Proof of Performance — Prototype of the world first blockchained index publisher, Part 2

Napoleon Group™
Proof of Performance
5 min readNov 5, 2017

As explained in a previous article, the blockchain, and Ethereum in particular, has an interesting role to play in an existing valuable business: index publishing.

What is Index Publishing?

Index publishing is the action to create and publish an index, from a trading strategy (or not but let’s skip this part), that represent the (theoretical) financial performance of such strategy.

As an example, the S&P500 is a famous index computed by S&P Dow Jones Indices a division of S&P Global. It follows a set of rules written on a methodology document published by S&P Dow Jones Indices and that explains very precisely how it is built and maintained. Basically, it has been designed to reflect the U.S. equity markets and, through the markets, the U.S. economy. In practice, it gives a set of weights on a group of US stocks, so that the resulting portfolio should represent a fair view of the U.S. equity markets according to its creator.

Example of the index representation of the S&P 500 for the past 10 years as of Nov, 3rd, 2017.

Proof of Performance and Replication

But what is very important is the fact that any index may be replicated. As explained in the methodology published by S&P Dow Jones Indices :

“The indices should be fair, meaning that an investor who buys all the stocks in an index with correct index weights can achieve the same performance that S&P Dow Jones Indices calculates.”

Saying so, S&P Dow Jones Indices guarantees that any investor able to trade the stocks following the set of rules disclosed in the methodology document, will achieve the same performance as calculated by them. Doing so, S&P Dow Jones Indices, as any other Index Publisher, guarantees two things about such computation: rightness and trust. Both can be brought by the blockchain and that’s why we did it at NaPoleonX!

It may be notable to say that replication may be affected by volumes. We have analyzed this topic in a dedicated article. That is also why, we focus on very scalable investment solutions, based on low to medium frequency trading algorithms, as explained by Stéphane Ifrah, NaPoleonX CEO, in a short video:

Why Low Frequency Trading ? By Stéphane Ifrah, CEO of NaPoleonX

Indices of daily trading strategies

At NaPoleonX, we have developed the world first Index Publisher based on the Blockchain. It may compute and publish quantitative trading strategies (aka trading bots) making a daily trade at closing, the more liquid time to trade. It implements the first example of Proof of Performance based on the blockchain. On the first version, only mono-asset trading strategies are available and for signals belonging in {-1 (short), 0 (neutral), 1 (long)}.

Computing the index of a mono-asset daily trading strategy from the signal it provides before the closing, is fairly simple. If Si is such signal, if Vi is the variation of the underlying asset at date i, and if Ip is the value of the Index for the previous business day, then Ii, the value of the index at date i will be : Ii = Ip x (1+Si x Vi). The complexity lies much more on the proof that Si was known by the strategy provider before the closing time Ti of the underlying asset on date i. This can be solved by sending such signal Si on the blokchain before Ti. For obvious reasons, such signal shall be obfuscated and revealed after a certain period of time called latency.

The following demo-video present our prototype and explains how we (crypto-)prove that such signal Si was known before Ti, whereas still be able to compute the index (with a latency):

Video-demo of the world first Index Publisher, designed by NaPoleonX

Basically, the demo is in two steps:

  1. Before Ti, NaPoleon’s oracle compute a random string Ri and concatenate it to the signal Si. Then it does the hash (through SHA256) which gives Hi. And then it publishes Hi before Ti, in a first transaction Tx1.
  2. After some day (latency), NaPoleon’s oracle publishes the signal Si and the random string Ri. At this stage, the smart-contract can compute the value of the index for the date i, and anyone can verify that the signal Si was known before Ti. Indeed, by computing H2, the SHA256 of the concatenation of Ri and Si, everyone can check that H2 = Hi (if not, there would have been a trickery from NaPoleon’s oracle). As there would have been impossible (from the features of SHA256) to find a random string R2, so that with another signal S2, the SHA256 of their concatenation would have given Hi, this is a (crypto-)proof that Si was known before Ti.

The example shown on the video can be summarize by the following picture:

Crypto-proof of a trading signal

This Index Publisher constitutes the first brick of the NaPoleonX platform.

NaPoleonX will be the first 100% algorithmic crypto asset manager. It is based on three pillars:

• A team of ex-trading professionals having managed several billions, used to go thorough due diligence of large investors, and able to get/acquire an Asset Management license from a EU regulator,

• A focus on low frequency algorithmic trading of highly liquid assets only, to achieve high scalability, low execution risks and best performing techniques. Our team has developed this expertise over the past 10 years more and we currently have than 20 performing absolute return algorithms,

• A technique to tokenize quantitative strategies trading on real assets, so that performance and liquidity are brought back to the crypto-world, enabling to offer highly scalable investment solutions providing a return on your BTC or ETH.

ICO will be launched in december with a pre-ICO next week. Go on https://www.napoleonx.ai to be sure not to miss it!

Please note that citizens and residents of several countries may not participate. As an example US persons are not allowed to participate.

--

--

Napoleon Group™
Proof of Performance

Fully algorithmic, scalable & decentralized crypto asset manager piloting trading bots, founded by former multibillion-dollar managers. https://www.napoleonx.ai