Forecasting Bitcoin Prices (APRIL 2018)

in #bitcoin7 years ago

A. Overview March 2018

Technical Analysis on March 2018

Bearish movement is still dominating the market. At first, buyers tried to turn the movement up, until on 5 March the movement went down as profit taking act. It is foreseen that at some price levels, the price has failed to breakout resistance points and then the price has continued to follow the major bearish trend. It indicates that some traders were doing such "hit and run" trade during the session in March 2018.
Also, the previous support level in Feb 2018, exactly at $9,280.4, has broken on 9 March 2018 and the price continued to make the lower low support level in March. Moreover, the price failed to break the highest price of Feb and only could reach its peak on March at $11,531, as seen on chart below.

Meanwhile, if we take a look at the Monthly Candlestick Form, the sellers has completed to move the price deeper than open price on March 2018 (around 90% drop), and then the buyers has successfully drove the price up in the last session (around 60% up). Thus, "the hanging man doji" is formed, the rule of thumb says that if "the hanging man doji" has appeared in a downtrend case, then this might indicates the bullish reversal pattern for the next bar ahead. However, this should be confirmed with another oscillator technical indicators for determining the oversold region, as well as the possible support and resistance levels.

Forecast vs Actual on March 2018

According to the previous simulation on March 2018, the result indicated that the probability of uptrend in March is amounted to 56.15%. It means that the uptrend is even chance to be occurred. While, the downtrend is amounted to 43.85% and it is even chance - unlikely to be occurred. Then, the probability the price would reach Highest of Feb 2018 is amounted to 99.84% which means it is classified as likely - most likely to be occurred in March 2018. Conversely, the lowest price of March 2018 would reach to Feb 2018 lowest is considered as rare to be occurred, due to the probability value is amounted to 2.891% (lower than 99.84%). Furthermore, according to the previous simulation, we notice that there are several price levels which possible to be reached in March 2018. I have drawn the levels on the below chart, in order to provide us a better visualization as well as the convenience in comparing with actual data.

Based on the above chart, we could say that the previous forecast doesn't satisfy enough to simulate the actual movement. Obviously, as aforementioned, the actual highest price failed to break the highest price of Feb 2018, even though the probability is counted at 99.84% to be occurred. However, we also notice that the lowest price of March 2018 failed to break the lowest price of Feb 2018, as its probability is having 2.891% chance to be occurred. Thus, at certain point of view, our simulation is not so bad for predicting price levels. But, I prefer to simulate the movement as well using extreme value analysis in forecasting the price levels on April 2018, in order to give us the different view based on best and worst case.

B. Input Parameters for Simulation

In this section, I would like to describe the modified parameters that I would use later on as input data for simulation. There are two parameters which should be updated before running the simulation for March 2018 : Distribution of Historical Daily Market Cap Growth and Decay (in percentage) and the forecasted total supply of BTC from 1 March – 31 March 2018. Also, I try to simulate the extreme value analysis using Generalized Extreme Value Density Function for simulating the best and worst scenario. The following sub sections would describe the parameters respectively.

Normally Distributed Historical Daily Market Cap Growth and Decay

The historical data of March 2018 has been collected, exactly from 1 March – 31 March 2018, and it would be utilized as input data for recalculate the Probability Density Function (PDF). The prior PDF shows that the mean of population is amounted approximately 0.19%, while the standard deviation is approximately 3.99%. As the latest samples gathered and incorporated into population, the mean of the updated PDF decreased to 0.16%, however the standard deviation is increasing into 4.00%. This indicates our sample variance are getting more vary than the prior one.

Generalized Extreme Value Distributed Historical Daily Market Cap Growth and Decay

The extreme value analysis would give us a new perspective for determining possible best and worst case in our simulation. Since the BTC bubble is considered burst, we have to consider the possible worst case of market movement. By simulating the extreme analysis, we also consider the lower price levels which possible to be passed by the price in April 2018. The extreme distribution has been generated, as seen on the below figure. The distribution has skewness factor (k) around -20.37%, this indicates that actually the distribution of growth/decay is classified as "fat-tailed" distribution. While, the mean and standard deviation of distribution are counted to -3.91% and 10.87% respectively.

Forecasted BTC Total Supply from 1 March – 31 March 2018

The other input parameter is forecasted total supply of BTC. Total supply of BTC would be assumed as linear time series function. The growth factor of its supply would be assumed as non-linear time series function, in this model I would utilize the lower incomplete gamma function for the growth factor.

C. Simulation Result

This section would summarize the result of 40,000 steps of iteration in simulation. The bottom line of the calculation has been summarized in the following table. The colorized bars in the table may give us some useful information for making decision to enter and exit the market. One thing to be noted, the colorized bars indicate how possible the price of April 2018 would mooning or drowning into certain range. Thus the colorized bars should be matched with the probability interpretation meter which I have also attached in the intended table. I have drawn in a chart that consist of possible range including some price levels which should be taken into account. The ranges as well as the price levels are based on the result of simulation.

Conclusion

Based on two performed analysis (technical and quantitative), we may conclude that:
  1. Most likely, the lowest price of April 2018, might fall in a region between $4,759 - $7,855. However, there is also a chance for the price to fall in a region between $2,835 - $4,759 on April 2018.
  2. Most likely, the highest price of April 2018, might climb in a region between $9,779 - $12,875. However, there is also a chance for the price to climb in a region between $7,855 - $9,779 on April 2018.
  3. It is a good opportunity to buy BTC when the price have reached $5,910, $3,263, and $2,835 (or lower than those).
    1. Enclosure

      I am not a financial or trading advisor, I am just an ordinary engineer who have interest in cryptocurrency. By posting this result to steemit, I just want to share what I have done in my leisure time. I do not enforce you to trade or investing in my way. Every people has their decision and unique trading way. Thank you for reading my post. Enjoy your crypto!
      • BTC : 1KZQipWvGBL6PijbeE1QEGMDpe18U9zJkY
      • ETH : 0x9f7d072e407d2f79ea165a02b38c8d05d1fcd3d2
        • Social Media :