Dev Update – 1st Sep 2020

Orbit Products

Here is the 1st development update on Sep.

The last two weeks are mainly for market sensing algorithm and mercury dashboard. To get details, please check the following.


Market Sensing Algorithm

  • Market Sensing Algorithm
  • There is no panacea for market sensing algorithm. So we have focused on the anomaly detection algorithm. The big financial event like the Corona pandemic leaves footprint before happening.
  • We tested several algorithms for the purpose so far and decided to choose probabilistic models rather than deterministic ones because it provides a solid background.
  • The probability distribution model is the core factor in the probabilistic one. Alpha, beta, gamma are names of well-known distributions.
  • We have tested alpha and beta distribution for the market sensing with few indexes and found out that it is good, but we are unsure whether it is coming from overfitting.
  • We need more proof in it for the effectiveness of those distributions.



  • The dashboard will present the charts telling the current status of asset management in a broad view. Orbit will manage various kinds of asset classes and many accounts, so we need a way to check Orbit’s overall status quickly and intuitively. The dashboard comes into the picture; by checking critical data through a chart, we can sense how it is going.
  • A portfolio chart has been implemented, which shows balances data, including actual and virtual accounts under management.
  • PnL(Profit and Loss) chart is done. We can see how much money we make or lose day by day.

Asset Class Classification

  • Asset’s valuation is deeply bounded to its characteristics, so it is a good source of data to improve the machine learning model’s accuracy by telling the asset’s category.
  • We have started asset classification for that purpose. It is a painful and lengthy process because everything will be done by humans, not by machines.
  • But we firmly believe that this information is likely to increase the accuracy of the model, we decided to dive into it.

Get early access to the latest updates on the MoA

Get early access to the latest updates on the MoA Investment Solutions