The NAV (Net Asset Value) of your DAF token would factor in this mechanism on a daily basis like normal traditional funds work, through the use of provisions, that can increase or decrease depending on positive or negative performance. Hence, when it is paid / detached, this has already been factored in. The real advantage we are seeing in this mechanism is that we avoid ETH transfer, hence risk, and the DAFs NAV do not change when these tokens are issued.

We have not designed specific arbitrage bots so far and have been more developing market timing bots. We do believe that on a long term basis, there should not be any arbitrage left when the crypto market will mature. But this would be a cool and almost risk free way to make money given the large discrrepancies between exchanges at the moment.

We have only initiated conversations with several exchanges at this stage and not conclude any agreement so far. Our wish is to have at least 2 exchanges for listing our tokens to provide some security nets and more convenience, bearing in mind that not only will we want to list NPX token, but also all the DAF tokens.

As far as I understand this project, they're goal is to gather, organize and provide data for algo trading but not provide actual money making strategies. To me, and without deep knowledge of their actual value proposal, they could be some kind of Bloomberg / Reuters for crypto assets, which would be great news for us as we would get raw material to design algos.

I have benefited, and still continue to do so, for years. However managing money is one thing, building a business around this capacity is another. I was lucky enough to work in an Advisory firm (EY), hopefully not too long, to get this business approach. I was also lucky enough to meet a few people of our team there with a great entrepreneurship approach. Launching this business is really the opportunity to write history in a brave new crypto world and contribute to this differently than just scalping the markets.

Our own existing bots do not currently use AI / Deep Learning as we have had a more physicist approach in our R&D. However, we are familiar with Machine Learning techniques as both Arnaud and myself met at an Executive program around Datascience at the Ecole Polytechnique a little almost 3 years ago. As long as it makes money, is systematic, has a low frequency trading and few operational associated risk, it would qualify for us.

It is true that you have a lot of projects that are promising to manage money for investors. Which is kind of amusing as the crypto community was kind of proud to have escaped the banking industry and one of the most crowded segment is in "finance".

One of our main differentiator is that we are proposing a library of existing algos and to monetize them through the launch of an asset manager to benefit from diversification. Also most of our algos will be getting their performance in the traditional asset space but with a very limited currency risk through our currency hedging strategy described in our Blackpaper.

Our intention is to have a dual listing whenever possible for all our tokens. We have initiated conversations with different exchanges but have not concluded any agreement yet. Of course, should we reach such agreement we would communicate on this.

You are raising an important point here on overfitting. We are very acute of any such risk and the strategies presented in our Blackpaper are for a large part the same strategy applied to different and unrelated markets. This is one of our selection criteria when we select an algo. We want it to be applicable to different markets so that we believe that we are closer to the modelisation of human behavior than overfitting.

Thank you for your kind words. We want to release some of our platform features in the next few weeks, before the ICO is officially launch so that people can judge by themselves on our delivery capabilities. You should have access to our strategies through this directly. Of course the actual trading strategy code will not be open source ...

Most of other funds charge:
1. upfront fee to subscribe to a fund. We have seen some fees being as high as 10% potentially
2. management fees as a fixed fee on asset under management. We have seen level of 2%, and has been a benchmark for traditional hedge fund, even though they have come under serious pressure in this industry
3. performance fee: paid out of positive performance only. We have seen 20% level as is usual for Hedge fund industry there too
4. Transaction fee: not completely sure what it covers

We are proposing:
1. performance fees only to align investors and asset managers interest. We have set the level @25% of positive performance using a high water mark feature. If performance is more than 8%, then we are earning more than the 2% management fee + 20% performance fee above. We are true believer in attracting volumes by aligning interest between all participants and making things simpler. 

No there is definitively a confusion with our Gas concept and we will change it. Here the gas = 25% of positive performance paid by the DAFs for having the right to use a performing bots. This is some kind of energy that is requested to make the DAF work. Any idea about rebranding is welcome. Someone mentioned oil.

Thank you for the tip on oil, we might give you some authorship rights on this. We could also call it petroleum, gasoline, electricity or whatever form of energy. But we take this confusion seriously and will try to change this denomination to avoid confusion.

You have a lot of ways to build a trading strategy depending on your background:
1. discretionary trading: people rely on your ability to read tea leafs trade after trade
2. AI: it is a systematic approach where you feed your data into a blackbox and get a trading decision in return. You need to train them with data set and so to work very hard on these data set. These need to be trained regularly from experience. When do you decide to train them again is more art than science.
3. traditional algos: they are based on market observations and a representation that you may have on the markets. This is very close to a physicist approach. These algos need also to be monitored on a regular basis as market dynamics may change over time.

You are right that our background is from non crypto trading, but to be frank, crypto trading is pretty recent. We are used to generate performance using one asset and to export it into another currency ("quanto trading") and that is essentially what we would be doing here using a "currency hedging strategy" described in our Blackpaper. However, we reckon that this crypto world will have its own dynamics and partnerng with longstanding crypto traders would open up possibilities for us. We are eager to learn every day. We have also designed a simple market timing algo between BTC and USD that is working quite well with a trade every 6 weeks on average.

Reading at all the comments, you are right that there is a confusion. Here gas = 25% performance fees = right by DAFs to use trading algos. Performance engines (=algos) need gas to function.

You are right. Our gas is the cost charged to the DAFs to use our trading algos. It is calculated as 25% of the positive performance and is calculated and paid on a monthly basis. Diversifying the DAFs, hence the trading bots, allow to get more regular revenues to NPX token holders.

No, you should not use gas or charge when your bot is making a trade otherwise we would be incentive to overtrade to earn more. That has been prohibited for traditional asset manager and rightly so. The gas concept could be replaced by performance fees, ie fees that are only payable if the bot has generated a positive performance for the investor. In this case, 25% of these performance would be paid to NPX token holders (85%) and Napoleon Crypto (15%)

Traders on their trading desks used to develop a lot of little Excel macro i order to have some decision making tool. This enabled them to get positive expectation when entering a trade and were repeating such trades so that probabilities would work. These were not true leaf reading people but just good statisticians. A little bit like professional gambler. We are just trying to industrialize this process and gather a large library of such strategies

The gas concept is more related to cost payable by DAFs to use the algos that would be used for them as if you needed petroleum to make them work. So it is definitively not he same gas as the one used to validate a block. Maybe that is a bit confusing, we reckon.

We are here to be as transparent as possible but we reckon we may have a different approach to providing investment vehicles than other projects.
1. anyone participating in our ICO would contribute to the launch of an asset manager that would only use algos to manage not 1 but a series of funds.
2. all the proposed algos would seek to provide absolute performance and try to track an index
3. all these funds would be distributed to investors holding cryptos and would be denominated in cryptos, run by smartcontracts
4. Napoleon Crypto, would be the entity advising the asset management entity and executing / implementing execution bot on behalf of this entity. Thus it would need to have a proper licence to operate
5. the asset management entity would be represented by the NPX token holders. They would be the ones taking strategic decisions, based on Napoleon Crypto advices.
6. The NPX token holders would get the bulk of the performance fees paid by the different funds (DAFs).
7. As these performance fees would be paid in DAFs tokens for operational issues, they would contribute to the creation and growth of a diversified portfolio that could be managed through a platform that will launch alongside our project to host several functionnalities (ex: mgt of tokens, voting sessions, index publishing for algo strategies ...)
8. The NPX token holders would also have a privileged access to DAFs launch should they decide to invest

I hope it clarifies a little our project