But there have been significant structural changes over the past few years, and we’ve changed our mind. Using horizontal drilling and fracking, drillers have basically learned how to squeeze oil (and gas) out of rock. They found a lot of these oil-rich rocks in places where there was little easily obtainable oil and thus little existing refining capacity: in North Dakota. Oil production in the Bakken Shale region of North Dakota has soared from 400,000 barrels (bbl) per day in 2011 to 1 million bbl per day in 2013. North Dakota now produces more oil than does Alaska and is second only to mighty Texas, where oil production has tripled, to 2.1 million bbl per day. (Read more: "Energy Independence in the U.S. ")
Since crude oil is useless to us and our cars — it has to be refined, and it doesn’t magically teleport itself to refineries — it is important to understand our existing refining infrastructure, which has developed over a long period of time, based on production and consumption patterns. Historically, about a third of imports arrived from Canada and Mexico and 40 percent came from OPEC countries, though most U.S. suppliers were outside of the Persian Gulf (Venezuela, Nigeria, Algeria and Angola). Half of the U.S. imports came to the Gulf Coast region, which also accounted for half of the total U.S. refining capacity.
The East Coast has imported mainly refined petroleum from Canada, Russia and Europe, because of a lack of cheap refining capacity (and high-cost, unionized refineries). Midcontinent North America has the least amount of refining infrastructure, but that is where a big chunk of new oil is coming from.
There are some additional oil industry constraints. U.S. law prohibits export of unrefined oil. Therefore, this new oil cannot be shipped to the coasts and put in tankers and sent overseas; its flow needs to fit into our existing transportation and, even more important, refining infrastructure. That is why the importance of refineries has grown so much. Building a new refinery costs five times more than adding additional capacity to an existing one. We probably will not see a lot of new refineries built, however — nobody wants one in their backyard.
We were looking at how to capitalize on the Bakken oil boom. The benefits to railroads had already been priced into their stocks. Makers of railcars and barges looked interesting but were threatened in the long run by the pipelines that are coming online in 2014 and beyond. Refineries are the biggest beneficiaries. There were many to choose from, but we put our money where our mouth is and bought Northern Tier Energy (NTI). It’s the least known of the bunch and had the best risk-reward.
NTI used to be owned by Marathon Oil but was sold to ACON Investments and TPG, a private equity firm, a few years back. TPG took it public late last year as a variable-payout master limited partnership (MLP).
What makes NTI unique is its location in St. Paul, Minnesota, making it one of the refineries closest to the Bakken Shale. There is not enough refining capacity in the region, and therefore about a fifth of refined products consumed in the region have to be imported into St. Paul from the Gulf Coast refineries. Texas or Louisiana refineries, however, will not send gasoline to St. Paul if prices there are below their total cost (of oil production, refining and shipping). This point is very important, because the price of gasoline in St. Paul is a function of what coastal refineries pay for the oil they refine. These dynamics get more complicated, because coastal refineries may be buying either Brent Crude — oil imported from other countries — or West Texas Intermediate (WTI). To keep things simple, I’ll focus on WTI (Brent usually trades at a premium to WTI, so I will err on the conservative side).
Since oil from the Bakken is shipped by rail or pipeline to the Gulf Coast, the price differential between Bakken and Gulf Coast is the transportation cost. NTI buys Bakken oil at a discount to WTI (a reduction of $9 per bbl if shipped by pipeline and $15 if shipped by railroad), refines it and sells it at a premium over the cost of the WTI that Gulf Coast refiners have to pay to ship gasoline to St. Paul (it costs them about $2 per bbl).
This unique dynamic allows NTI to earn outsize profits that should be sustainable in the long run. It is able to capture an additional $11 to $17 in gross margin profit, versus the coastal refineries selling gasoline in St. Paul, and thus should earn about $3 to $5 per share. NTI’s earnings will likely be volatile going forward, but we can accept this volatility considering that we are paying six times its worst-case earnings. Unlike traditional MLPs, NTI pays out all its earnings in distributions, giving this refinery a dividend yield somewhere between 15 percent and 26 percent.
NTI is misunderstood for many reasons. First, it just completed a planned five-year turnaround (every five years the refinery is shut down and fixed up). Thus output was down, and the company cut its dividend, which scared a lot of retail investors who owned blindly on dividend. TPG, the largest shareholder, just had a secondary offering of its shares, and that might have put additional pressure on the stock. It may be down in sympathy with the MLP sector, which took a dive when interest rates spiked. But NTI pays a much higher dividend than almost any other MLP, and it is significantly underleveraged.
In the market that is hitting all-time highs, most of your errors will come from forced-buy decisions. Constantly rising stock prices create an enormous pressure to be fully invested. They force you to compromise on your valuation standards. You start defaulting to relative (instead of absolute) valuation to justify your buying decisions: that is, for example, this stock is trading at 18 times earnings but its competitors are at 22, so it is cheap. Bull markets don’t last forever, nor do the premium valuations that they bring. The value in NTI is not a mirage that will dissolve with this bull market. It should do well in any market environment, as it is dirt cheap in both a relative and, more important, an absolute sense.