ECT's natural gas production over the past year has absolutely cratered. But so has the share price. Is there finally some value here?
Recently, I considered purchasing a set of oil and gas trusts as part of a periodic portfolio update. As part of my research, I took a hard look at the numbers; I reviewed years of SEC filings and built an engineering-caliber model to forecast trusts' future distributions based on individually-tailored forecasts of production, prices, and expenses.
This article discusses the results of the model as applied to ECA Marcellus Trust I (ECT). It includes a review of key assumptions and a summary of the fair value for ECT for various rates of return and risk. ECT's relatively young wells, irregular and selective reporting of well production, extensive use of price hedges, and previously subordinated shares make this trust particularly difficult to value; your comments and challenging questions on the assumptions are highly encouraged.
For reference, you can find a previous application of the model to Enduro Royalty Trust/NDRO here.
A General Model of an Oil and Gas Trust
Oil and gas trusts are a class of commodity investment in which an investor purchases the right to future net profits from a set of wells. ECT, for example, owns "royalty interests in fourteen producing… and 52 development… natural gas wells drilled to the Marcellus Shale formation in Greene County, Pennsylvania." (Source: ECT 10-Q filed 11/8/13)
ECT (though gas-only) is one of many such oil and gas trusts. Others include:
- Chesapeake Granite Wash Trust (CHKR)
- Hugoton Royalty Trust (HGT)
- Mesa Royalty Trust (MTR)
- MV Oil Trust (MVO)
- Enduro Royalty Trust (NDRO)
- SandRidge Permian Trust (PER)
- Pacific Coast Oil Trust (ROYT)
- SandRidge Mississippian Trust II (SDR)
- SandRidge Mississippian Trust I (SDT)
- Whiting USA Trust I (WHX)
- Whiting USA Trust II (WHZ)
The value of ECT, as with all such trusts, is entirely based on the net present value, or NPV, of its future distributions. While other methods, such as trailing yield and proven reserves, are often used as the basis for valuing a trust, these other methods may significantly mislead the investor. Only the NPV of future distributions measures what matters to the investor: it's the value of the cash they will put in your hand.
To estimate future distributions, I built a generalized model that considers forecasts of 1) gross revenues from the sale of oil and gas from properties underlying the trust, 2) expenses incurred by well proprietors, 3) the trust's expenses, and 4) the structure of the trust, which including the share of underlying revenues that is owned by the trust and the number of outstanding shares. This generalized model is shown in the two-step calculation below.
Of course, in the case of trusts, the devil is in the details. As discussed in my earlier article, "Tricks that Inflate the Value of Oil and Gas Trusts," trusts and well proprietors use a number of accounting and contractual techniques to manipulate net income and, indirectly, the distribution. ECT in particular has used many of these tricks, so it is wise to be aware of them before investing.
ECT Trust Distribution Model
To determine a value for ECT, I adapted the general distribution model discussed above to ECT's unique circumstances, used it to forecast future distributions, and applied an NPV calculation.
ECT owns royalty interests in two sets of wells, 14 "PDP" wells and 52.06 "equivalent PUD" wells. Both sets are notably young; most wells were drilled after the trust came into existence, and all are less than 3.5 years of age. As with all young wells, production is greatest at inception and quickly declines, ending in a long tail that may last decades. ECT's NPV depends highly on the level of production at which its wells will ultimately stabilize.
Given the sensitivity of ECT's production to the age of its wells, I broke down ECT's reported production numbers by well. ECT did not make this break out easy; in its SEC filings, the trust only reports aggregate production numbers for its share of production(!), buries well inception dates, and glosses over previous pipeline issues. However, with a fair amount of Excel grease, I developed two power curves, one for each set of wells, that describe the production for a well over time and that, together, comprise a "best-fit" function for historical production.
The chart below shows historical production, combined with three forecasts of future production, a "medium" forecast based on the power curves and "best" and "worst" cases that reflect a 10% margin of error. All forecasts include significant decline in well production from current levels.
ECT gas sales are largely through closed contracts. Contract terms are unavailable to the investor, but published historical sales numbers roughly reflect NYMEX Henry Hub prices. Therefore, for the dividend forecast, future gas prices are set equal to the value of NYMEX Henry Hub futures, shown below. As of 11/19/13, the price is similar to those for spot sales today and gradually moves to a long-term value of $5.34/MCF.
ECT deducts a variable cost for compression and transport that is accrued based on production volume. This charge varies by quarter and previous charges have ranged between $0.58/MCF (Q3 2010) and $0.77/MCF (Q4 2012). The dividend model assumes that the charge for the most recent quarter, $0.74/MCF, will continue in perpetuity.
In addition to the gas royalties, ECT owns a series of natural gas hedges that run through Q1 2014 at $5.00/MCF. The profits related to the settlement of the hedges are included as a separate line item in the quarterly reports. The value of the remaining hedges is calculated based on NYMEX Henry Hub futures and included through Q1 2014 (and $0 thereafter). As a result of this assumption, the dividend model forecasts that ECT's dividend will drop by 30% sequentially into Q2 2014.
ECT trust administration deducts a single quarterly charge for trust expenses. The trust's quarterly reports indicate that this charge includes constant expenses for "accounting" ($15,000/qtr), "administration" ($37,500/qtr), and trust management ($600/qtr). However, the bulk of expenses, averaging nearly $300,000/qtr, are not described. The dividend model assumes that all quarterly fees remains constant, and that the undescribed charge is taken as the average of past charges.
ECT will dissolve on or before March 31, 2030. Given the best-fit production curves discussed above, and forecasts of future sales prices and expenses, ECT will have a non-trivial residual value of between $2 and $5 per share. (My calculation assumes a 10% annual ROI on the ongoing value of annual production, a 1.5 "fudge factor," and a 50% cut taken by the trust proprietor, ECA.)
So what's ECT worth? Well, that depends on an investor's desired NPV and risk level. Using the model, I calculated the NPV of future distributions for the best, middle, and worst cases (with residual values) and using a variety of possible rates of return. The results are summarized in the following table:
As I read the table, it suggests that the current market valuation of ECT ($7.22, as I write this), assumes that production declines will tend towards the worst-case forecast and have an expected rate of return of around 10%.
That said, although ECT's byzantine SEC filings, accounting "trickeration," and recent share price catastrophe suggest that a high-risk premium is warranted, the table also suggests that the market may be overstating this risk. A more fair valuation might assume the middle case and a return of 10% or worst-case at 8%, which would place the value of ECT between $7.97 and $8.23 per share. Not only does this valuation assume continued and significant production decline (near term) and an 8-10% annual ROI, it also offers a 10% premium to the share price for the investor.
The dividend model suggest that investors may wish to consider initiating a small stake. (I did, and plan to expand it if the share price continues to drop.) Alternatively, investors might also consider waiting until May 2014, when ECT will announce its first distribution that no longer includes income from the natural gas hedges. The distribution model forecasts that ECT's distribution will fall another 30% sequentially at this time, which might cause skittish investors to abandon ECT once again, possibly yielding a buy opportunity.