View by Day

Prof. Martin Ellison

University of Oxford, United Kingdom

Authors: Martin Ellison and Andrew Scott

Abstract: Using a new historical database derived from monthly statistics for each bond issued by the UK government we construct a series for the market value of UK government debt for the period 1694 to 2016. By constructing zero coupon prices we use the analysis of Hall and Sargent (2011) to correct the National Accounts treatment of public finances so as to consider the role of bond price fluctuations in achieving UK fiscal sustainability over the last 300 years. Over this period the government has twice achieved major reductions in government debt - in the 18th century, through a long sequence of fiscal surpluses, and after the second world war, through inflation and low bond returns. Comparing 20th century experiences the UK has used primary surpluses much less than the US and inflation much more. In general, over our whole sample bond holders have achieved a good rate of return but in the 20th century bond holders have been poorly rewarded. The two exceptions are the 1930s and the most recent period since the financial crisis. High rates of return for bond holders, especially on long bonds, mean that by the end of the sample period the market value of debt is at its highest relative to par value since the early 1700s. Because our dataset contains prices and quantities for each bond the government has issued we can perform a number of counterfactuals to investigate i) the potential importance of debt management in influencing the level of government debt and ii) how different debt management policies affect refinancing risks and the degree of fiscal insurance achieved. We find that if the government had just issued short bonds then government debt in 2015 would have been lower by nearly 50% of GDP. The over-performance of short compared to long bonds in our dataset is due to the limited fiscal insurance offered by long bonds. In particular we find that after financial crises long bonds rise sharply in value leading to large increases in government debt. This explains why current levels of UK government debt are at historical highs relative to their face value. We find that the ex post optimality of short bonds over our sample remains even when allowing for auction effects through issuance. However issuing large quantities of short debt does increase rollover risk. Taking this into account and calculating efficiency frontiers which trade off refinancing risk and the level of debt does reinstate a minority role for long bonds with an ex post optimal portfolio of 70-80% short and 30-20% long.

Full Text:  WILL BE AVAILABLE LATER