Derivatives are used to hedge against uncertainty in macroeconomic factors influencing project cash flows, in particular interest rates, exchange rates, and to a lesser extent CPI and sometimes fuel prices.
Projects make use of such hedges to immunise their cash flows against adverse changes in the underlying factor, thereby decreasing the volatility of cash flows available for debt service and allowing the project to leverage higher than they would otherwise be able to.
Corporates and Parastatals engaged in infrastructure projects also make use of hedges if they have significant borrowing programs, and more commonly, to lock in exchange rates when they purchase significant capex in a foreign currency.
Hedges, however, come at a cost, with the result that the banks offering them to the projects earn a considerable return. In particular, with the implementation of Basel III, and the Credit Risk Transfer (“CRT”) requirement, also called Counterparty Valuation Adjustment (“CVA”), spreads have widened considerably.
As an indication of how profitable a hedge may be to the investor, banks recognising PnL from derivatives in their trading books up front may earn as much as 2.0% – 3.5% of the notional debt amount on day 1 of the trade. (This is because profit and loss on the trading book is measured as the change in the mark-to-market, or NPV, of each derivative in the book.
Spreads on commodities (fuel) and on CPI are significantly wider than on interest rate swaps and fx swaps, with CPI swaps in particular being of interest because of the investor demand for CPI-linked debt.
Hedging instruments incur both market risk and credit risk. Market risk can be hedged by the investor in turn by entering into an equal and opposite position with its own trading counterparties. However, credit risk is a function of the unknown, but statistically predictable, future fluctuations in the underlying macroeconomic factor(s). Unlike the derivative trading activities which an institution typically enters into with banks, derivatives with a project will not be governed by a collateral agreement or CSA, with the result that the credit exposure can spike. Helpfully, hedges typically rank similarly to the tranche of debt which they are hedging, and are therefore typically senior in both the pre- and post enforcement payment waterfalls. This results in a probability of default (“PD”) and a loss given default (“LGD”) very (very) similar to that of the senior debt. However, the exposure at default changes in accordance with a probability distribution function, as touched on above.
Hedging derivatives can create a significant return kicker for traditional debt instruments, and may constitute an opportunity for a debt investor in that:
- The work to determine the project’s credit quality may already have been done to evaluate more traditional investments; and
- The credit exposure can become negative if the underlying moves in the “right” direction, leading to an implied funding benefit.
The investment banks typically reserve credit lines to cover credit risk associated with a hedging instrument provided to a project in the amount of the potential future exposure which will not be exceeded with a 98% probability, sometimes referred to as the swap’s PFE. This is calculated quantitatively by running, for example, 10 000 monte carlo simulations on the underlying asset, for example the interest rate curve. For each simulation, the hedge is revalued at each duration. The exposures at each duration are then sorted in, say, descending order and the 200th largest exposure at each duration is plotted to form the PFE curve. Similarly, expected exposure can be modelled by using the 5000th (or median) exposure at each duration. This process is relatively straight-forward to carry out for interest rate swaps using every-day tools, provided that the original market data is available and the investor is comfortable to rely upon simulated spot rates and forward curves. The same exercises for fx swaps would be more challenging because of the need to maintain a realistic arbitrage-free forward exchange rate curve in simulations.
Using market data available at the time of writing, the PFE for a 12-year vanilla amortising interest rate swap was approximately 15% of the original notional loan amount and decayed rapidly towards zero at later durations as the number of remaining payments contributing towards the NPV reduced.
More details on project finance hedging to follow in later posts…