Description

What is TMS?

• In their words: “We are the largest provider of outsourced industrial services to steel mills in North America…We offer the most comprehensive suite of outsourced industrial services to the steel industry. Our employees and equipment are embedded at customer sites and are integral throughout the steel production process other than steel making itself.”

• TMS was taken public in April of 2011 for $13.00/share by Onex, the large Canadian private equity firm.  Onex continues to own ~60% of the company and controls ~85% of the voting rights through its class-B shares, which do not trade publicly.

Steel? Isn’t that a bad business?

• TMS is a great business with several highly desirable characteristics and good growth prospects. For various reasons, it has “fallen through the cracks”, and now trades at a large discount to a conservative estimate of its intrinsic value.

• It’s not hard to see why TMS is not on the radar of most investors:
  • It’s small—~$413mm market cap—and since Onex owns 60% of it, it only trades an average of $650k/day.
  • If you just look at the GAAP financials, it appears to be a low-margin (~5% EBITDA margin) business with volatile revenues, and thus would be excluded from many screens. As we will see, this is highly misleading.
  • The steel industry is not generally known for generating good shareholder returns. Indeed, with some notable exceptions, the steel business in North America has been a bad place to invest for a long time, and the same is true for the scrap steel business. I suspect that TMS has been incorrectly tarred by the same brush, which has prevented it from getting the attention it deserves.

Why is TMS a good business?

• Basically, TMS is insulated from most of the issues that plague the steel business as a result of its differentiated business model.

• TMS provides various services for steel mills (scrap procurement, slag recycling, surface treatment, etc.) on an outsourced basis more cheaply and effectively than the mills could do on their own, while also giving the mills greater cost flexibility. TMS saves its customers money and makes them better operators.

• Importantly, ~88% of TMS’s revenue comes from long-term, iron-clad contracts with an average duration of 7 years. The critical feature of these contracts is that they depend only on the volume of steel produced, NOT the price.

• Furthermore, TMS is able to protect itself from sudden downturns in the steel industry by using tiered-pricing based on production volumes and contractual minimum monthly fees regardless of volumes.

• How is TMS able to get such attractive terms from its clients? In short, it’s because they have a great track record with their customers and have performed reliably and safely for many years.

• The best evidence for this claim is that they have worked with the same customers for a very long time, and they keep coming back:

“In 2011, we renewed and extended 11 contracts globally for various terms at multiple locations—a 100% renewal rate for the year. Our aggregate contract renewal rate since 2005 is now over 97% based on Revenue After Raw Materials Costs.” (Source: 2011 Annual Report)

• That’s not all—in addition to tiered pricing and minimum fees, TMS enjoys a few other characteristics that protect the company from macro volatility:
• Highly variable cost structure: The company claims that “approximately 80% of operating costs are variable”  (Source: 9/21/2012 Company Presentation). This is the opposite of what one might expect in the steel industry, and is one of the ways TMS provides value to its customers, which generally have limited labor flexibility because of the strong presence  of labor unions in the industry.
• Capex flexibility: As the company explains, “Maintenance capital expenditures [are] tied to equipment utilization”—that is, tied to the volume of steel produced by their customers, and thus in sync with TMS’s revenues. So if business is slow, they are not “burning up” the useful life of their assets. (Source: 9/21/2012 Company Presentation)
• Almost no inventory/price risk: We will investigate this in more depth later, but for now, the important thing to note is that TMS takes very little commodity price risk. This is a huge problem in the steel business, and companies like Nucor and Sims Metal Management often get burned when there is price volatility in steel markets— like when the price of scrap steel declined by over 25% in a few weeks earlier in 2012.

It’s Not “Build it and they will come”

• The other really nice thing about TMS is that “growth capex” is really for growth—they define this term to include only capex that is tied to specific new contract wins.

• Unlike most companies, which invest capital up-front and hope to earn a return on the capital in the future, TMS doesn’t spend capital until it has a rock-solid, signed contract with guaranteed revenues for a term of anywhere from 5 to 12 years. This significantly de-risks the business and allows for disciplined capital allocation.

• Basically, TMS is able to figure out what sort of returns on capital they are likely to earn on any new contract, and if it isn’t high enough, then they don’t do the deal. But generally, it isn’t hard for TMS to hit their hurdles since a large part of new contract wins are the result of cross-selling to existing customers (11 of 13 wins in 2012 came from cross-selling)—thus leveraging the existing infrastructure investments at the client site.

Misleading Accounting

• I claimed earlier that the accounting treatment for TMS leads to a distorted view of their margins and revenue stability. This is a result of the way TMS’s scrap procurement business operates.

• TMS uses its know-how, proprietary logistics software, and global network of metal traders to source scrap steel more cheaply and efficiently than their clients could do on their own. But (in yet another indication that TMS is delivering value-added service to its clients, rather than simply offering a balance sheet like a typical distributor), TMS is able to mostly avoid taking price risk:

“Because of our model, where we say, 90— over 91%, 92% of our transactions are back-to-back, basically, where we lock in a sale and a buy at the same time. So that, basically, protects us against large write downs that we see with other competitors in this business, where they hold a lot of inventory. So the inventory that we hold is minimized by our model.” (Ray Kalouche, President & COO - Mill Services Group; 8/2/2012 Earnings Call)

• Incidentally, TMS does not suffer from the working capital demands that often accompany a large amount of “pass-through” revenues, and typically is able to more than offset growth in inventories and receivables by extending payables.

• Since TMS does in fact briefly take title to the scrap they procure for their clients, GAAP accounting forces TMS to recognize this on a gross rather than net basis.

• Thus, for example, in 2011 TMS had total revenue of ~$2.7 billion— but the “effective” revenue, or as they call it, “Revenue after Raw Materials Costs”, was just $539mm. So about 80% of their revenues are simply a pass-through of costs for which they assume almost no risk.

• Still, this makes the company look like it earns an EBITDA margin of 4% to 5%, when in reality, it earns closer to a 24% to 26% EBITDA margin. On top of that, TMS’s top-line is injected with spurious volatility as a result of the price changes in scrap steel. This makes it look on the surface like it is much more levered to the industrial cycle than it really is.

Growth Opportunities

• As mentioned earlier, TMS is able to grow purely from cross-selling additional services to existing customers at current operating sites, and of course can also win additional sites from existing customers.

• The bigger opportunity, however, is international expansion. Luckily for TMS, this expansion can be done with significantly less risk because they can simply follow their existing customers that already have a global footprint, such as ArcelorMittal.

• The following management quotes from TMS’s 8/2/2012 earnings call give a sense of the company’s rapid progress in realizing this international growth:

“Year-to-date, revenue after raw materials costs from international operations, which we define as non-US and Canada, is now approximately 26% versus 6% in 2007. Even though we continue to win new contracts in North America, we expect this percentage to further increase in the long-term, as we selectively grow internationally.”

“We continue to have success penetrating international markets, both through our mill services group and our raw material and optimization group. We expect this trend to continue, as our mill services contract pipeline remains very healthy…”

• Although North American represents about three quarters of TMS’s business today, the North American steel market is actually not all that large compared to some other international markets. The following chart (from an ArcelorMittal presentation) lists the largest steel producers in the world. Existing TMS customers are highlighted in yellow:

• As shown by TMS’s continued cross-selling success, there is plenty of growth runway within existing customers. But there is also a good opportunity to forge new relationships with large global players.

• As international steel producers become more sophisticated and competitive, it is not unreasonable to expect that they will follow in the path of the North American producers and embrace out-sourced services for the performance and efficiency reasons. 

Normalization

• Clearly, the North American steel industry has recovered substantially from the low levels reached during 2008 and 2009. After all, steel industry capacity utilization  reached a shocking low of 36% in early 2009, after hitting 90% in early 2008.

• But as the chart below shows, industry capacity utilization is still significantly below the average level for the past decade (of course, this could get significantly worse before it improves):

• One might argue that the last decade was not truly representative, since there was a credit bubble (although residential housing does not consume a very large amount of steel). But the average capacity utilization rate going back 25 years is only slightly lower, at 82% versus ~85% for the last decade. (Source: TMS prospectus)

• Thus, TMS should be able to grow sales organically (i.e., through existing contracts) if utilization rates return to the longer term average levels.

The Proof of the Pudding…

• So far, I have claimed several times that TMS’s business model protects it from the vagaries and brutal swings of the steel industry. If that is really the case, then TMS should have gone through the 2008 – 2009 financial crisis without getting beaten up too badly.

• To see that this is the case, let’s first get a sense of just how clobbered the steel industry got during that period. The chart below shows the EBITDA of 3 steel companies (Nucor, U.S. Steel, and ArcelorMittal— all customers of TMS) and Sims Metal Management, the largest dealer/processor of scrap steel in the world, normalized to a scale of 100 beginning in 2005:

…is in the eating.

• The chart is pretty amazing— it shows that all of these companies suffered massive deterioration in business performance during that time, with EBITDAs declining over 50% from peak to trough, and in some cases much more. 

• So what happened to TMS? As you can see below, the impact on TMS was extremely limited in comparison— adjusted EBITDA fell just 16.5% from 2008 to 2009, and then more than recovered in 2010!

Valuation

• We have seen how much better TMS fared during the 2008-2009 downturn than its customers in the steel industry. While these other firms faced unprecedented difficulties (in addition to the plunging capacity utilization that we saw before, scrap steel prices declined 81% from July to November of 2008), TMS suffered a ~16% decline in EBITDA.

• Clearly, there are plenty of indications that TMS is a much better business—the long term contracts with high renewal rates, the lack of commodity price risk, the favorable working capital characteristics, the lower risk associated with growth capex, etc. Thus it seems reasonable that TMS should trade for a much higher multiple than its steel producer customers.

• Indeed, I would go much further and suggest that TMS should trade at a valuation commensurate with its attractive characteristics and growth prospects, and should have little to do with the valuation of steel companies. But for now, let’s just take a look at where these firms trade. Note that the multiples below (from Bloomberg) are based on analyst consensus estimates for the next two fiscal years. We can see that they trade for at least 5-6x EBITDA on average (Note: these multiples are a couple weeks out of date):

• Now let’s look at the valuation for TMS. To start with, the company has given detailed guidance for 2012, as shown in this slide from their 9/21/2012 investor presentation:

• It’s important to note that the EBITDA guidance from 2012 does not fully reflect the contribution of the new contracts signed during the year:

• Still, we can use the 2012 guidance as a base to show that TMS is extremely cheap without assuming the additional contributions from new contracts:

• There are a few things to note about the numbers presented above:
• The company refinanced its 9¾% bonds with a considerably cheaper LIBOR based term loan in March of 2012. The interest expense assumed above annualizes the amount paid in Q2-2012, which is reflective on the current run-rate.
• To be conservative, the maintenance capex assumed above is slightly higher than the $38mm annualized number arrived at by taking the Q2-2012 total capex less the amount specifically described as “growth capex”.
• The company gives the 32% target tax rate for 2012 in its August 2, 2012 conference call.

• One could argue that I only assumed maintenance capex of $40mm/yr, while the annualized depreciation expense from Q2-2012 is ~$55mm/yr. Since the “effective” depreciation of TMS’s equipment is a function of customer production volumes, it is not unreasonable that the true maintenance capex would diverge from the depreciation schedules adopted at the start of the contracts, when production levels were higher. The company should be in a reasonably good position to estimate this.

• Furthermore, one could argue that it isn’t fair to value TMS without assuming significant growth capex, since their revenue growth in recent years is clearly a direct result of this contract related capex. This is certainly true, but I would say that a high-quality company without significant revenue growth is still attractive at a 15% free cash flow yield. And a high-quality company like TMS that can steadily grow—especially given the low-risk nature of the capex that is spent only after a contract has been secured—merits a much higher multiple.

• But even if we ignore all of that, and just look at the EV/EBITDA multiple—which is of course a misleading statistic in that TMS is  “capital light” compared with the steel producers—we still reach the absurd conclusion that TMS is trading for a much lower multiple than these much inferior steel companies, with their volatile cyclicality, exposure to commodity price risk, high fixed costs and inflexible labor unions, and, lest we forget, their tremendous pension and healthcare benefit liabilities which are not reflected in the computed TEVs.

Another Indicator

• Aside from the compelling valuation case that can be made simply by looking at TMS’s guidance for the current year—that is, without resorting to various manipulations of the numbers, some of which are quite reasonable, such as reflecting the full contribution from signed contracts—there is also some circumstantial evidence that the company is under-valued, and that is to look at the behavior of Onex, the largest holder.

• Take a look at the preliminary prospectus below. Here you can see that Onex originally planned to sell over 6 million shares of stock for a mid-point price of $16/share:

• Partially as a result of the emerging European debt crisis at that time, the price of the offering fell significantly to $13/share just a few days later, as shown in the final prospectus:

• The important thing to note here is that Onex slashed the amount they were willing to sell, from over 6mm shares to just 2mm shares, while increasing the shares sold to the public. This is a pretty good indication that they felt the valuation at $13/share was too low and didn’t reflect the intrinsic value—that is, that they would be better off holding onto their shares. One might object that they shouldn’t have done the IPO at all if they felt that way, but they may not have had a choice (remember, they are a PE fund, and the funds have end dates).

• In some ways it is even more strange that the stock has continued to fall from the IPO price to its current level of ~$10.67– this despite the fact that they have continued to growth revenues and earnings, to sign new contracts, and to insist that their sales pipeline is “very healthy”.

Management Alignment

• Another nice thing about TMS is that the management team is well aligned with shareholders, since they are fairly large shareholders themselves (current executive are show in yellow in the following excerpt from the most recent proxy):

• Furthermore, the options that were awarded to top executives at the time of the IPO are not exercisable unless the stock is at least $14.95/share— over 40% higher than where it is currently trading.

Risks/Issues

• The biggest problem with TMS is that Onex still controls it completely. This is obviously not an ideal situation in terms of corporate governance for an outside shareholder, and it also has other negative side effects, such as the low liquidity of the shares and the potential overhang from Onex selling more of its stake.

• At the same time, there is also a positive side to it: Onex is a very smart owner and can be expected to prevent the company from making stupid acquisitions or other bad decisions. Furthermore, Onex is likely to act swiftly to take advantage of opportunities, such as the following coup described in the prospectus: “In April 2009, we acquired $2.0 million of face value of our senior subordinated notes due 2015 for a cash payment of $0.4 million.” These aren’t necessarily bad guys to have on your team!

• Another issue is the concentration of TMS’s revenues among a small number of customers (top 10 customers were 79% of revenues in 2011). I would argue that this is offset in a couple ways. For one, TMS works through site-specific contracts, not customer-specific contracts, which means that each contract is negotiated separately. The other thing is that TMS has been working with these customers for an average of 30+ years, and they seem to renew the contracts virtually all the time. This shows that TMS’s customers view it more like a valued partner rather than as an adversary that they are trying to squeeze at every opportunity.

• Finally, steel is obviously a highly cyclical and often troubled industry. There are currently some indications that the US steel industry is facing near term issues in terms of over-capacity and price competition, as well as pressure from cheap foreign steel producers, some of which may not be economically motivated (e.g., China). While I have argued that TMS is insulated in several ways from the brunt of these factors, it would be naive to suppose that it won't have a negative impact on TMS. The question is whether the current valuation for TMS more than prices in this risk with a margin of safety, and in my view, it does.

Catalyst