Investing Is Hard.
Most asset managers require an investment of more money than 95% of Americans have.
Even worse, investing in a quantitative fund is nearly impossible. The investment requirement is more money than 99.9% of Americans have, or the fund is so successfull that it is closed to additional investment.
Trading Pipeline
Retrieve Data
Each day, our rebalancing process begins with data collection. Our universe is the S&P100, so we get updated pricing for each of those stocks.
Once we've recieved the data, we then use a prediction process that employes a technique called Multivariate Singular Specturm Analysis. This works by first using a Singular Value Decomposition process to filter out unimportant price movements, then using linear regression to predict on this transformed data.
Price Predictions
Portfolio Optimization
We then pass the output of the MSSA, along with a covariance matrix associated with the previous year daily price movements of the S&P100 stocks into our optimization program.
This program utilizes Quadratic Programming to find the set of weights that give an optimal risk-return trade off for a certain risk threshold.
We currently bound our weights between 0 and 10%, meaning no short-selling and no over concentration of a single security.
After optimal weights are created, we send our orders over to our broker to place trades. We currently use Alpaca as our broker. Liquidations and sells are placed first to ensure no liquidity issues.
Trade Placement