Not only were the prices paid for the French concessions significantly lower (on a price-earnings basis) than the prices for the U.S. concessions, but also the spread of the bids received for each of the French roads was much narrower than the bids received for the US roads. One fact that makes this outcome particularly interesting is that there was significant overlap among the bidders for the French concessions and the US concessions as shown in table 7.
(Insert table 7 around here)
To explain the large variation in the bids for the US concessions, one needs to look at the key assumptions that the bidders made in order to construct their respective valuation models. These assumptions are: (a) Traffic growth; (b) Toll schedule (c) Operating costs; (d) Discount Rate.7 It is reasonable to conclude by observing the very small spread between the winning and second place bids in the French auctions that the assumptions used by the bidders were similar. Conversely, the large spreads in the US auctions were likely caused by disparate assumptions.
There are several reasons why there were different distributions of assumptions in the French vs. the U.S. concessions. First with respect tolls, the toll setting formula for the French concessions is based on a fixed percentage (70%) of CPI after an initial period of years during which the actual tolls were established by the French government. Given the fact that the European Central Bank has an established policy of maintaining CPI at 2%, there is very little opportunity for the bidders to have different projections of the tolls that can be charged.
In both Skyway and the Indiana Toll Road, tolls can be adjusted annually by the greater of 2%, change in CPI or change in nominal GDP per capita. Further, there is significantly more volatility in the US indices as shown in the graph 1, below. A likely consequence of the resulting uncertainly is widely divergent views on the part of bidders with regard to the future levels of these indices and, in turn, the projection of toll rates.
(Insert graph 1 around here)
With respect to traffic projections, the bidders had access to historic and projected traffic data prepared by or for the public owner. It was at the discretion of each of the bidders to adopt the projections or to prepare revised projections. In the case of the Skyway it is known that the winning bidder commissioned its own traffic study that projected dramatically higher usage of the road, as shown in table 8.
(Insert table 8 around here)
Using Skyway again as an illustration, the winning and second place bids were $1.83 billion and $700 million, respectively. The $1.1 billion difference can be explained by the winning bidder using more aggressive traffic projections, higher toll rates or a lower cost of capital, or a combination of these.
Although we do not have inside information about the assumptions that underpin the respective bidders' valuation of the concessions or the rationale used by each bidder to make their assumptions, there are two plausible and complimentary explanations for the winning bidder of the US concessions to be an outlier. The first of these theories is the Winner's Curse,8 which observes there is a tendency for one or more bidders in an auction to overestimate the value of the asset being auctioned. The Winner's Curse suggests that there are a number of possible reasons for this overestimation-both conscious and unconscious, rational and irrational.
The second possible explanation is that in a market that is the early stages of privatization, bidders are willing to make abnormally high bids in order to achieve a leading position for subsequent privatizations (Miralles, 2006). Certainly, the privatization markets in France and the US are in very different stages. Private tollways in France are well-established and represent a mature investment opportunity. In fact, no significant privatizations are expected in the foreseeable future. In contrast, privatization of toll roads is an emerging market in the US, and most industry observers expect that there will be significant privatization activity over the next several years. Skyway and the Indiana Toll Road have played the role of 'opening' the US market, and winning these concessions has given Macquaire and Cintra visibility, and positioned them as leading firms in this new area.9
In addition to the different nature of the two markets, there is a procedural reason why the bidders for the French concessions were not as susceptible to the Winner's Curse (or at least to the same degree as the bidders for the US concessions). In the French process there is the requirement for the bidders to disclose their assumptions to the government as part of the bidding process. We expect if bidders know that the referee of the auction will review and evaluate the reasonableness of the financial and operating models as part of its decision-making process, bidders will be more circumspect with respect their assumptions and less prone to overestimation.
There is an additional observation that can be made about respective outcomes of the French and US privatization approaches. The Skyway and Indiana concessions were structured to maximize the concession price. This was done by establishing the bid parameters with the objective of generating the highest possible discounted cash flow and by deciding the auction only on the basis of price. With respect to the latter, this "high bid" protocol provides the opportunity for the Winner's Curse phenomenon to play out with the consequence of a possible over-valuation. The French approach, on the other hand, is not structured to maximize the concession price.
___________________________________________________________________________________________
7 As discussed previously, the appropriate discount rate is the weighted cost of capital that is used to finance concession price. In the absence of leverage constraints, this cost of capital is a function of the bidders' perception of the risk of the project; the higher the risk, the higher the cost of capital. This assumes that all of the bidders have essentially the same access to the capital markets.
8 The concept of the Winner's Course was first discussed in Capen, Clapp and Campbell (1971). Thaler (1988) contains a useful explanation of this concept and its applications.
9 Even if Macquarie and Cintra "over paid" for these concessions, their downside risk, from a corporate perspective, is minimal. This is due to the relatively small size of the deals and the fact that in the case of Skyway, they were able to structure the financing so that it shifted much of the financial risk of the enterprise to the lenders through the use of leverage, bond insurance, and liberal dividend provisions.