Sigmoidal Success Story


Annualized return


Higher annualized return compared to other benchmark strategy

6.7 mln

higher revenue in Q1 2019

What was the business problem?

Financial risk management systems specializing in specific assets are ill-equipped in analyzing complex, multi-asset class portfolios. Furthermore, many managers do not provide security-level transparency, rendering security-based solutions nonviable.

To make sound investment decisions, an asset allocator needs to be armed with an understanding of the total portfolio positioning. Conventional portfolio analytics that focuses on managers’ returns seek to approximate historical manager-to-benchmark relationships but provide no insight as to the managers’ current exposure.

Our Client, a Singapore-based Hedge fund, wanted Sigmoidal to design and develop an intelligent asset allocation system.

What savings and profits does the client achieve?

By leveraging the novel AI-based approaches, our client was able to create an investment strategy that generated an 8% annualized return, which was 23% higher than any other benchmark strategy tested over a two-year period. In addition, we were able to estimate risks, as well as the uncertainty of our estimates using a novel technique called Variational Dropout.

How did we accomplish it?

In one of our projects, we designed an intelligent asset allocation system that utilized Deep Learning and Modern Portfolio Theory. 

The task was to implement an investment strategy that could adapt to rapid changes in the market environment.

The base AI model was responsible for predicting asset returns based on historical data. This was accomplished by implementing Long Short-Term Memory Units, which are a sophisticated generalization of a Recurrent Neural Network. This particular architecture can store information for multiple timesteps, which is made possible by a Memory Cell. This property enables the model to learn long and complicated temporal patterns in data. As a result, we were able to predict the asset’s future returns, as well as the uncertainty of our estimates using a novel technique called Variational Dropout.

In order to strengthen our predictions, we used a wealth of market data, such as currencies, indices, etc. in our model, in addition to the historical returns of relevant assets. This resulted in over 400 features we used to make final predictions. Of course, many of these features were correlated. This problem was mitigated by Principal Component Analysis (PCA), which reduces the dimensionality of the problem and decorrelates features. 

We then used the predictions of return and risk (uncertainty) for all the assets as inputs to a Mean-Variance Optimization algorithm, which uses a quadratic solver to minimize risk for a given return. This method determines the allocation of assets, which is diverse and ensures the lowest possible level of risk, given the returns’ predictions.

How did we boost the project with Sigmoidal DNA?

Design Authority

Sigmoidal employed the model-centric approach, developing experimental research to improve the machine learning model performance. We selected the most applicable model architecture and training process from a wide range of possibilities, to maximize efficiency.

Product-focus Method

We used metrics definition and monitoring, which sets the ​​gold standard for monitoring machine learning models and tracking accuracy in real-time. We expanded the model performance, adding a 3+ year lifespan of the product.

Deployment Strategy

We utilized the Blue-Green deployment method driving towards the ability to rapidly put software into production.

What technical stack has been designed and implemented?