In searchofatheoryofdebtmanagement

The completemarketapproachtogovernmentdebtmanagementarguesthataportfolio of non-contingentbondsatdifferentmaturitiesshouldbechosensothatfluctuationsin marketvalueoffsetchangesinexpectedfuturedeficits.However,thisapproach recommendshugefluctuationsinpositions,enormouschangesinportfoliosforminor changesinmaturitiesandnopresumptionitisalwaysoptimaltoissuelongandinvest shortterminawidearrayofmodelspecifications.Theseextreme,volatileandunstable featuresareundesirablefortworeasons.Firstlyfragilityofportfoliostosmallchanges in assumptionsmeansthatitisoftenbettertofollowabalancedbudgetratherthan issue theoptimaldebtportfolioundersomepossiblymisspecifiedmodel.Secondlyfor even minisculetransactioncosts,governmentspreferabalancedbudgetratherthanthe large positionscompletemarketsrecommends.Thecompletemarketrecommendations conflict withanumberoffeatureswebelieveareintegraltobondmarketincomplete- ness,e.g.transactioncosts,liquidityeffects,robustness,etc.andwhichneedtobe explicitlyincorporatedintotheportfolioproblem