Cindicator Bot: January Summary

Crypto in January
January 2018 proved to be just as volatile as December 2017. On 14 January we witnessed a new all-time high in the price of Ethereum — the currency then underwent a 40% correction only three days later, on 17 January. Ripple, however, which had a market capitalisation of $144 billion at the beginning of the month, had fallen by more than 70% to $39 billion by mid-January.

By the same time, the overall capitalisation of the crypto market grew to all-time high at $830 billion at the beginning of the month, but saw an almost 50% collapse in under a fortnight, falling to $430 billion.

As for the five leading cryptocurrencies there was no change by the end of the month, although Ethereum regained its second position, overtaking Ripple. On top of that, Ethereum is now closing in on the once unquestioned leader of cryptocurrencies, Bitcoin.

31 December 2017

28 January 2018


Indicator stats

All things considered, January was probably the most dramatically volatile month in the history of the cryptocurrency market. In times like these, our analytical products acquire a special value. In this section we will go into more detail on our statistics for the month.

We sent out 226 indicators during this busy month — almost twice as many as in December. The number of indicators grew in each of the three tiers relative to the previous month.

In terms of the distribution of indicators by type, we saw once again that the highest number of indicators were sent to users with access to price level indicators. Our research has shown that this type of indicator sees the highest demand from our users.

It should be noted that January saw a marked increase on the previous month in the number of indicators sent to Beginner tier users, thanks to the increase in Market Events indicators. Looking at the distribution of indicators by type, we can say that this month, the number of fiat indicators increased somewhat as a result of more company report indicators being sent out. The overall crypto-fiat ratio remained practically unchanged, however, standing at 7:3.

Perhaps the most interesting aspect of this report for the majority of our users will be the figures for our models’ accuracy. We have already discussed the dramatic levels of market volatility in January, and inevitably the accuracy of our models was affected by this.

It’s worth pointing out that times like these are very useful for the Hybrid Intelligence ecosystem as a whole, and Cindicator’s machine learning models in particular. This is because the preceding period didn’t give our models the chance to accurately assess the market in times of such volatility and an overall bearish trend.

Of the January indicators, at present 171 indicators have now ended. Around 50% of all binary choice indicators (Price Level indicators, Market Events indicators, community choice) fell in the 35% to 65% accuracy range, clearly showing the high degree of uncertainty we saw in January. Uncertain indicators can also be of use, though it would be incorrect to use them to assess accuracy levels.

It should also be noted that our criteria for calculating the accuracy of support and resistance levels this month meant that we considered an indicator correct if at least one of the levels was predicted correctly, taking into account the actual volatility of the asset.

The overall value for the accuracy of our indicators in January stood at 62.6%, taking into account all types of indicators except ICO ranking. In spite of the fact that this figure is 10% lower than what was achieved in December, we consider this result to be relatively impressive. The main result for us was the opportunity for learning which January gave our models. Cindicator now has a good chance of surprising our users next month.

As always, we have put together a rundown of the most memorable indicators of the month for you.

Market Events indicator

This indicator regarding the capitalisation of the entire crypto market was sent on 24 January at 14:00 UTC to all token holders starting with the Beginner tier (to receive these indicators you simply need to have at least 5,000 CND tokens in your wallet).

Weekly S/R levels for fiat/crypto

One of the most interesting support and resistance levels was the one calculated by Hybrid Intelligence for Bitcoin in the midst of the market crash. The levels given (support at $6,130 and resistance at $8,800) were sent to all token holders from the Trader tier up on 5 February at 14:00 UTC.

The indicator proved to be extremely accurate, considering the huge levels of volatility seen during that period. Forecasters were able to accurately predict the minimum price level Bitcoin would reach, and even the high point it could climb back to after the fall!

As for fiat questions, look at the gold futures support and resistance levels sent on 29 January at 14:00 UTC:

Price level for fiat/crypto

One of the most lucrative indicators for cryptocurrencies turned out to be the Tron (TRX) indicator. We sent this indicator to Expert tier token holders at 14:30 UTC on 1 January, when the TRX price stood at $0.045. In three days TRX surged by more than 120% and reached its target price!

For traditional market indicators, take a look at our Facebook indicator. Facebook shares fell on 12 January, and so we sent out a question asking if it would recoup its losses in the following month. We sent out the indicator on 15 January at 14:20 UTC, and the price actually rose by 3.3% to its target of $188 in the next ten days.

ICO ranking

In January we published three ICO rankings. Usually the question is set out in such a way as to clarify Hybrid Intelligence’s predictions for the growth of the token in the three months immediately following the ICO. It is still too early, therefore, to sum up the ratings published in January (most of the companies either still have to finish their ICO, or to release their tokens).

We can however look at the results of the December ratings — the Verify and Bounty0x tokens, given top positions in the rankings, grew by 50 and 30 times their respective ICO prices, and in less than a month and a half! The Ins Ecosystem and Bread tokens, both given second position in the December rankings, grew by 4.7 and four times respectively.

We are very proud of the fact that many companies holding ICOs express their desire to use our ranking to promote their tokens. Indeed, a favourable assessment from Hybrid Intelligence increases the likelihood of a startup successfully acquiring its funding target, as more and more crypto investors begin to trust our indicators. It’d be important to notice that we define the tokens to question our analysts on the crowd forecasting platform about relying only on our internal team research. So that our research remains independent.

Development: recent updates and further plans

The whole of January was devoted to work which may be undetectable from the outside, but we managed to significantly increase the stability of all our internal systems. Besides that, in January we successfully brought the third version of our new ML model into production, and even carried out research on a fourth iteration. We redesigned the architecture for our information processing system, allowing us to take a big step forward in terms of calculation speeds. This allows us to test new models more quickly and in future will make processing large volumes of information easier.

Looking ahead to February with optimism,

The Cindicator Team

This report was created by:
Nodari Kolmakhidze, Chief Investment Officer
Kate Kurbanova, Head of Analytics

Vlad Kazakov, Product Manager