Investment Analytics

Traditional Investment analytics have relied singularly on market data (mostly drawn US equities research) and linear mathematical models.  However, Sapiat believes that these tools can be enhanced significantly with the application of new machine learning methods that take into account non-linearity, non-stationarity, and fundamental data sparsity.  The next generation investment analytics will incorporate more dynamic, sensitive, and robust approaches to everything from dimensionality reduction (e.g. factor extraction) to optimization and allocation under changing scenarios.


Risk and Alpha Analysis

Separating and decomposing the systematic and idiosyncratic components of investment performance has been a fundamental preoccupation of modern finance theory.  Using contemporary machine learning approaches help to better separate and estimate the factors that lead to commonly recognized systematic returns and their residuals which may or may not be related to other sources of return.  While alpha is an elusive concept, our analytic toolbox helps to reduce estimation error and remove bias in the estimation of risk premia from a return stream.


Performance Attribution

Sapiat’s architecture delivers transaction-based performance attribution that fits the fund manager investment process. Combined with Sapiat’s High Performance Computing Scenario Simulation this module allows asset owners to analyze each fund manager’s investment style and how it fits into longer term asset allocation.


Scenario and Allocation Exploration

The high-dimensional exploration of multiple scenarios has posed various computational challenges.  However, the advent of high performance computing (HPC) and the use of alternative data couple with machine learning allow us to expand our ability to number-crunch and simultaneously perform dynamic conditioning to reduce the need for the overly burdensome simulations.  Particularly, the appropriate modeling of tail risks are of fundamental importance to longer-horizon investors.  Clients are able to explore the full distributional implications of various allocations and  scenarios via a visually compelling Dashboard which allow close-to-real time analysis and debate  of various investment decisions.


© 2019 Sapiat Pte. Ltd.