Resource Companies: NPV vs. Transaction Values

In this tutorial we looked through the construction of various economic studies, at their main components, and why they are usually a more accurate way of estimating the value of an asset compared to in-ground resources. One of the key components of economic studies is the Net Present Value (NPV). This is a single number - that uses a range of assumptions - that is intended to reflect the combined value of all future cash flows from an asset, in terms of what they are worth today. However, whilst the NPV is a useful method in determining the value of a project, it is not necessarily the best method. In many cases, real transaction values are more useful; after all, an asset is worth what the market is willing to pay for it.

The Limitations of the NPV figure

The limitations of the net present value are inherent in that they are because of the assumptions that are involved in arriving at the figure in the first place. These assumptions we use when creating the NPV figure are generally less accurate in the case of resource assets too, given their lack of predictability in many cases. Consequently, sale prices of assets can vary substantially from their touted NPV values. A range of limitations of NPV are below:

- It is based on a pricing assumption. For example, an oil and gas asset NPV may be based on an oil price of $100/barrel. If the oil price was to rise or fall significantly, so will the NPV value as resource companies NPVs are highly sensitive to their commodity price. Given that oil and gas assets usually have a lifespan of many years, the pricing assumption is likely to be inaccurate, especially if the underlying commodity price is volatile.

- It is based on output, operating expenditure and capital expenditure assumptions. These are difficult to predict too, particularly the former two figures, and particularly for oil production. This is because issues can arise, such as the need to purchase new equipment to tackle production issues. Unlike mining a mineral resource, which is fairly predictable since you can map areas of high and low grade, oil production is also more difficult to forecast given that it is not always easy to forecast production decline.

- It is based on a discount rate. We base a NPV calculation on a discounted annual cash flow. The discount we apply is ultimately subjective.

- Non-numerical factors are ignored. Since the NPV is a single figure, it will never capture all non-numerical factors that influence a project. For example, the precise location of the project and how appropriate the location is, cannot be included in the figure.

- It is based on a resource assumption. The NPV uses a resource assumption to estimate how much of a certain resource is extractable. Whilst this can overestimate the resource commercially extractable, it cannot include resources not yet known about or discovered. Therefore it can just as easily underestimate the commercially extractable resource.

- Unforeseeable events cannot be included. Similarly to some non-numerical factors, it is impossible to foresee some future events. For example, although Canada is seen as an extremely attractive environment for mining companies, there are no guarantees that the Canadian government will not hike taxes on mining companies in the future. That would significantly impact upon a NPV figures. Other examples may include events such as natural hazards, which could impact upon future production.

- It does not account for the Opportunity Cost. The opportunity cost of developing an asset is that of the next best alternative and the benefits that could have been derived from that second best project. This is especially important for resource companies and here is why. Most start off with an early stage asset and develop it towards production. Once they near production, they may produce an attractive NPV. But is that the best NPV on offer, or are there many other projects on sale for lower prices that are more attractive? That would impact upon the market value of your project and what's more, you are usually bounded to your project since you have already spent vast sums of cash on developing it to the NPV stage.

Two NPV vs. Transaction Value Examples

It is important to not take NPV values at face value if possible, and to correlate them with real transaction data. Whilst we will briefly look at that concept in the next section, we will look a two examples where transactions of an asset took place at values substantially different from their net present values. Notice that this applies to both pre-production and production assets.

Example 1: Hummingbird Resources & their purchase of the Yanfolia Gold asset

This is a straightforward example where we can clearly see that in June 2014 - during a recent period of depressed gold prices (when the gold price was ~$1,150/oz -> $1,250/oz), Hummingbird managed to purchase the Yanfolia gold asset for $20m paid in shares. However, at $1,250/oz and an 8% discount rate, the gold asset has a stated NPV of $65.5m. Therefore, Hummingbird was able to purchase the project at a near 70% discount to the touted post-tax NPV.

Example 2: Aminex & their sale of various US oil and gas production assets

In April 2011, junior oil and gas exploration and production company Aminex released a resource and reserves report for three properties in the US where they outlined their resource potential and corresponding NPV figures. These numbers were as per the table below.

Notice that the table above is split into two sections; 1P = proved oil and gas resources, 2P = proved + probable oil and gas resources. Ultimately, the total 1P value - at a 10% discount rate - of the three properties was touted to be $46.4m, or $135m at the total 2P value, also as a 10% discount rate.
It's fair to say that these figures were far wide of the mark when it actually came around to selling the assets between August 2013 and May 2014. Let's start with the South Weslaco property, which was said to have a 1P value of $1.56m two years prior to its sale. After factoring in resource depletion over the two years, the sale value of $0.45m is not catastrophic as a percentage, but it's certainly not close to $1.56m

However, this was a far larger percentage than was achieved for the Shoats Creek and Alta Loma properties. These were said to have a combined 1P NPV of $44.88m or a combined 2P NPV of $133.02m. Before there are any thoughts as to whether the oil price collapse dented the eventual sale price, this idea can be dismissed. The US crude oil price was broadly flat between 2011 when the report was produced and the second sale in May 2014; in fact the sale was a couple of months prior to the start of the sharp oil price downturn of 2014.

So how much those two properties sell for?

The two properties sold for just $5m

Moreover, only $500,000 of that $5m was upfront, with the remainder likely to take years to recoup. Even being generous and using the $5m figure, that is a near 90% discount to the 2011 1P NPV or a whopping 96% discount to the 2011 2P NPV. Being less generous and using the upfront $500,000 figure and those percentages drop to discounts of 98.9% and 99.6% respectively. That's a far cry from what some investors may have been expecting based on the NPV. The reasons for the low price included poor production figures, Aminex being in a weak negotiating position and little buying appetite for onshore US producing assets, at the time.

The discounts to NPV can be extreme and it is worth acknowledging that. What may appear to be look like a seriously valuable asset for a resource company, can quickly be revealed as a far less valuable asset simply because of the intrinsic problems that a NPV figure has.

What sort of Transaction Value models can you use?

The most important part of using historic transaction value models is that you find a close analogue to the company you are looking at. By that, you should use another project/asset that is very similar to the one you are looking to value. It may be similar in location, production, resource size or a range of other ways. It is important to recognise immediately that the problem with transaction value models is that it's incredibly difficult to find a perfect analogue, hence discrepancies can arise.

Once you have found a comparable set of figures, the main figures to compare are often:
- Output of the resource (e.g. gold ounces per year or oil barrels per day)
- The financial details of the above output (e.g. operating costs, breakeven levels)
- The in-ground resource (e.g. the total sale value of the comparable company / total resource, and use this number to calculate a pro-forma resource figure for the company you are trying to value)
We will look at these in future tutorials.

It is abundantly clear that the NPV of a project or asset is not necessarily even close to its tradable value. It is often useful to use transaction values to construct a better 'likely value' framework for a project. Although NPVs remain useful tools when a comparator transaction cannot be found, it is worth bearing in mind that the market rarely pays the figure as the NPV in an economic study and to not simply assume that the net present value of an asset is what the market should value it at. In the vast majority of cases, especially pre-production, it will not come close.

Using real tangible and forecast cash flows for producing assets, and discounted NPVs or transaction values for non-producing assets (where possible) should be preferred in most cases.