Topics: Project Finance / Interest Rate Risk / Foreign Exchange Risk

Financial models are the backbone of project financing, enabling the banks to check the borrower’s strength and monitor compliance with the various covenants throughout the financing period.

They also allow the project sponsors to measure and stress-test their internal rate of return and validate their equity investment. They are audited and enable the parties to enter contractual commitments with a reasonable degree of confidence.

As the investment decision relies heavily on the financial model, it is essential for the model to make it possible to both identify the key assumptions regarding outcomes and measure the outcomes’ sensitivity to these assumptions and their interactions.

What are the risks of IRR variations linked to exchange rates, inflation and interest rates, once correlations have been taken into account ?

Where certain risk parameters are not subject to hedging, the justification of the parameters used for the base case and of the capacity of the model used to stress-test the assumptions in order to measure their sensitivity is essential. However, a model based on significant market assumptions (e.g. interest rates, inflation, exchange rates) sometimes fails to provide a sufficiently clear view of the degree of uncertainty of the base case if the assumptions are not fulfilled during the life of the project.

What form of hedging would be optimal for these risks? How does this hedging affect the expected average IRR and how does it reduce the uncertainty of the investment’s IRR? What other risks does the hedging introduce? How should the implementation of options-based hedging (such as caps or swaptions) be taken into account in the IRR analysis?

Identifying and reducing financial risks in advance

This is no small matter: it concerns essential aspects of the return vs. risk analysis, and the fact is that traditional financial models are often inadequate when it comes to measuring these risks and the sensitivity to them.

Market risks need to be identified in advance of the project just as industrial risks would be. Identifying them too late will make them harder or more expensive to manage.

The first step in carrying out this risk analysis is of course to study the project’s financial characteristics: possible borrowing currencies, capex currencies, future income currencies, whether the income or capex will be index-linked (contractually or economically) to inflation, the fixed/variable rate mix, constraints imposed by the lenders, refinancing and liquidity risk, the capacity of the project to set up hedging instruments, limitations on hedging instruments, etc.

Once this analysis has been performed, the study of the sensitivity of the project IRR to the various underlying parameters will throw light on the risk management strategies that can be implemented.

Let’s take two examples to illustrate this point:

Example 1 : A motorway project where changes to the tolls are economically indexed to inflation. The project contracts 100% of its debt at a fixed rate, either by choice or on the lender's insistence.

• Risk: Inflation is lower than represented in the model. Project income increases less quickly than modelled but the cost of debt does not change.

• Result if the risk materialises: Destruction of the value of the equity, restructuring of the project (equity cure, cash sweep etc.).

• Possible solutions: Borrow more at a variable rate, hedge the inflation risk to assets, or a combination of the two.

• Expected outcome of the model: Challenge the assumptions behind unhedged variables, perform stress testing and guide the sponsors in their decision.

Illustration : Optimizing the interest-rate hedge ratio for a project with inflation-indexed income

Analysis : Due to the positive correlation between interest rates and the inflation rate, and in the absence of inflation hedging, 100% hedging of interest rate risk is not the most effective solution for reducing uncertainty about the IRR.

In the context of this project, the rate hedging ratio that minimises uncertainty about the IRR lies in the 50%-70% range. It is therefore strongly recommended to negotiate with the lending banks to ensure a minimum hedging ratio of between 50% and 70%.

Example 2 : An infrastructure project whose income is denominated in Brazilian reals but whose debt is in euros because the sponsors decided that a loan in BRL was either too expensive or impossible to raise given the maturity of the loan.

• Risk: The BRL depreciates faster against the euro than represented in the model.

• Result if the risk materialises: Destruction of the value of the equity, restructuring of the project (equity cure, cash sweep etc.).

• Possible solutions: Borrow partially or entirely in BRL, use cross-currency swaps, hedge all or part of the expected income with foreign exchange futures or options over part or the whole of the project’s duration.

• Expected outcome of the model: Challenge the assumptions behind unhedged variables, perform stress testing and guide the sponsors in their decision.

Illustration : Impact of currency risk on the IRR of a foreign-currency project

Analysis : Unless hedging is used, the exchange rate risk on the IRR is greater than the project risk (business plan unfavourable compared with the base case). If currency hedging is used, the risk on the IRR is essentially linked to the project risk. It is therefore advisable to set up the envisaged currency hedging.

The limits of deterministic modelling

Without becoming excessively sophisticated, it is important for the model to take account of the asymmetries and convexities intrinsic to the project’s structure.

Let’s take a simple example in which a project decides to hedge its interest rate risk with a cap rather than a swap. If the model means that the project will both pay a cap premium and pay the ceiling rate without modelling the possibility of the project benefiting fully from a reduction in rates, the solution of cap hedging will always be deemed to be too expensive and less effective than swap hedging.

However, with hindsight, given the almost continuous fall in rates for more than 10 years, hedging with caps would have been much more advantageous for the project than hedging with swaps that made it impossible to benefit from falling rates.

How many projects have deprived themselves of a rate reduction of 3% or even more because they baulked at paying a cap premium of 0.6% per year ?

During the decision-making process, did the model represent scenarios in which rates fall properly, or only the cost of the cap in addition to the swap rate?

A simple stochastic valuation module (Monte Carlo simulations) can represent these valuations and the financial risks with sufficient detail and robustness to support decision-making about the structuring of financing and the benefits of using hedging instruments to address market risks.

The model will also make it clear whether the reduction in risk around the central scenario is accompanied by a reduction in the IRR in the central scenario and in what proportions.

It is clear that financial models are rarely designed to perform these analyses in a relevant and exhaustive manner and that certain projects are particularly vulnerable to assumptions that are made statically in the financial model; it is not easy to understand the significance of these assumptions, the rationality of the assumptions used in the central scenario, or the magnitude of the risk associated with them.

The sooner this modelling exercise is carried out, the more likely it is that the right questions will be asked, and the project will be structured so as to minimise costs and risks. Even for simple projects, a deterministic financial model which only considers forward scenarios without being able to correctly represent shifting asymmetries, options and scenarios only provides a partial and sometimes truncated picture of the expected return on equity. The marginal cost of improving models is reasonable and the increase in visibility can be considerable.

Conclusion

Whether within or outside the financial model, it is possible to take account more accurately of the different scenarios for changes in unhedged variables and to demonstrate the potential value to the project sponsors of structuring and hedging solutions.

Even aside from unhedged variables, decisions about how to index and model different variables can have unsuspected significance. For example, an indexation factor in a multi-year programme, an external reference for fixing an exchange rate, a given inflation factor, a method of extrapolating market data if a price is not directly listed are all factors that can have a decisive influence on project profitability. The model must make it possible to understand the importance of these variables and the significance of the risk that they may shift before the closing of the deal or during the life of the project.

ESTER can provide expertise in these additional analyses of financial models, enabling projetct sponsors to make their decisions with even more confidence.