Past performance may not be indicative of future results…

by Tim Whittaker, Associate Research Director, EDHECinfra

Investing in Early Public Works: Financial Returns in English and Welsh Turnpikes 1820-1845 September 2016, by Dan Bogart, Department of Economics, UC Irvine

This paper examines the history of financial returns of privately provided infrastructure in the England and Wales.  Employing a unique sample of turnpikes, Bogart (2016) examines the risk and returns of private infrastructure and compares it to other investable assets of a similar era.

Turnpike trusts, according to Bogart (2016), were created in the UK to improve specific sections of the road network in the early years of the industrial revolution.  The trust was usually given responsibility by an act of parliament for 21 years. The author notes that it was also common for these trusts to be renewed by subsequent acts.

Trusts had the right to levy tolls on the road and issued bonds that were secured by the toll revenues.  Most acts creating the turnpikes explicitly forbade the creation of shares and retention of profits: all revenues had to be devoted to road maintenance, as well as repayment of the principle and interest of its debt.

By 1830, 20 percent of the total road network was managed by approximately 1,000 trusts according to the author.  By 1883 however, competition from railroads made many turnpikes uneconomic and a new act of Parliament led to the winding up of the remaining trusts.

As with all studies of infrastructure returns, data is difficult to obtain.  The author used information provided in response to a series of mid-19th century Parliamentary Select Committee requests: revenues, expenses, cash balances, interest payable and the amount of debt outstanding. From this information, the internal rate of return, as well as a return on capital employed for the up to 1020 trusts in England and Wales can be estimated.

The interest rate for the bonds issued by turnpikes varies between 3.44 percent and 4.09 percent for an average of 3.37 percent.  Consol bonds (government bonds) of a similar period paid an interest rate on average of 3.44 percent.  From that perspective the turnpike bonds, with the additional project risk, are not a desirable investment opportunity.  After 1848, returns on turnpike bonds started to fall, this is due to the development of railways reducing the demand for these roads and as a result, their revenues and ability to pay interest.

Bogart (2016) also finds that investors viewed newer roads as ‘riskier’ as younger trusts paid a higher interest rate than older trusts.  Furthermore, when comparing the realised returns of the bonds issued by turnpike trusts, then tend to have a lower Sharpe ratio than other contemporaneous assets.  Only farmland exhibited a lower Sharpe ratio.  The mean return for the bonds measured in 1820 was 3.70% for bonds and 4.26% for bonds measured in 1838.

This paper provides an interesting piece of analysis of early infrastructure investment. However, when Sharpe ratios are measured, the estimation procedure of return volatility for each turnpike is not clear. The author states that prices have been used from the Charity records, but does not explain the time horizon for such changes.  This reduces the value of the Sharpe ratio findings.

Turnpike investment in the 1800s seems to have been better than investing in farmland, but compared to other infrastructure like assets such as canals and railways, they seem to have been a poor long-term investment choice.

The paper can be found here.