LQM Registry / optimization
Constrained Mean-Variance Optimizer
Long-only capped portfolio optimisation using projected-gradient optimisation on expected returns and covariance.
Model key
lq.optimization.mean-varianceVersion1.0.0
Runtimepython-process
Model contract
Inputs are explicit.
{
"type": "object",
"required": [
"expected_returns",
"covariance"
]
}Output contract
Outputs are explicit.
{
"type": "object",
"required": [
"weights",
"expected_return",
"variance",
"volatility",
"objective"
]
}Default parameters
{
"risk_aversion": 5,
"max_weight": 1,
"iterations": 5000,
"learning_rate": 0.05000000000000000277555756156289135105907917022705078125
}Example input
{
"expected_returns": [
0.1000000000000000055511151231257827021181583404541015625,
0.08000000000000000166533453693773481063544750213623046875,
0.040000000000000000832667268468867405317723751068115234375
],
"covariance": [
[
0.040000000000000000832667268468867405317723751068115234375,
0.01000000000000000020816681711721685132943093776702880859375,
0.005000000000000000104083408558608425664715468883514404296875
],
[
0.01000000000000000020816681711721685132943093776702880859375,
0.0299999999999999988897769753748434595763683319091796875,
0.0040000000000000000832667268468867405317723751068115234375
],
[
0.005000000000000000104083408558608425664715468883514404296875,
0.0040000000000000000832667268468867405317723751068115234375,
0.0200000000000000004163336342344337026588618755340576171875
]
]
}