Be Careful with "Comps"

I've previously written about valuation methods, especially discounted cash flow analysis, used in the upstream oil and gas business. Today I'd like to touch on those that rely on metrics like multiples of production rate, monthly cash flow, reserves volume and acreage. These "short-cut" methods use values of these multiples from prior sales of comparable properties ("comps") to estimate the value of another property. As an example, if you know the current net production of a property and use comparable transactions to estimate the production rate multiple, then multiplying the two will give you an estimate of the property value.

Multiples from comparable properties are a handy way to get a rough estimate of value, but they have many limitations. Users of them need to understand these limitations and be careful in relying too much on these techniques if getting an accurate valuation is important.

For starters, it can be difficult to even identify "comparable" transactions. To illustrate, recall that the value of an oil and gas property is a function of several factors, including -

  • the current production rate,
  • the production decline rate,
  • the mix of products,
  • the operating cost structure,
  • the expectation of future prices,
  • the amount of undeveloped reserves and the capital needed to produce them,
  • the cost of capital, and
  • the various project-specific technical risks.

For properties to be "comparable", most of these parameters need to be the same or very similar. This usually means that the properties need to be in the same geologic and geographic setting, be at the same relative stage of maturity, and the prior transactions must have occurred in a time with similar future price expectations.

An obvious issue is that if an evaluator has all of the data necessary to decide that properties are comparable, then they probably have enough data to make a rigorous valuation using discounted cash flow analysis.  Using discounted cash flow is always preferable over multiples. And, if the properties are determined to not be truly "comparable," there is no way to accurately adjust the metrics to account for their differences.

Also, even if comparable properties can be identified, the set of them is often small and the range of multiples from them is often too wide to be very reliable. For instance, knowing that comparable properties have traded in a range of 6 to 9 times the prior twelve months' cash flow doesn't help very much in determining the actual amount to bid.

For production rate multiples to be useful, the comparable transactions must have the same decline rates, cost structures, levels of development, and the price outlook must be the same. Also, since production multiples are usually expressed on an oil or gas equivalent basis, then the product mix must be the same. If the product mix of the comparable property is projected to change over time, for instance in a field where the gas/oil ratio is increasing, then it must change in about the same way in the comparable and subject properties.

Reserves multiples suffer from essentially the same shortcomings as production rate multiples. Sometimes, evaluators will attempt to assign reserves multiples to different reserves classes, but this creates a new problem. Since it is very rare to know how the value of the prior sales was allocated between the various reserves categories, someone must first make that assumption and then calculate the multiples. Occasionally this is possible when there have been property sales whose reserves are almost all producing or almost all undeveloped, but these instances are rare. By using only these cases, evaluators risk relying on very small data sets in which outliers can have too mush influence.

Acreage multiples are very difficult to use effectively for all the same reasons as reserves multiples, but they also are made less useful since none of the factors that truly impact value are strongly correlated to acreage. One place they are useful is comparing transactions that are entirely undeveloped, in the same geologic setting, and are very close to each other. In some cases, they are used with production rate multiples, but here again lies the problem of allocating value for the comparable transactions between their productive wells and their non-producing acreage.

In my opinion, the cash flow multiple is the most useful because it reduces the impact of price and product mix differences. It is still strongly dependent on producing decline rate, level of development maturity, and risk, however.  Properties that have no undeveloped potential, with similar decline rates, in the same geologic setting and location can often be considered comparable. Another issue here is whether to use recent past cash flows or expected future cash flows.

If multiples from prior sales are so problematic, then why are they so often used?  Mostly because they are easy to calculate, seem easy to understand, and they can be used as a quick, rough indication of value. Often, it is sufficient to merely know that a property value is most likely in a particular broad range, and they are very useful for that. 

I also find them useful in comparing trends in valuation over time, and making comparisons between different geographic regions and ownership interest types. They can also be used as a check after a discounted cash flow analysis has been prepared to determine how it compares to prior transactions.  Its very important to be careful here, however, because there are often good reasons for a property's metrics to lie outside the range of the comparable ones, and even if these metrics are within expectations, it doesn't necessarily mean the valuation is accurate.

By and large, using "comps" is handy, but can be misleading.  Be sure to understand their limitations, and remember to focus on the details of each property, and they won't trip you up.