Before the fact

On the 4th of September, the Ministry of Industry and Information Technology of China issued a ban on the ownership and trading of ICOs in china. This had a huge impact on the capital raise for many small businesses as well as the value of the Ethereum token. After the dust started to settle financial analysts did what they do best and wrote articles on the economic impact of this ban. Usually this analysis goes hand in hand with some nice graphs and predictions. The key word here is “after”. For an investor, knowing why they suffered a 30% loss of capital “after the fact” isn’t relevant anymore. In this article, an analysis is made through a standard Bollinger band method. Trade volume analysis is added to support our claims. This methodology is compared with the ICODAO sentiment analysis and the prediction that follow from that. We apologise that we wrote this article after the fact. 😉

Bollinger bands

The concept of Bollinger bands was introduced by John Bollinger in 1980 and trademarked in 2011. Bollinger bands are an indication of the relative “highness” or “lowness” of an assets price based on previous price observations . This is calculated in the following way:

Upper band = MA + kσ

Upper band = MA + kσ

Where MA is a N-period simple moving average, σ is standard deviation and k is a multiplier depending on the volatility of the asset. In our example, we take the value N = 15 and k=2. Data is gathered from coinmarketcap.com with a daily interval:

The line graphs are the closing price of Ethereum and its Bollinger bands, the bars are the trading volume on the given day. Each trader uses the Bollinger bands in different ways. We use it to buy when the price is close to the lower band and sell when it is close to the higher band. For our own piece of mind, we use trade volume to check the saturation of the market before we begin to act. In the graph, the first arrow is when the first buzz happened on China planning to put a ban on ICOs. The second arrow is the official statement as mentioned in the introduction. As can be seen, the price isn’t near the upper or lower band and there is no unusual trading volume, with exception of the two marked days.

Sentiment analysis

Obviously, the question is now, how the ICODAO can help with better predictions. The ICODAO does what any experienced investor would do, check all the relevant sources, check the prices and the sentiment in the market. Based on this the DAO makes an investment decision. ICODAO does this, without any inherent bias when making its decision.

In the Graph below you can see the price of Ethereum, normalised sentiment index and trade volume. The arrows still represent the first buzz on the ICOs ban by China and the official announcement. From this graph, various interesting observations can be made:

1. In circle 1 & 2 the sentiment gave a strong indication that the price would go up and later would fall. Predictions happened roughly 3 days before the event.

2. There seems to be a strong correlation between sentiment and volume.

3. After the first arrow, sentiment clearly went down while the price went up, after prices kept on rising sentiment went up. Then 4 days before the official ban, the sentiment went down again.

In retrospect, we can say that the ICODAO sentiment analysis has good indicators to predict price changes two or three days ahead of their local maximum or minimum. The correlation between sentiment and volume was to be expected because volume is a different metric for “interest”. In the specific case of the Chinese ban, we can see that investors were already anticipating this action on the 30–08–2017, 4 days before the actual announcement.

In the future, more research has to be done on this first sentiment index. In following the articles we’ll dig a bit more into questions like the impact of momentum in our sentiment index. Can we distinguish between bullish and bearish sentiment? What does positive sentiment say about people’s readiness to invest?

As of this morning (22–09–2017) sentiment around Ethereum is still relatively low. We’ll send out a tweet (follow us @ICODAOFUND) when the sentiment goes up again.