org.ojalgo.finance.portfolio

Class MarketEquilibrium

java.lang.Object

org.ojalgo.finance.portfolio.MarketEquilibrium

public class MarketEquilibrium
extends Object

MarketEquilibrium translates between the market portfolio weights and the equilibrium excess returns. The only things needed to do those translations are the covariance matrix and the (market) risk aversion factor - that's what you need to supply when you instantiate this class.

This class performs unconstrained optimisation. For each set of asset returns there is an optimal set of weights. It also performs reverse (unconstrained) optimisation producing the "optimal" expected returns given a set of weights.

The name MarketEquilibrium is actually a bit misleading. By altering the risk aversion factor this class can/will describe the weights/returns equilibrium for any investor.

Author:

apete

See Also:

calculateAssetReturns(BasicMatrix), calculateAssetWeights(BasicMatrix)

Constructor Summary

Constructors

Constructor and Description
MarketEquilibrium(BasicMatrix covarianceMatrix)
MarketEquilibrium(BasicMatrix covarianceMatrix, Number riskAversionFactor)
MarketEquilibrium(String[] assetNamesOrKeys, BasicMatrix covarianceMatrix)
MarketEquilibrium(String[] assetNamesOrKeys, BasicMatrix covarianceMatrix, Number riskAversionFactor)

Method Summary

Modifier and Type	Method and Description
BasicMatrix	calculateAssetReturns(BasicMatrix assetWeights) If the input vector of asset weights are the weights of the market portfolio, then the ouput is the equilibrium excess returns.
BasicMatrix	calculateAssetWeights(BasicMatrix assetReturns) If the input vector of returns are the equilibrium excess returns then the output is the market portfolio weights.
static Scalar<?>	calculatePortfolioReturn(BasicMatrix assetWeights, BasicMatrix assetReturns) Calculates the portfolio return using the input asset weights and returns.
Scalar<?>	calculatePortfolioVariance(BasicMatrix assetWeights) Calculates the portfolio variance using the input instrument weights.
void	calibrate(BasicMatrix assetWeights, BasicMatrix assetReturns) Will set the risk aversion factor to the best fit for an observed pair of market portfolio asset weights and equilibrium/historical excess returns.
MarketEquilibrium	clean() Equivalent to copying, but additionally the covariance matrix will be cleaned of negative and very small eigenvalues to make it positive definite.
MarketEquilibrium	copy()
String	getAssetKey(int index)
String[]	getAssetKeys()
BasicMatrix	getCovariances()
Scalar<?>	getRiskAversion()
void	setRiskAversion(Number factor)
int	size()
BasicMatrix	toCorrelations()

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

MarketEquilibrium

public MarketEquilibrium(BasicMatrix covarianceMatrix)

MarketEquilibrium

public MarketEquilibrium(BasicMatrix covarianceMatrix,
                         Number riskAversionFactor)

MarketEquilibrium

public MarketEquilibrium(String[] assetNamesOrKeys,
                         BasicMatrix covarianceMatrix)

MarketEquilibrium

public MarketEquilibrium(String[] assetNamesOrKeys,
                         BasicMatrix covarianceMatrix,
                         Number riskAversionFactor)

Method Detail

calculatePortfolioReturn

public static Scalar<?> calculatePortfolioReturn(BasicMatrix assetWeights,
                                                 BasicMatrix assetReturns)

Calculates the portfolio return using the input asset weights and returns.

calculateAssetReturns

public BasicMatrix calculateAssetReturns(BasicMatrix assetWeights)

If the input vector of asset weights are the weights of the market portfolio, then the ouput is the equilibrium excess returns.

calculateAssetWeights

public BasicMatrix calculateAssetWeights(BasicMatrix assetReturns)

If the input vector of returns are the equilibrium excess returns then the output is the market portfolio weights. This is unconstrained optimisation - there are no constraints on the resulting instrument weights.

calculatePortfolioVariance

public Scalar<?> calculatePortfolioVariance(BasicMatrix assetWeights)

Calculates the portfolio variance using the input instrument weights.

calibrate

public void calibrate(BasicMatrix assetWeights,
                      BasicMatrix assetReturns)

Will set the risk aversion factor to the best fit for an observed pair of market portfolio asset weights and equilibrium/historical excess returns.

clean

public MarketEquilibrium clean()

Equivalent to copying, but additionally the covariance matrix will be cleaned of negative and very small eigenvalues to make it positive definite.

copy

public MarketEquilibrium copy()

getAssetKey

public String getAssetKey(int index)

getAssetKeys

public String[] getAssetKeys()

getCovariances

public BasicMatrix getCovariances()

getRiskAversion

public Scalar<?> getRiskAversion()

setRiskAversion

public void setRiskAversion(Number factor)

size

public int size()

toCorrelations

public BasicMatrix toCorrelations()