, tag, mark, you name it, a certain record in you database to a certain “meaning”. For example, to the days lacking to the next turn. The unique ability of the neural networks is to project values in the future, not in base to an abstract theory, but crunching the real numbers.
Artificial Intelligence can analyze cycles, you just have to pose the correct question, as to the Speaking Mirror: you need to extract meaning from the patterns. One of the r.Virgeel’sindicators, the rv.NC Next Cycle evaluates when and where the next cycle will take place: how many bars into the future and at which price level. It is not a triggering indicators, it is aimed to focus our attention to the incoming events. I consider the rv.NC an alert indicator.
Today I would like to talk about waves. The markets express themselves through waves. But first, let’s look at waves. When on the shore, look at waves. look at them for a long time and pattern will arise behind your eyes. When I first attempted to put a neural network at work, the real first time, after having prepared the database and applied the simplest conditions possible, I was expecting a big awful inevitable crash in the code, or an anyway “impossible” error to underline that I was trying to make something not available. Whooo. It passed and produced a meaningful output. In a fraction of time, all the theories I’ve read about random walks went in the bin. Trash. Market activity is actually pattern-based, and patterns arise from the wavy nature. At that time, I was working using the SPX as the input of the network and it took me many years to understand that everything you want a neural network gives you, it must be on the output side. Self-referential networks may work for a while, but they are going to crash.
Waves generate because many forces are applied to the market. On the water surface, it is the wind that generates waves. In the market, the summing of all the operative setups generate waves. Waves that have intensity (volume) and volatility (price delta). Return to the shore: you will note that waves into the sea have moments of intense activity and other moments of almost flat water. This happens because waves have the characteristic of summing them up, either building giant waves or reciprocally annihilating.
I saw these images by Dave Sandford and, whether I suppose that heavy editing may have been applied, I’m sure you will be pleased to watch:
We learn a very important lesson from these pictures: the dynamic of summing and annihilating each other generates exceptional events, explosions, overcharging, unsustainable conditions that collapse in a fraction of time. If you now return to a long term chart of the SPX,
Semi-log SPX monthly chart
I’m sure you are looking at it with a different eye. You know that enlarging the chart up to a 1 minute realtime, you have the same fractal behaviour. Waves build them up in any time frame.
Now, I would like your to return to “patterns will arise behind your eyes” and consider that, with a long experience, you may easily detect patterns in events. Simpler or more complex, patterns represent a probable outcome of an event, they fuel our ability to “foreview”. Patterns are actually the concept behind our self-defense attitude, they represent the conditions of a series of different components and their contribution to reaching a certain result.
Patterns inside a neural network are subject to the correlation between the inputs, as set up in the code. The network cannot guess something that it is not in the realm of what it already knows. No intuition. Correlation, instead. Correlation is the equivalent of the summation/annihilation process in water waves. In a neural network-based environment, the recipe of the elements that shape it is absolutely fundamental.
Can you correlate three inputs? Can you correlate twelve? many dozens? We do not have even the charting possibility of over imposing many dozens of prices and detect the available patterns. Here comes useful the neural approach. Consider that when analyzing the SPX, r.Virgeel does not “see” the SPX, it has no input from the index. What does r.Virgeel sees, instead? It sees the fluctuations of a selection of markets and econometric data from more than 200. The brain of r.Virgeel works at full steam for daily time frame with a distinguished size of 12 millions of neurons, and each one is a correlation. We are in a territory where human brain is by far over and artificial intelligence is able to produce meaning from what we humans perceive as chaos.
‘Cycles rise naturally into a system into which energy is added or subtract’
The website offers access to the numbers crunched by the neural networks: targets, stops, etc. Even if signals are provided, I recommend you to consider r.Virgeel is NOT a trading system, but a set of weapons you may add to your arsenal.
Since months, and it’s well documented in the free blog, the daily forecast is NOT SENT BY EMAIL, but accessible via the Alpha Chart page on this website. In this way, the forecast is much more efficient and flexible to be updated if necessary (sometimes it is).
Weekly and monthly forecast and analysis are sent by email as usual.
Latest months have been plenty of improvements in the structure and behavior of r.Virgeel. It now outputs a standardized 3d matrix of the indicators, that is easy to chart and interpret. A new interpretation and writing module is now active and r.Virgeel is able to write comments on its own. Two new indicators have been introduced: rv.Swing and rv.Rank. The Alpha chart page (premium users only), reflects all these improvements and it’s now easier than ever to read, now that r.Virgeel has started to write.
Latest months evidence is confirming that r.Virgeel is fine-tuned with the S&P 500 index and very responsive to market turns. It anticipates the turning point with simple signals and progressive degradation of the indicators, plus a warning system. A simple interpretation let’s you discern between corrections and major turns, to get maxim profit from the running position. r.Virgeel follows the market either long or short and with the aid of the FastTrack indicator, it can be useful even to intraday traders, at least for figuring out how the market is positioned. By personal experience, given an affordable background description and diving in the intraday arena, with a couple of averages and oscillators, it’s not difficul to follow the waves profitably, even down to the 2 minutes charts.
I often receive questions about rv.Signal and rv.Position indicators and supposing they generate a bit of confusion, I hope the following may help.
In origin, it was Signal. It was the first attempt to generate inversion warnings. Its model was hand trained. Then, with various steps, it has developed into the first version of the Position indicator. It was not just trying to catch inversions, but to recognize an “open-hold-close” sequence, that for sure is often coincident with the inversion sequence of the Signal indicator, but being able to provide much more complex information. The present rv.Position indicator is the development of the original and has demonstrated long term reliability. It has grown to a group of indicators that work together in reciprocal confirmation.
The basic rv.Position, it evaluates both long and short positions, splitted in
the rv.+Position, it evaluates long positions only,
The rv.-Position, it evaluates short position only
the rv.Target, the extreme to be reached by the current position, calculated as close and high/low
the rv.Stop, the value that must not be broken on close to validating the position.
the rv.PositionColor, a colour code that visually feedbacks: light green for accumulating, light red for distributing.
Here a sample of the training screen for the rv.Position indicator. r.Virgeel learns to take and hold a position until the reversal occurs. rvPositionColor, rv.Targets and rv.Stops are set as consequences. All these indicators are calculated and evaluated independently, without knowledge of each other. The process is now partly handmade, partly algorithmic and supervised. Hundreds of positions concur to setup r.Virgeel’s pattern recognition, in many different market conditions. And every day a new bar adds to its knowledge.
After rv.Position has been developed, the Signal indicator was temporarily dismissed. After a while, I decided to revive it and make it another confirmation tool. r.Virgeel is like a son to me, but I cannot trust one single indicator. Even if rv.Position gives an articulated insight in the market position, I always search for confirmations from a different logical approach.
Now, rv.Signal is composed of hand taught signals, plus other signals generated by a cyclical analysis (here a sample of the minimum detector) at four different levels (short term, medium term, long term, extra long term). Cyclical analysis is not yet complete, but working. It sorts out that rv.Signal is a sort of freestyle turning point detector, that works also for short term swings and mainly as a confirmation for the rv.Position system.
The ideal final charting of the mix of rv.Signal and rv.Position indicators is something like this. A complex of bars evaluation, target, stop and triangles that mark the relevant turning points, mostly in the direction of the underlying position. My attitude is to limit as much as possible to take position against the current trend, even if sometimes the sudden corrective move holds some real value.
rv.Signal ideally fires +1/-1 values. In the real world, the values range from +1/-1 to a fraction of them, and, even if the .3 or .4 values means a very weak warning, I prefer to chart also smaller values to have the visual feedback.
rv.Position ranges from +1/-1 to zero. Ideally, a +1/-1 (Open Long/Short) warning is fired, then its value slides down to a +0.5/-0.5 (Hold) value, and in the final bars of the position, the value goes toward zero, that marks the “Close. Out of the Market” position. In the real world, it often happens the same.
Here you may see an enlargement of the chart that the subscribers have receive one week ago, last Friday night (CET), with overimposed the real weekly action: r.Virgeel, in its weekly version, has nicely caught this week range and direction, shown in the rightmost bar, confirming the turn signal generated by the daily forecast.
In circumstances like this, the power of artificial intelligence shows at its best, being able to provide an information that no technical analysis can even imagine.
Has r.Virgeel access to reserved infos? No. The model is built with the numbers publicly available about the world markets and no mind reading ability has been injected in it. 😉
It is just the power of correlation and pattern recognition: if the model is correctly setup, it sees things that we humans… It’s not magic. It’s number crunching.
Following a subscriber’s hint, I have re-brewed the D-FT historical values, with all the four values that are daily provided and for a longer time span, 60 bars instead of 30. Also, there is the date for every record, to make it easier for you to check out past values (please, remember the D-FT is always calculated in anticipation, for the next bar).
Just a final reminder: the D-FT is a very sensitive indicator and I think it is more indicated to provide a frame to intraday to short term traders, unlike the other indicators from r.Virgeel, that are more directed to position traders and investors.
This is a sample of the file format. Fields are:
date, resistance1, support1, resistence2, support3, bars back from day of calculation (in this case 2019-05-08)
Eight days ago, on the 30th of April, r.Virgeel has fired the first “Close Long Positions”. As usual, since it trained this way, we needed a second signal to confirm, that arrived the following day, on the 1st of May. So, in the next one/two bars, depending on your trading style, the spxbot users have closed their long position, opened back in the first days of January, with a good 16/17% of gain in exactly four months.
r.Virgeel has correctly detected the entry and exit points and it is now evaluating a new target for the current wave downward. I may say, it is interesting to note that usually, at the beginning of a move, r.Virgeel has a more precise outlook of the target, that during the development of the wave gets fuzzier and more undefined (I understand that this may seem against the logic, but it is what I have observed during the years, in many occasions).
Does r.Virgeel have mind-reading powers? No, I just think that some patterns do appear in anticipation of relevant events (someone knows in advance and acts in consequence), impossible to be detected by humans, but not impossible for brute force artificial intelligence.
Solicited by a subscriber, I’ve built the D-FT (Daily FastTrack) data for the latest 30 market sessions, for all the followed symbols:
S&P 500 -SPX
Dow Jones Industrial Average -DJIA
Nasdaq Composite -NASDAQ
DAX Frankfurt -DAX
SHC Shanghai -SHC
GOLD
OIL WTI
10Y US Bond Yield -TNX
I’m happy that someone helps me to check the validity of the output from r.Virgeel: I follow daily the SPX, but time is never enough to check out everything. Please, let me know what is your opinion about the numbers and the quality of the resulting operativity. Comments are open for your results.
Today’s SPX D-FastTrack
Data is provided as CSV files (easily manageable with Excel or with any programming language), packed in a ZIP file. Every file has three lines of general info, followed by the table formed by three values, the two main FT levels and the distance in the past of the bar they refer to. The last line refers to the values calculated yesterday, valid for today.
Following the requests coming from some users that were initially confused about how to interpret the signals coming out from r.Virgeel, I think that nothing is better than to explain how the indicators are produced. I have never done it before so extensively.
The process is very different from the calculations for a technical indicator as RSI or MACD, which involves the direct extraction of values from the price historical data.
Using neural networks follows a different logic: what to I want to know? Think to a medical sample: doctor, I have these symptoms, which is the disease? Ideally, if I have a database of symptoms associated to various diseases, querying the database can give an easily probable range of possibilities, if not the correct reply. This happens now in many major hospitals around the world.
If you have ever used a spreadsheet (who doesn’t?), think to a table made of columns of homogeneous data, each line a single “experience”. You may add one or more columns and associate to each experience a characteristic, a category, a value, a disease, something that give sense to the line of data. Great, you have created an a.i. model. It is a table, where some columns are input data that represent an event or an experience, and other columns that represent a meaning that we want to associate to each event or experience.
Usually, by habit, input columns are placed left in the spreadsheet and associated values are placed right.
r.Virgeel left columns does NOT include any data from the S&P 500 index. All the inputs come from dozens of other market indices, that covers all main stock markets, commodities, forex, metals, yields, bonds, etc.
I may say that r.Virgeel evaluates the S&P 500 as a consequence of all other markets. The influence of each market is directly related to its dimension, so the bond market and the forex are the most relevant.
At the moment the daily model has 74 indices on the input side and the weekly model has 123 inputs on the input side.
At this point, how the r.Virgeel indicators are built? What is placed in the right side columns?
I will cover the main indicators, the rv.Stop, rv.Target, rv.Position and rv.Signal indicators.
First, the market action is broken in “positions”, either long and short. Here you may see a sample of a long position, from the training program. This code is one of the three legs that support r.Virgeel (the others being the data maintenance and the number cruncher): here is where the “experience” is transferred into the model.
The position is marked by the thick white line.
rv.Stop (red dots) is placed to accommodate the whole movement. This is a traditional stop, a value that, if perforated, does invalidate the position.
rv.Target is calculated from the highest high and from the highest close of the final bar of the position.
Two series of signals are placed at significant bars (let’s call them “arrows”): the rv.Position puts arrows at the first and second bar of the move, to mark the entry points, and at the second-last and at the last bar, to mark the close of the position. Then it marks all the other bars of the position with a “Stay long” marker. The same signals are placed in the rv.Signal indicator, except for the “Stay long” markers, plus other arrows are placed at meaningful bars, to highlight minor corrective movements and new entry points available, as the main position is holding.
These four indicators work in accord almost always. Almost. When they point in different directions, it’s a nice warning that something crucial is happening. Big attention is suggested then. After recent improvements, it seems that indecision/uncertainty usually last one/two bars, then r.Virgeel returns under normal conditions.
In the background, you may see coded the rv.PosColor: bright red at major long entry points, bright red at major short entry points, a gradient in between. It is a confirming tool.
All indicators are calculated separately and r.Virgeel is totally unaware of its own past outputs. Every time it produces an output, let’s call it a forecast or a diagnosis, it evaluates present values on the left side against the archive of experiences it has recorded. It is called PatternRecognition and is produced by applying a reverse logic algorithm. The less improbable result is output.
This is how neural networks work. Now you may understand better when I’ve written “A.I. is different“, it has nothing to do with technical analysis by any means.
One recent improvement has been the introduction of the FastTrack indicator. It is the first indicator that has been suggested directly by r.Virgeel. The model has a strange behaviour, probably depending on the fact that I’m not a mathematician I cannot explain it, that produces exaggerated forecasts under defined conditions. Leveraging on this defect, I verified extensively that the output was a fantastic trend follower: false signals are near to nil, considering that even if when they occur, usually you may return on the correct side with a minimum loss. The FastTrack output is a series of four levels. What matters are the two levels that define the central neutrality area.
Let’s make an example. Today’s D-FT (Daily FastTrack) is the sequence 2837.79, 2797.362 | 2808.145, 2788.749
The ordered sequence is 2837.79, 2808.145, 2797.362, 2788.749
The central neutrality area is defined by the numbers 2808.145, 2797.362
Above 2808.145, we have confirmation of a rising dynamic under development,
Under 2797.362, we should have confirmed of a declining action under development.
The FastTrack is an end of day indicator, it is very sensible and I find it useful placed on the short term intraday real-time chart. Actually, using the r.Virgeel’s framework, the D-FT levels, some rudiments of Elliott’s Wave’s theory and a couple of moving averages seems a good setup. I’m testing it at the moment with a certain satisfaction and I will report the experience when it will be mature. Simply said, I’m moving the first steps towards a wealth building machine, using intraday action, based on a weekly staircase. Each week target is based on a stable compound capitalization model. In a few month I will go bankrupt (on the CFD account) or the machine will be snorting and puffing and chugging.
r.Virgeel’s indicators somehow mimic some traditional t.a. indicators and this may generate confusion or misunderstanding. It’s my fault: as the model construction has developed, I’ve got from my previous experience of investor and trading scholar, fishing ideas trying to shape r.Virgeel. Many indicators have been developed and few survived. Some brilliant ideas have been demonstrated them dumb, others have a long track of solidity and sharpness. The quest is not over.
Since a few weeks, subscribers have a new tool, the AlphaChart page.
This page on the website has substituted the daily email and post, it has a larger chart, almost all model’s outputs are represented and it integrates the weekly and the monthly forecasts on the same chart. Most recent bars are enlarged in the bottom right corner, with the projected D-FT levels for the incoming bar. It has all relevant numeric data reported on it. The chart is clean and easy to be read, it has room for improvements and additions, easy to be accessed by your smartphone, tablet or pc.
The AlphaChart is updated every 15 minutes during trading days, to let users have visual feedback in (almost) realtime.
The first chart of the final version of the AlphaChart, dated 8th of February.
By the way, this chart indicates an rv.Target at 2805, exactly where we are almost one month later. Targets are dynamic, new targets have been opened during the past month.
The AlphaChart is still under development: it has reached a first stable step and I will refine and complete it. Also, the chart page is growing with additional analysis and the new real-time updates, that, I have to say, is a bit tricky: it is in alpha testing, at the moment, meaning that it may become a stable feature or die soon for lack of interest. But it may be important, during reversal days, to track the market during the close phase, to get alerts in advance.
I have published the usual monthly update and in the post I have included a significant minority report. One subscriber was surprised by the existence of a “minority report” and asked how does it work. It is an interesting question, that I cannot reply exhaustively, without revealing some well-kept secret about the building of the model. But, I can try to explain.
Like any software tool, r.Virgeel’s code is plenty of variables. One of my long term efforts has always been to try to reduce the parameters of the model to the minimum, to avoid any possibility of over-optimize the networks. Neural networks have their ability to generalize, inducing replies out of sample, inside their realm of comfort.
Finally, I arrived in the latest versions of the model, to just one variable. The one and only that affects the model sensitiveness. Let’s call it sensibility. Low sensibility produces more volatile analysis and indicators; higher sensibility produces results more stable, day by day. If sensibility goes too high, r.Virgeel gets stuck for long periods, inside a sort of trance. There is an interval of best response.
I’m interested in a reactive and adaptive response, so I usually select a value of sensitiveness inside the well-tested range and change it only occasionally. And I also get a look to the forecasts generated with different sensitiveness and if I find a particularly persistent minority report I share it with the subscribers, to warn of any possible incoming event.
Also, consider that, to work, neural networks must be trained. Training means that experience is transferred into the network and what sorts out is that very similar conditions are trained for opposite outputs. It is not an error or a limit, it is inevitable: every market is an alive entity and r.Virgeel brews its reports from a huge correlation matrix between dozens of markets. Altogether, it’s gigantic. Bifurcations are inevitable, are part of the alive thing. Bifurcations and minority reports are different aspects of the same datascape. I’m working on this aspect, but it a long way. Anyway, bifurcations have reduced their aggressiveness and really interesting minority reports are rare.
If you come from technical analysis, you are used to consider the price bars as your primary source and you are used to self refer your data to generate some significance. Inside the model, the SPX is absent from the correlation matrix, being placed on the learning side. Yes, it’s different. The whole configuration of dozens of other markets generates the SPX inside the brain of r.Virgeel – by the way, the most relevant markets are the biggest, not surprisingly. The process is known as “pattern recognition”: find where some similar data is in the archive and learn from it to process the current moment. Once the model works, and it has been real-time tested since 2013, you, me, we are not requested to do much. Through the indicators, r.Virgeel gives a variety of different reading of the present status of the market, designed to be in reciprocal confirmation.
The use of the output we get from r.Virgeel is up to each one of us. I’m sure everyone uses it differently, with different time horizons. The a.i. tools offer a huge potential to enhance many trading styles. At the moment I’m testing the intraday use the FastTrack indicator and the results are really nice. The FastTrack levels on the intraday chart help me to have an immediate frame and it works from 4h to 5mins. You do not have to use r.Virgeel in a specific way: find how it fits your plans.
A final word about the jargon: I understand that sometimes you may be confused by technical terms, but I must use them to try to be clear. Artificial intelligence has a plethora of dialects and terminology, it’s exploding just now and I’m sure that in the next future many concepts will be of public domain.
Here you may see an enlargement of the chart that the subscribers have receive one week ago, last Friday night (CET), with overimposed the real weekly action: r.Virgeel, in its weekly version, has nicely caught this week range and direction, shown in the rightmost bar, confirming the turn signal 
Eight days ago, on the 30th of April, r.Virgeel has fired the first “Close Long Positions”. As usual, since it trained this way, we needed a second signal to confirm, that arrived the following day, on the 1st of May. So, in the next one/two bars, depending on your trading style, the spxbot users have closed their long position, opened back in the first days of January, with a good 16/17% of gain in exactly four months.
I have published the usual monthly update and in the post I have included a significant minority report. One subscriber was surprised by the existence of a “minority report” and asked how does it work. It is an interesting question, that I cannot reply exhaustively, without revealing some well-kept secret about the building of the model. But, I can try to explain.