Description

            Freightcar America (RAIL) is the #1 manufacturer of coal freight cars in North America, with an 80% market share. Currently, RAIL sells at an EV/EBITDA= 2.7 (based on 2006 EBITDA; EV/EBITDA (normalized) < 5), with more than 27% of its market cap as Net Cash. 2006 EV/FCF= 3. This is a very simple, easy to understand, investment idea. Mr. Market is concerned about Q3—Q4 2007 and the quarter to quarter volatility of orders, but I see the opportunity to buy a business at a surprisingly low price where the visibility for at least the next 5-10 years is clear. 2007 and 2008 are unlikely to be as strong as 2006 (but should still be nicely profitable), but sometime in the next 1-2 years RAIL should be the subject of a feeding frenzy among the Cramer’s of the world.

RAIL was established in 1901. For much of its life it was part of Bethlehem Steel. It was privately owned from 1991 until the April 2005 IPO. For much of that time it was known as Johnstown America. RAIL also manufactures other types of freight cars, but 80-90% of their shipments are coal cars, at about $80k each. RAIL sells to the major railroads (such as Burlington Northern, Union Pacific, Canadian Pacific), electric utilities (such as TXU and AEP), and leasing companies (Mitsui Rail Capital and GE Equipment Services). RAIL, unlike most of its competitors, does not have a leasing business. This is both a weakness and a strength. They do not have the option of smoothing out their production schedule with orders for their own fleet. On the plus side, they are not in competition with a large segment of their customer base.

            92% of US coal consumption is used for the generation of electricity. Traditionally, railcar shipments are strongly cyclical but over the next 5-10 years, the market for coal cars will be very strong based upon the normal growth of electrical demand (which can only be economically filled by increased coal shipments), the increased average mileage from mines to utilities, and the replacement cycle for freightcars.

RAIL has 6-9 months of firm backlog with raw material cost escalator provisions. Contracts include cancellation clauses under which customers are required, upon cancellation of the contract, to reimburse RAIL for costs incurred and lost profits. However, customers may change requested ship dates and there are the usual other T&C, such as inspection rights which could prevent or delay backlog from being converted into sales.  

Q3 shipments were 5027 cars, resulting in quarterly revenues, earnings, FCF, and EBITDA of $396, $37, $40, and $61Million, respectively. Maintenance Capex is minimal; but  sometime soon they expect to spend $60 - $80 Million for an additional low cost factory (which mostly will replace higher-cost existing capacity; might not happen depending on negotiations with the Union) and up to four Joint Ventures (in Russia, China, and 2 undisclosed countries).

            Management expects Q4 shipments to be slightly higher than in Q3. Based on backlog and customer requirements, management may then dial back the production schedule (which is what the market is concerned about).

Quarterly orders are very lumpy. Q1, Q2, Q3 new booked orders were 1031, 3763, and 357, respectively.  This looks like a bad trend compared to the Q3 shipments of over 5000 cars and a booked backlog of slightly more than 12000. However, as of October 26, there were already 2073 units ordered in Q4. In addition, RAIL just announced an unbooked order for up to 7650 cars by TXU, contingent upon the utility getting the necessary permitting for several new coal plants. Delivery is planned for the second half of 2008 though 2009. Once the railroads approve their 2007 internal budgets, the ordering rate should pick up again. Norfolk Southern, for example, has declared that their CapEx budget will increase by approximately 10%.

Regardless of the quarter to quarter order rate, we can be pretty confident in developing a normalized order rate. There are 250k coal cars in the North American fleet. Coal cars have a useful life of about 25 years in that function. After that they can be used to carry other cargoes, such as aggregates. After 40 years, the cars are typically retired. The 25 year life results in an estimated replacement rate of 4%, assuming that cars have been added in the past to the fleet at about the same annual rate.

In fact, the order rate has historically had lots of peaks and valleys. There were peaks in total freight car construction in 1979-1980 (after a 15 year period of strong shipments), 1998-1999, and 2001-2003. 1982-1993 represented a sustained period of weak shipments.Shipments between 1984-2003 were only 60% of the prior 20 year period. The coal cars delivered in the sustained strong shipment cycle of 1970’s are due to be replaced. Validating this conclusion is Norfolk Southern’s announcement last month that they planned to replace “most” of their 33,000 coal car fleet over the next 10 years. This yields a replacement rate for NS of 5-10%, depending on the interpretation of “most”. In my analysis, I use a replacement rate of 4% over the next 5-10 years, which may well be too low.

Besides replacement, there is also a growth component for the normalized order rate.  The long term annual increase (both past and projected) in US electric power demand is 1-2%. Most of this will be met by coal since coal is much cheaper for utilities than natural gas. NG needs to be less than $5 per MMBTU to be anywhere close to competitive even after taking into account pollution credits Given the structural issues relating to NG supply, this kind of pricing seems unlikely on a sustained basis. Barring a sudden commitment to conservation by the American public, growth in the coal market seems assured over the next 5-10 years.

Increased coal production is expected to come from the Powder River Basin (PRB) in Wyoming and Montana (source of 40% of US coal production). Most coal from the PRB is transported via a rail line jointly owned (called either the ‘Joint Line’ or ‘Orin Line’) by Union Pacific and Burlington Northern. This is the principal bottleneck for coal output. In fact, you may remember that because of difficulties in 2005 on the Joint Line, many utilities’ coal inventory dropped to an uncomfortable level. The UP and BN have been pouring money into increasing the reliability and capacity of the Joint Line. The next major increase in capacity will take effect in 2009 (for a capacity of 400 million tons per year vs 325 million in 2005). Utilities (such as TXU) are basing their building plans of new coal plants based on this schedule. Utilities then need to start building their coal inventories 6-9 months prior to the plant going on line and for that they need new coal freight cars.  Thus we’d expect to see a surge in new coal car shipments beginning in the 2^nd half of 2008, consistent with TXU’s order.

Coal from the PRB tends to travel more miles to get to utilities than coal from other coal producing regions. This also increases the number of coal cars needed because they are able to make fewer round trips per year (currently about 40) as the mileage and congestion increases. The effect of this is probably a 1-2% annual increase in the required coal car fleet.

There is also a proposal by the DM&E railroad to upgrade existing routes out of the southern PRB directly eastward. This could increase capacity from the PRB by another 100 million tons annually. However, my analysis assumes this expansion never happen. If it does, that’s great for RAIL

Taking into account that most spare coal car capacity will soon be (or already is) minimal, and adding the replacement rate (4%), increased coal usage (1-2%), and increased coal car mileage (1-2%), one comes up with a normalized order rate of 6%-8% on a base of 250,000, or 15k-20k. Assuming that RAIL continues to maintain its 80% market share, this yields a normalized coal car order rate for RAIL of 12k- 16k. Assume the lower end, but also assume that 15% of RAIL’s orders are non-coal, as history suggests. This results in a normalized total order rate for RAIL of  14k – 19k. Based on RAIL’s EBITDA in Q3 2005 (when shipments were near the lower end of this range), this should result in an annual EBITDA north of $100  million. This is even more conservative when one takes into account that since then RAIL has improved its cost structure significantly. RAIL now believes that it is the low cost leader, with competitive delivery responsiveness.

So, even after taking into account the most conservative assumptions I could reasonably justify, I come up with RAIL having an absurdly low normalized EV/EBITDA of 4.9. Now, let’s add the unfunded pension ($27 million) and 5 year operating lease obligations ($5.3 Million) to the EV. One still ends up with a normalized conservative EV/EBITDA of 5.2. It’s hard for me to come up with a reasonable scenario where this stock is not far too cheap (although see the Risk Factors below).

            One of RAIL’s biggest problems is what to do with all their cash ($185 million, $14.40 per share). Management believes that they will be able to announce their plans for this cash by their next conference call. Some of their cash will be used for the Joint Ventures and they also think there is the opportunity to buy some businesses within the industry, but I think a good deal of the money will end up being used for a stock buyback. I also think that they are as likely to be a target for an acquisition by a leasing operation as they are to be a hunter. As you can see from the Comps below, RAIL is a bargain on a relative basis also.

Comps: The other railcar manufacturers all have leasing and/or railcar management services and serve a more diversified customer base than RAIL. Still, it’s hard to justify the difference in valuation. (These are TTM from a database, but I don’t believe any distortions affect the conclusions you’d draw)

Catalyst