This post outlines the valuation pitfalls that need to be avoided when competing for the acquisition of a new generic Top-Level Domain (gTLD).

The most widely used financial tools to determine the economic viability of a capital investment project, such as bidding for a new gTLD, are the Internal Rate of Return (IRR) and the Net Present Value (NPV), which measures the expected additional value a project would create if undertaken. On the other hand, the Return on Investment (ROI) primarily measures the performance of financial assets over a specific period of time. I focus on the NPV, which is the most widely used and appropriate for valuing new gTLDs.

Below are the valuation components that must be taken into account when deciding on the acquisition of any asset, including gTLDs.

1. There are two independent biases in estimating revenue that need to be avoided. The first is a general cognitive bias and the second arises when there is heavy demand for an item being sold, i.e., in “hot markets”. This can represent the second round biddings for new gTLDs, with each being unique, i.e., a scarce item. Under such circumstances executives and consultants seeking to acquire the item often overvalue such items.
2. The riskiness of the cash flows that a gTLD is expected to generate is not necessarily equal over the wide range of industries representing the gTLDs. Risk affects NPV decisions, whereby the higher the risk, the lower the NPV of the cash flows.
3. The risk tolerance of gTLD applicants varies between those of new entrants and incumbents. For new entrants, the investment to acquire a new gTLD represents a strategic decision, while it is a tactical decision for the incumbents. Strategic investments entail the possible loss of the entire investment.
4. When estimating revenue, real options tend to be ignored. Real options represent future opportunities that might arise due to the ownership of an income generating asset such as a gTLD. The value of these options needs to be included in the final NPV calculation; otherwise the benefits of a gTLD would be underestimated. One example of such options is the possibility of a registry deciding to become a registrar at some point in the future and/or a registry opting to invest in future gTLDs.
5. NPV does not always lead to a black and while decision: buy, don’t buy options. The inputs into the NPV formula always have uncertainty and include measurement errors. Thus, one solution is to assume a range of outcomes for each input variable, then estimate the probability distribution of NPV. Hence, for the same probability distribution different people can come up with different acceptance decisions based on their risk aversion.
6. Applicants for a particular gTLD have access to the same information and capabilities, save some variations due to:
   1. Differences in applicants’ country tax rates,
   2. Portfolio risk diversification considerations,
   3. Experience in running a registry, and
   4. Having a better vision and value estimation of what the real options may be.
   Nevertheless, these differences might not be enough to have a big impact on some of the gTLDs, and thus, auctions can misallocate resources, a phenomenon I have called allocation myopia. (One option to avoid this path is for ICANN to consider allocations via lotteries to qualified applicants.)

Before you start bidding, you must have a maximum bid in mind, which is independent of your competitors’. Without sticking to such a price, you can easily end up with a winner’s curse, an auction situation whereby the winner ends up paying in excess of the value of the acquired object. However, although you may have a pretty good idea of what the value of the gTLD may be to your competitor, your decision should be independent of such a value. Nevertheless, your bidding strategy depends on the type of auction being held, i.e., whether an ascending English auction (a la eBay) or a descending Dutch auction or other bidding mechanisms.