Once Upon a Time

History repeats itself time and time again makes it possible to discover hidden order in the apparent chaos in financial markets. So what happened once upon a time will happen again when intertwined time cycles indicate the same energy to cause panics or bull markets. People respond to energy whatever the source is.

History repeats itself

The research we have complete in the last decennium was only possible because nowadays we have available a vast amount of price data as well as the fact that computers are fast enough to process this data using complex algorithms. In addition, programming software has evolved to such a good quality that programming complex models and its maintenance is possible today. Also science and mathematics have made big steps in understanding and modeling complex interrelationships using big data.

Consequently,  we have been able to use history to test our model and improve it further empirically.

In simple words, we can test our time cycles and see what happens any time when certain cycles are active and how the markets respond. 

The last step, was to statistically research every period and every cycle for its performance in the past and determine its significance.

For this purpose we have calculated time patterns, which are dynamic and fractal in nature, for the Dow Jones from 1900.  What a pity it is that this data mostly only is available on a daily basis. However, it is good enough to chart the workings of cycles and compare this to the period of the last 20 years of which we have intraday data.

The results of the analysis of the Dow confirm our statistical evidence of the intraday data, making clear what cycles work positive or negative on average. 

By selecting time cycles that are minimally are 70% correct and/ or deliver an exceptional performance, it is possible to select Long and Short opportunities for the near future which our models predict using time cycles and its patterns. This way we have been feeding our DeLorean product that pictures the openings in several markets day by day for the next month.

Also we have developed proprietary trading systems that trade intraday when our software signals to trade long or short.

Some examples of Dow history

Below we show 2 very important periods with a great impact on the financial market.  The first is one of the longest bear markets in history during 1929- 1932, the second totally different being a flash crash of only 1 month or so in 1987.

Dow Jones in 1929:

This example shows how our time cycles strongly forecast a bear market for the years 1930-32 in which period an astonishing number of negative time cycles are present, which are the clustered purple bars with “Bear”above it. Also the indicator below the chart shows how red(=negative) dominates for a long period with the lowest value ever. Before this period green(= positive) is much stronger , leading to the top in 1929. In 1932 there is a strong positieve time cycle (big green bar), exactly where recovery starts.

Dow Jones in 1987:

Very interesting is the fact that the chart for the Dow shows a large positive time cycle (green bar) exactly at the top in 1987, followed immediately by a very large cluster of purple bars, which is very negative and predicted a decline.  Green time cycles dominate since 1983 and thereafter from 1985 until october 1987 as indicated by a green curve in the chart. Red cycles (red curve in chart) are prominent in 1982 and after 1987 until october 1988. Then positive cycles took over again until 1990.

Statistical evidence

below we will show you some examples of the output of our statistics. Firstly Time cycles that correlate highly with negative markets:

These cycles have a probability of in between 70 and 100% (win/loss, last column). The pattern/cycle that gives the most total negative performance (-11.38% ) and most observations (20) has been highlighted, at the left bottom every occurence is listed. A very significant cycle with 75% hitratio (1-25%).

Secondly time cycles that correlate highly with positive markets:

These positive cycles have a probability of in between 89 and 100% (win/loss, last column). The pattern/cycle  that gives the most total performance (+13.30% ) and most observations (20) has been highlighted, at the left bottom every occurence is listed. A very, very significant cycle with 100% hitratio (1-25%).

These statistics are available for any time period from 1900 – 2018 or the total period. Also, we can take into account that positive and negative time cycles can be active at the same time,  and determine when which time cycle is stronger or will follow up on the other.

Once strength and performance has been attributed to time cycles, we can use this to predict the nearby future.

Devising a Trading systeem

By means of a far-reaching quantitative analysis we have been able to make a model that reflects the energy quite well, positively or negatively, in the financial markets.

The software has been developed in such a way that it has calculated in advance when the system is going to buy or sell. This is based on our time cycles which have discovered a hidden order in the apparent chaos of the markets. The nice thing about this is that my original idea of ​​how these patterns work was confirmed, not a  complete surprise of course.

In this hidden order time patterns are found that are either positive, negative or mixed and in this way indicate whether the market is going up or down.

Of course, the question was whether and to what extent the model would predict fluctuations in the market correctly. As said previously, we tested this extensively on price data from 1900 of the Dow Jones and on futures data of  the Dax from 1997 with minute data.

The price files from 1997 already showed that the model is working well, although we had not yet been able to investigate all underlying data. That has now been accomplished. In addition we have now 100 years of statistical data from 1900 onwards at hand.

Having statistical research means that we know which return and risk the different time cycles bring about positively or negatively by analyzing price movements in the past when these cycles were active.

Since we can calculate these time patterns with our model and know when they occur, we can select cycles that are successful as well as having a high probability in order to use them for future actions ( buy / sell) in the stock market.

To give an example, in the Dow Jones over a period of 100 years from a number of 50000 combinations, about 3000 time patterns have been found, which on average have a superior predictive power. We look up these patterns found in history for the coming period in the near future and draw these green (up trend) and red (down trend) areas in our price chart. At these points in time the trading system will automatically buy and sell, without having any connection with that period itself. This means that historic prices are not correlated with the prediction.

The system has been running live on the Dax futures in a trial setup (not yet in the Fund) that approaches reality for the last 6 weeks. The return has been excellent and the risk is relatively low, namely a return over this short period of approx. 5% with a maximum drawdown of approx. 2%. Here, on average, has been traded  without leverage. This was also a good test if the software worked flawlessly.

The risk is limited, which is only possible by being in the market for a relatively short time when the chances are the best. In addition, the use of a policy that protects profits and cuts off losses is necessary.

Above the graph of the last month where the system was executed in real time on a simulated account. In the table on the left you will see red arrows for the result achieved on a portfolio of Eur 100,000, as well as below the maximum drawdown. The chart below the price chart indicates the equity curve, the development of the return, going from 100 to approximately 105%. Finally, the green and red areas are visible, also for the near future.

Then the question remained what would happen to the return and risk if we simulated the entire period from 2016 to the present in a good way that approaches real trading. The short-term result is significantly good, will it remain so over the longer period?

Over the longer period of approx. 2.5 years, this simulation gives a result from 2016 of approx. 60% at a maximum risk of approx. 3.7% drawdown. Much better than we expected, but a welcome surprise. more importantly, it is in line with the shorter period.

The longer-term simulation confirms the picture of the short term. A good return with a low risk. Here, too, you see the Equity curve that rises very evenly and in 2018 goes sideways in spite of a declining market.

Our conclusion is that the system is ready to be introduced gradually. The question is always how the start will be, but the limited risk certainly makes this acceptable.

Back to the future

Previous article we said:

To finish we will give some forecasts for future events. Firstly we are expecting volatile, mostly negative markets in the months of July (WEAK RISE) and August (CORRECT) to begin with. Other sensitive dates for negative markets are around:

August 16 (CORRECT) around low) and around August 30-31(CORRECT), 25-26th of September(CORRECT), The week of 23/24th of November 2018.(WAIT AND SEE)
Also we have projected the time waves, the specific waves that build cycles, into the future for positive events. We mention the following. July 12
(CORRECT, up trend started),16(CORRECT) – August 6 (CORRECT start up) -27-28(CORRECT)

0