Binance Russia AMA with CIO Nodari Kolmakhidze
Last week, Cindicator’s CIO Nodari Kolmakhidze joined an ‘ask me anything’ or AMA session with Binance Russia, one of the exchange’s fastest growing communities.
The session was conducted in Russian, so today we’re sharing a writeup, edited slightly for brevity and clarity.
Please introduce yourself. What do you do at Cindicator? What is your background and how did you enter the world of cryptocurrencies?
My name is Nodari Kolmakhidze. My background is purely financial – I studied at the Financial University in Moscow, where I got both my BSc in Financial Markets and Financial Engineering and my MSc in Financial Markets and Financial Mathematics. My whole life has been spent in finance and the markets.
Before Cindicator, I worked at a hedge fund, we were doing event-driven trading, which is trading based on the news and various events. While I was working there, we covered pretty much all kinds of events: financial, political, and economic. In 2016, three years ago, I moved to Cindicator to be one of the leaders of the newly created financial markets department.
My first superficial immersion in crypto was in 2012. That was when I first heard about Bitcoin as an oddity, some kind of “internet money” and so on. It was valued at $10 back then, maybe a bit more or a bit less.
The first serious foray was together with our company, Cindicator, in 2017. That’s when we had our token sale. Actually, it was even earlier – we had wanted to tokenise our internal economy for some time. Since 2017 I’ve been full time in the digital asset world. My knowledge has greatly increased since then. And the market as a whole has developed a lot.
What is the goal of Cindicator? What problems does it solve?
Cindicator is developing a Hybrid Intelligence ecosystem. Our goal is simple: we want to change the way people work with uncertainty, forecasting the future and responding to future challenges.
We’re building Hybrid Intelligence, and it’s called ‘hybrid’ because there are people (who forecast) and machines (which help to predict the markets).
In short, our goal is to create Hybrid Intelligence to help answer questions, helping investors and traders to act amid uncertainty and serving as an additional source of insights about markets.
Cindicator was founded four years ago and has passed through several stages of development. It started out as a form of new media. The founders were fascinated by ‘wisdom of the crowd’ technology. You’ve probably heard that a group of people can forecast more accurately then the best experts. One of the first researchers of this theory was Francis Galton in the 19th century. He had an experiment where at a county fair a crowd was asked to guess the weight of an ox. After collecting the forecasts, he found that the average estimate of the crowd, which included both people with no education and some ‘professionals’, was quite accurate. The error was less than 1%.
Our founders wanted to use this approach to create a media platform of a new type. For example, let’s take a certain event – some elections being held in a given country. By asking the subscribers of this platform, it’s possible to predict how the event will unfold, what the next developments could be and so on. This gamified platform could end up being interesting for users. That was the first stage in 2015.
As with any startup, Cindicator has passed through several pivots.
In 2016, after the founders met with hedge funds in New York, they thought that it would be good to initially focus this technology on financial markets. That made sense because the financial markets provide feedback very quickly, helping to judge the accuracy and value of forecasts. It’s a lot more efficient: we can also forecast the winners of the Oscars, but to get five data points, we’d need to wait five years… In financial markets everything is a lot more dynamic and we can train our algorithm faster.
Around the same time we understood that just processing collective forecasts was not enough, so we added machine learning to process these predictions – that’s how we came to Hybrid Intelligence.
In 2017, the third major step was the tokenisation of our ecosystem. Forecasters started to receive rewards in CND tokens for forecasts, which were used to provide indicators for token holders – traders who wanted to get the indicators. It’s a kind of a circular economy: traders and investors pay for tokens from our company and we redistribute the tokens to forecasters. As a result, forecasters have both monetary and non-monetary motivations, which increases the accuracy of forecasts.
What are the goals for the near future? What news should we expect?
Hybrid Intelligence is our core technology and we will keep working on that. We anticipate that the accuracy and volume of indicators will increase. All else being equal, the community will evolve to be more accurate. Our machine learning models will also improve with more data.
We have ambitious goals – we want to greatly increase the number of analysts on the platform. Right now we have over 130,000 analysts and we think this could increase to a million or millions. This would be very interesting. It could be a different scale – not yet a state, but something equally large. This would be a large ecosystem with a lot of people engaged in analysis, making forecasts and receiving rewards.
In terms of more technical things, we’re also expecting to release a trading platform that could help hedge funds and individual investors trade on exchanges. It’s a system that takes prices as inputs along with other data, for example from Cindicator, and transforms them into trading strategies executed on exchanges. It will mostly likely be made available only to hedge funds as it would be a high-end, valuable product.
Cindicator completed its token sale almost two years ago. Were the stated goals reached?
Firstly, I’ll talk a bit more about the token sale. It was not just a fundraising process but also a way of organising the ecosystem to make it valuable to all participants.
We thought about our own token way before the 2017 ICO hype. Of course our early ideas were a lot different from what we have right now, because our experience was limited and there were far fewer cryptocurrencies and projects with tokenised ecosystems. We thought it would be a demi-state that would issue its own currency and pay it out to analysts as rewards. But as you know, we now have a fixed emission that was set during the token sale two years ago. Since then no new tokens have been created. The supply is fixed and tokens are circulated within the system.
Our token sale was different. We were among the first (if not actually the first) to introduce a system of whitelisting. Some of you might remember that those days were known as the age of ‘gas wars’: whoever paid the highest gas fees for a transaction got into the token sale.
We didn’t think that was the way to go for us. Instead we asked every token holder to fill out an application, telling us what they could contribute to our ecosystem and to themselves as part of the ecosystem.
It helped us to build a healthy community that is still active, rather than just attracting token flippers who wanted to speculate and sell right away. Almost two years later, people are still engaged. We collected about $500 million worth of applications. From that we picked token holders for just $15 million. I think that was the correct decision for greater decentralisation and participation.
This was quite different from what others were doing. There’s even a funny story: our core development team was based in St Petersburg in a tiny office. As one of our developers said later, he was sitting and coding in a really small room, crammed with 10–12 people. In an even smaller meeting room next door a co-founder was on a call with potential contributors: “What, $5 million? No, we can’t sell you that much.” I think that this strict selection was sensible in the long term.
Are the indicators we pay for really that valuable?
That’s probably a question from one of our analysts.
In short – yes.
Cindicator is not a holy grail where you can just take indicators, start trading and become a millionaire. These things don’t exist.
Yet Hybrid Intelligence indicators indeed have value, which is unique because nobody can copy it. To recreate it, someone would need to develop the same ecosystem, gather 130,000 forecasters, and build over 50 machine learning models. In theory, it’s possible to create this system from scratch, but for us it took four years. It would take at least 2–3 years for a potential competitor to build something like that. That’s why this source of data is unique.
In the markets, if everybody knows about your source of data it stops working. This is because the markets are leaning towards efficiency and all signals and patterns are priced in by all traders and the market doesn’t react to them anymore.
In the case of Hybrid Intelligence, the tokens are limited, so the number of people who could receive indicators is limited. This means that access to alpha is limited and its value is retained.
We also trade based on these indicators. First it was a test, then increased volumes. Indeed we see that they make money if you know what to do. Of course, you need to do additional research and select indicators that really fit your strategy, using Hybrid Intelligence indicators as a supplementary data source. For example, we have internal traders who use Hybrid Intelligence to help determine the current trade, whether it’s bullish or bearish. By adding this layer to the process of identifying the trend, a trader increased their ROI by about 30%.
Token holders also find it valuable. You can see in our Telegram chat how people use indicators. For example, it was interesting for me to see how some members of our community use indicators to trade stock options. Their results are very good. Several members are doing that and one of them even publishes the results. Personally, I don’t really like those indicators, but if someone can make money with that it’s great.
Whose algorithm is winning? Yours or the United Traders’?
I don’t know about the United Traders algorithm, but we have some strategies that beat all benchmarks in crypto: the indices and Bitcoin.
So, investors who want to get exposure to cryptocurrencies have several options. They can either buy and hold some assets directly, or they could buy and rebalance manually. One of our strategies is rebalancing weekly and this significantly improves performance compared to just investing in Bitcoin or one of the broader indices.
What are the roles of advisors in your company? Or were their photos included just for the token sale?
Any company that conducted a token sale, especially towards the end of the boom, indeed had temporary advisors for a month while the sale was ongoing.
Some of our advisors are still active and we’re really thankful – they help with business development, introductions, advice on markets, legal questions, new ideas for ML models and more. So I’d say that most of our advisors are still active. It’s true that some of them are less active, but that’s the nature of any tokenised company.
How are you different from REP?
REP is Augur – they’re doing prediction markets. With Augur you can make a bet on some event happening. For example, you can make a prediction that the Federal Reserve will increase the rate next month. You can buy or sell a share in the event. If the event happens, you get a reward. If it doesn’t happen, you lose your bet. Essentially, it’s a market for events.
With Cindicator, our forecasters and our analysts don’t lose money if they forecast incorrectly. Instead they can win or lose points. At the end of the month, we finalise a rating and our forecasters get rewards in CND. So there is no financial component, no gambling where you can lose money. With us it’s purely forecasting. If you’re correct you can get rewards, but you can’t lose anything.
Also, in Augur there is an exchange, as well as market-making strategies that aren’t connected to forecasting an event.
What are the advantages of the token?
CND is the only way to access Hybrid Intelligence indicators. We have some access tiers. If you have a certain number of tokens, you can access the corresponding number of indicators.
For example, at 700,000 CND tokens you can get access to Expert level indicators, one of the highest tiers, including price indicators. For example, we might ask whether Bitcoin is going to increase to $10k next week. You get an indicator with a certain probability. Currently this happens via Telegram. Knowing the long-term accuracy (right now it’s about 62%, which is high), you can use it in your strategies.
What are the criteria for evaluating the project’s success? What would indicate that the goals have been reached?
You can review the white paper, which was published ahead of the token sale in 2017. We’ve reached all of the targets, except for the one that’s planned for this year – the trading platform, which is ready. The release is several months away, as there are still some things to finalise.
Is there a plan to move to the BNB blockchain?
Strategically, there is a possibility that Cindicator might need to move to another blockchain. This would depend on how far we need to scale. We’re currently on the Ethereum blockchain as an ERC-20 token and at some point Ethereum might not be enough.
We’d then evaluate the possibility of moving to another blockchain and would also look at other methods of using blockchain in our projects. For example, our internal analysts are currently formulating the questions. It’s possible that in future we could give this opportunity to some external oracles to create or influence the questions by holding a certain number of tokens – that’s just a hypothesis, and isn’t even in our plans.
BNB could be one of the options. We don’t really like hype or making rash decisions. Right now BNB is in demand from companies that are on Binance. We’ll decide based purely on functionality and how it could help our project.
What do you want to ultimately achieve with the project?
We want this new format of interaction within the ecosystem to tackle big global challenges that are not solvable right now: environmental issues, the global North–South divide, even space exploration.
Who are your future users?
I’d say it’s mostly institutional players, the funds that would use indicators and the trading platform. But our current active traders would still remain our core audience. We’re trying to make a range of products for different levels of capital.
Has anybody at Cindicator bought a Lambo?
No, not yet.
Is it possible to sell the same indicators to several funds?
That’s a very good question.
Indeed, this could be an issue. Funds typically want exclusivity. Currently, we’re sending all indicators to all token holders. But there is a high probability that we might create exclusive indicators in the near future. They would be more expensive than regular indicators, whether it’s decided by some auction system, a revenue share or something else.
How does the team work to arrive at a final decision?
Since our project is based on the ideas of decentralisation, we also use this in making decisions. We don’t have a single authoritarian organ. Our team has different views and it’s very interesting to arrive at consensus among a large pool of people. Possibly that’s why our decisions are better compared to a situation where a single individual, even the smartest one, is making all the decisions.
Do employees get paid in tokens or as bonuses?
There are cases where employees get bonuses in tokens. Of course, nobody is getting their salary in just CND.
How could Cindicator help crypto reach mass adoption?
Forecasters have financial and nonfinancial motivation. The get rewards in crypto. An analyst who has made accurate forecasts for a month would climb the rating and would get a certain number of CND tokens.
That could be someone’s first step in the crypto world if this person just found our app in the store and got interested.
Our app is fairly captivating, and even if they haven’t had experience with crypto they’ll be nudged to create a wallet in order to receive the tokens. For some countries that could be a major push towards adoption.
What difficulties did you face after the token sale? How did you tackle them?
There were plenty of difficulties. In particular, there was a certain mismatch between the desires of some token holders (a vocal minority) and our plans. For example, the first official listing – on Binance by the way – was two months after the token unlock. Of course, flippers didn’t like that.
Our goal was not to list on an exchange just to let people speculate. CND is a utility token that provides access to indicators. So once we completed development, we opened access and got listed. Of course, some ‘serious investors’ wanted the listing immediately to get their 2x and forget about the project.
I’ve read that the core business of Cindicator is in applying Hybrid Intelligence in the financial sector and banking and not in trading?
Actually, that’s not correct – our core business is in creating indicators that could be used for profitable trading. In a few years we might move to some other use cases, possibly making financial decisions within a company. That’s not our current focus – currently we’re about reducing uncertainty in trading financial assets, including crypto assets.
Should we expect any interesting partnerships or updates in the near future?
Yes, there is something big coming fairly soon: can’t exactly when, but it could be a matter of months. Hint: it will show the practical value of Hybrid Intelligence.
How would you explain Cindicator to someone who has never heard of it?
I’d say it’s a symbiosis of humans and machines for forecasting the financial markets.
How many analysts do you have on the platform?
We have 130,000 analysts who have registered and made forecasts. You can also do that, we have apps in Apple’s App Store, Google Play, and a web app that you can use on desktops.
As I’ve said, we have about 50 machine learning models that process the forecasts. Models look at different features of analysts’ behaviour. Some are simple, giving greater weight to analysts who are higher up in the rating. Others are more complex and look at how the analyst answers the question. We’re tracking all models, and some are working better than others – that’s normal. Of course, the best models are used for indicators that we send to token holders.
What would Cindicator 2.0 look like? How much time would you need for it?
An answer to this is coming up very soon. It’s connected to the upcoming update that I mentioned. Let’s wait and see.
I’ve noticed that some forecasts were marked as ‘financial’. Are other types coming up, for example political or tech forecasts?
As I said, we’re focused on financial markets because it’s easier to get data, train technology, and monetise it.
It’s not out of the question that we might look at other use cases after reaching the full potential in finance. We already include politics if it affects the markets.
What returns do you guarantee?
We don’t. Nobody can give any guarantees.
Over the past few years, historical accuracy is at about 62% across all indicators. We’re trying to reach a balance between accuracy and value. It’d be easy to forecast where Bitcoin might be in 10 seconds, but you can’t make money with that. We’re focusing on forecasts for several days to weeks, which is where we have a stable accuracy and it’s possible to be profitable.
How much money is left from the token sale?
That’s a trade secret that we won’t share. We probably have more funds left than 90% of companies because we carefully managed all our positions. If you look at treasury sales statistics, most teams sold at the bottom. We haven’t done that and I’m really happy about it.
What other markets apart from cryptocurrencies can you analyse?
We can quickly move into new niches, for example in cyber sports assets, by attracting analysts who are already experts in those markets. Our team won’t need to learn new skills, so this would be an advantage.
You’ve said you don’t like certain indicators? What could be changed to improve them?
To clarify, I was talking about a strategy for options based on indicators for US company earnings. But overall I like the indicators – I can see that we have alpha.
If analysts sell their tokens that they receive as rewards, how could you be confident that supply and demand would balance?
Not all analysts move to fiat. Some view CND as investment and hold it. Some forecasters are returning the CND back into the system by gaining access to indicators.
It’s not possible to be absolutely sure of anything, but I believe that demand will equal supply because we moved to CND rewards in September 2018 and haven’t observed a single case of any influence on the price of the token on days when rewards are paid out or even several days after that.
Are you planning to create a robot trading platform based on indicators?
Yes, this already exists as a trading platform. It’s unclear right now how access would work – it would probably be a more complex system.
That’s it for now!
If you have any other questions, please ask them in our Telegram chat and our team will get back to you.
Make forecasts with Cindicator:
- Improve analytical skills;
- Stay up to date with market movements;
- Win part of the monthly prize of 1.25 BTC and $7,500!
130,000+ analysts already signed up!