Efficient Allocation of Risks in Infrastructure Projects

One of the basic premises of risk management is that risks should be assumed by whoever is most efficient in its assumption.  Even if this seems to be quite obvious, it is not always the case, and sometimes not following this rule has important consequences.  Let’s analyze the case of large infrastructure projects in which a usually powerful party, usually a government or a large company, requires several sellers to submit an offer for the construction of an infrastructure project like a bridge, an airport, a building, etc.  These projects usually involve a multistage procedure, in which the perception of risks is not always symmetric across all the parties involved in the transaction.  This can cause inefficiencies.  I order to illustrate this issue we will describe the process and the risks involved in it.

Assume that the government is requiring the construction of a bridge across a river.  As a first step several companies submit technical proposals. Only a few of them are accepted and are required to submit a financial proposal for the completion of the project.  Let’s assume that all the proposals have to be received by March 1, 2016, and the government will make a final decision of which seller will be awarded the project by November 1, 2016.  By March 1st each seller has to submit a financial proposal for the construction of the bridge, in which –since the government does not want to bear any risk- they are required to offer a fixed price for the completion of the bridge.  By doing this, the government knows the final price of the bridge at the moment of receiving the offers.  But this certainty comes at a cost.  In any project there are risks to be borne and these do not disappear, they are just translated from one party to another.  In this case, the risks that are not assumed by the buyer are going to be assumed by the seller.  This is not necessarily a problem if the seller is able to assume these risks efficiently, but if this is not the case and inefficiencies generate a cost, then the society as a whole will pay for it.

To illustrate the nature of this problem lets continue the discussion of our example.  Let’s assume that the construction of the bridge needs the purchase of several tons of commodities, just for the sake of discussion assume that 100 tons of copper and 100 tons of aluminum are needed for this particular project, and as we all know these commodity prices are volatile (their annual volatility can be well above 30%).  Given that the government is not willing to assume the risk of these volatilities increasing the price of the bridge, the seller is required to provide a fixed price for the whole project as of March 1st.  This fixed price is essentially transferring the exposure to commodity prices volatility to the seller.  The seller that ends up winning the bid and getting the concession, in November will have an open short position in Copper and Aluminum, since they will have promised to build a bridge for a given price, that will have a given assumed Aluminum and Copper price. This short position is easy to hedge using several hedging strategies involving the purchase of derivative instruments in the financial markets or with suppliers or using operational hedging strategies.  The problem is that, in order to have an open risk position that can be hedged, the seller needs to know that it has been awarded the contract, which might not alway be the case in a competitive bidding process.  The company that is awarded the contract will have a short position in November, while the other competitors –the ones not getting the deal- will not have a short position.  So, by March 1st nobody has a hedgeable short position; by the time they submit their financial offer all the sellers have a contingent short position, that will become a short position for the winner of the project by November 1st.  The problem is that between March and November commodity prices can change significantly, leaving the sellers exposed to potentially devastating volatility.  In other words, the problem with this approach is that the sellers are forced to assume a contingent risk position, that is essentially impossible to hedge between March 1st and November 1st.  A company participating in the offer for the bridge cannot hedge the position in March because if they get the deal in November, then the hedge purchased in March will offset the November exposure, but if they do not get the deal, their hedge purchased in March will not match any exposure, resulting in a risky position.

In order to overcome this problem firms willing to participate in the deal have three possible strategies: (i) increase the offer price in order to have some of this potential volatility mitigated by a “price cushion”, (ii) assume a serious risk of losing money in the project, and (iii) not offering to participate in the construction of the bridge.  None of these alternatives is particularly appealing for the potential participants in the offer, for the government and for the society as a whole.

The problem is that contingent risk positions cannot be hedged.  Fortunately there is a solution for this problem; a solution that does eliminate this non hedgeable risks.  The only party that has a certainty that will be exposed to these commodity prices fluctuations is the government. The government knows that in November they will grant the construction of the bridge to one of the sellers.  So, and based on this certainty, they government should change the bidding process and require the sellers to offer the bridge according to a price equation that allows commodity prices (in this example, just aluminum and copper prices) to vary before November 1st hence firms will not need to include a contingent "price cushion" to account for future volatility.   In other words, the government should ask the sellers to offer something like:

Bridge Cost= 100 tons of Aluminum at Nov 1st spot price + 100 tons of Copper at Nov 1st spot price + FIXED AMOUNT OF $$

As we can see, this procedure forces the government to face an exposure to the volatility of aluminum and copper prices.  More specifically, the government is taking a short position in both commodities between March 1st and November 1st.  But, unlike in the case of the sellers, this position is not contingent, hence can be easily hedged by the government.  In order to hedge this position, the government could buy futures in the financial markets to cover 100 tons of each commodity, with expiration date November 1st.  When the government awards the project and assigns the contract to the winner, the Bridge Cost shown in the previous equation will be calculated as a fixed amount of money since by then the commodity prices as of Nov 1st will be known.  If November's commodity prices result higher than March's prices, the cost of the bridge will be higher than expected but the futures on the commodities will be sold at a profit that will exactly compensate the loss, and if November's commodity prices happen to be lower than March's prices, the bridge will be cheaper but the losses in the Futures contracts will wipe out the profits.  In other words, the government will be able to keep the cost of the bridge at the same level expected in March.

This approach requires that the entity that is requesting the project generate a risk map of the project and then decide which party is better equipped to assume each risk.  In the example we discussed above, it is easy to hedge the risks in the financial markets, and the only issue was which risks where confirmed and which where contingent, but in some other cases the problem might be that one party is better equipped than the other to assume a given confirmed risk because of the firm’s structure, informational advantage, financial structure, etc…  By following this approach projects should be more efficient from a pricing standpoint, with lower levels of extra costs and should attract more firms to the deal.