This thesis considers the link between financialisation and emerging market government policy autonomy. It analyses the government bond markets of three case study countries: Brazil, Lebanon and Turkey. Using extensive interview data in the three countries, and interviews with financial market actors in London and New York, the study explores the investment behaviour of a range of investors: commercial banks; individual investors; mutual funds; pension funds and hedge funds. The thesis uses the framework of financialisation – measured by the ability to trade risk – to analyse both international and domestic investors. The study shows that increased financialisation, of both financial market actors and the structure of government bond markets, generally serves to reduce loyalty and therefore reduces government policy autonomy. However, it is demonstrated that initial financialisation – the development of pension and mutual funds – serves to increase autonomy. This is captured by the construction of an ‘autonomy curve’. The conclusions suggest an updating the use of Hirschman’s concept of voice, exit and loyalty in the analysis of financial markets, to give a greater emphasis on loyalty and to include the use of ‘disloyalty’, the ability to short securities. It is also argued that financialisation is the appropriate framework to analyse processes of change in financial markets. The thesis also makes observations as to the true extent of government policy autonomy in emerging market countries, and policy recommendations regarding those governments’ attitude to financialisation.