Description

LONG: TimkenSteel Corp (NYSE: TMST)

TimkenSteel Corporation (“Timken” or “TMST”) manufactures alloy steel bars, tubes, and other products using electric arc furnace technology. Their products are used in the automotive, oil and gas, and industrial equipment industries. The company was founded in 2014 as a spin-off from The Timken Company (“TKR”), which makes bearings and power transmission components. Timken is known for its expertise in using nearly 100% recycled steel to make high-quality products. Its equity market cap is about $820 million.

Timken has been on a roller coaster since the spin off. In their first year of independence, they earned $217 million in EBITDA on $1.7 billion in sales. Then volume declined 24% and EBITDA fell below zero in 2015. It rebounded to $127 million in 2018, fell to $32 million in 2019, and recovered to $246 million in the steel boom of 2021. More recently, gross profit fell from $80 million in 2Q22 to negative $20 million in 4Q22. The latest earnings collapse is due to tragedy. Their only furnace facility suffered a fatal accident in July 2022. They had to shut it down, relying on finished goods inventory and third-party work to satisfy customers.

Analysts are underestimating their potential to recover. Consensus has Timken earning less in a year of production in 2023 than the half year of production in 2022. This makes little sense given their exposure to three economic bright spots: auto volumes, oil and gas production, and capital equipment. We think the company could earn $190-$200+ million in EBITDA in 2023 versus the sell side outlook of only $127 million. This estimate puts them at sub 4x forward EBITDA. TMST is currently trading at $18 per share but we believe it is worth at least $29 with 62% upside and far less downside.

Because the accident arguably resulted from strategic efforts to increase utilization on a single furnace, longer-term history is important to the setup.

The Spinoff

Henry Timken founded the Timken Roller Bearing Axle Company in 1899. Fast forward to 2012, and the Timken Company (NYSE: TKR) had a sprawling portfolio of products across four segments. They earned over $900 million in EBITDA on $4 billion of sales. The Timken family owned 10% of the stock and three descendants of the founder were on the board. This included Ward Timken Jr., the great-great grandson of the founder, who was made Chairman at the age of 37 in 2005.

In August of 2012, activist investor Relational Capital teamed up with CalSTRS to control 7% of shares. They filed a 13D and agitated to spin off the steel division. They argued over the following months that a split would unlock value, since the stock was trading at a discount to peers on a sum of the parts basis and failing to demonstrate synergies from disparate businesses. Buttressed by accusations of poor governance, the centerpiece of which was $9.4 million of 2011 pay to Chairman Timken Jr., the campaign was successful. Shareholders approved the proposal on May 7, 2013.

For the soon-to-be-independent TimkenSteel, the ordeal had awkward timing. Their division president, a 40-year company man named Salvatore Miraglia, announced his retirement in the same month the activist campaign began. In the meantime, the division was supervised by two recently appointed “Group Presidents”, each with a wide variety of TKR’s operations under their scope. In the end, Ward Timken Jr. was made Chairman, CEO, and president of the spin co.

Post Spin

Former employees describe the strategy post spin as “go big”. Timken’s initial investor day in September 2014 highlighted a $200 million investment in a jumbo bloom vertical caster, $35 million for an in-line forge press, and $25 million for a ladle refining system. Together, the investments were intended to expand capacity for the booming oil and gas business and deepen their competitive advantage in large diameter bars. The collective investment amounted to ~10% of their $2.3 billion TEV on the investor day and over a third of year end book value. The timing could not have been worse as the oil and gas industry, their most profitable end market, collapsed. The issue spilled over into broader industrial weakness.

Capacity utilization is critical for Timken. The volume declines led to flatlining profits. Enterprise value bottomed out near $350 million in January 2016, 85% below the post-spin peak just 15 months prior.

Clearly “go big” was unlucky at best and mismanaged at worst. Timken Jr. led an efficiency driven recovery to nearly half of 2014 gross profit levels by 2018. But as gross profits declined again, the company announced a CEO transition in October 2019.

Interim management guided the company through most of 2020 and hired current CEO, Mike Williams, in December of that year. No Timken family member remains involved at TMST today.

New Management / Tragedy Strikes

Williams decided to reverse course and “go small”. When he took over, TMST ran three manufacturing facilities called Faircrest, Harrison, and Gambrinus. All are in Canton, Ohio, advantageously close to customers. To increase capacity utilization and get profit recovery from overhead absorption, the company indefinitely idled all assets at Harrison, condensing operations from two electric arc furnaces to one. Aided by the post-covid steel boom, this worked until it didn’t.

By Q2 2022, the company had achieved $80 million per quarter of gross profit, a post-spin record, just before the fatal accident in July 2022. After sadly losing a young father, the company is taking steps to improve safety including $7 million of training, equipment, and process expenses plus an additional $45 million in capex for “enhancing our manufacturing excellence and asset reliability programs”. We don’t have the expertise to judge if this is adequate, or if overworking the single melt facility increased the odds of an accident. But it certainly didn’t help.

From a business perspective, moving to a single point of failure made Timken’s profits more fragile. Recent results deep in negative gross profit territory demonstrate that it is in the financial interest of shareholders to invest in safety. Proper maintenance will cost some utilization, but it is worth it for everyone involved.

Going Forward

The company has not yet capitalized on the recovery in their end markets currently underway:

• Automotive (“Mobile”) is 45% of their volume. IHS forecasts 5.5% year over year growth in North American production to 15.1 million. This remains significantly below pre-covid averages and reflects an improving semiconductor supply chain.
• 46% of volume is for “Industrial” end markets, which include a broad basket of subsectors. Named customers include Caterpillar, Eaton, General Dynamics, Reliance Steel, former parent TKR, and a subsidiary of Park-Ohio. Only Reliance is expected to have lower sales in 2023. We’re told the largest customer is CAT, growing at mid-single digits. The combined basket is expected to grow sales by 4%.
•  “Energy” markets generate 9% of shipments. Although profits are not disclosed by end market, our interviews consistently indicate energy is the most profitable sector. Timken remains a leader in large alloy tubes used in fracking. Although rig counts are slightly down in recent months, production is still growing year over year. OPEC cuts will support demand for North American energy.

Industry level strength is all well and good in cyclicals. The real reason to doubt consensus profits is the simple fact that the single steel producing furnace will be running normally for 12 months instead of 7.

We don’t know how the analysts covering this stock can get to this profit figure without a time machine to witness a complete reversal of end market trends or egregious management misstatements. Modeling this company is simple. Sales = tons x (base rate + raw material surcharge rates). Raw material surcharges follow steel scrap prices.

IR tells us the company can reach 800k-900k tons of production when capacity recovers. This is believable, since they did around 200k per quarter for six straight quarters prior to the incident with weaker end markets:

They should approach this pace by 2Q23 if we believe their statement that capacity utilization reached the 70% range in February. Utilization averaged 77% in the six quarters from 1Q21 to 2Q22.

Timken’s gross profit bridges are pictured below and show four key components to profit changes. Here is what these looked like before and through the furnace shutdown in Q3:

As we look toward 2023, we believe the table is set for all four components to be flat to up:

1. Base price should be up since Timken has not had trouble taking price to offset inflation. Management forecasts a mid-single digit base price increase, which historically drops to gross profit at 45%. This works out to +$20 million gross profits in 2023 from base price, taking “mid-single” to mean a 5% rate increase.
2. Manufacturing efficiency should be negative for one more quarter as utilization ramps, but positive for the rest of the year as they lap the furnace shutdown. If the first quarter is roughly half the y/y headwind of 4Q22, and they only get back half of the $100 million in shutdown efficiency losses, due to running a more conservative utilization and safety related scheduling, there should still be +$30-35 million in gross profit increases for 2023 on overhead absorption and efficiency.
3. Raw material spread tends to rise and fall with steel scrap prices on gross profits, because they surcharge slightly ahead of when the rising input costs are recognized. The inverse is also true. This adds up to zero influence on gross profit over many years. Steel is rising now so this is certainly a near-term sequential tailwind, but it is difficult to map onto quarters since the bridge data is year over year. Although it is more likely to be positive than negative, we model no help from raw material in gross profit changes this year. 
4. Volume is obviously going to be positive while lapping a furnace shut down. It should be up about $17 million.

Add this all together and we get a 2023 gross profit that is $74 million higher than in 2022.

SG&A hit a record low as a fraction of sales in 2022. It should probably go higher by at least $7 million as they invest in safety-related items. Other operating income is going to come down on pension related changes. Neither is likely to undo the huge increase in gross profit. This should add up to $190 million in EBITDA, 50% higher than consensus. With a net cash position and an active buyback, EPS should be about $2.75 against consensus of $1.27.

A Special Catalyst

As you see in the table above, the company received an insurance payment related to the accident of $34.5 million. Critically, these inflows are not adjusted out of headline results. They are not done pursuing insurance payments.

“We continue to seek additional insurance recoveries related to the unplanned downtime experienced in the second half of the year, although the timing and amount of potential additional recoveries are uncertain at this time.”

 – Kristopher Westbrooks, CFO, 2/24/23

Management refused to provide specifics for timing and amount, both on the call and in our conversations with them. However, they insist that the amount is not trivial. This potential inflow is not embedded in consensus. If this payment happens, it will boost profits even further above expectations.

Balance Sheet & Valuation

Management’s best decision was to de-risk the balance sheet by paying off almost all interest-bearing debt and cutting the unfunded status of the pension by over 30%.

With $257 million in cash on hand and $200 million in undrawn credit, Timken has ample resources to invest in safety, rebuild inventory, contribute to the pension plan, and withstand cyclical shocks. The market cap is $805 million, and the enterprise value is $730 million including the pension. This is sub 4x our F23 EBITDA estimate and a FY1 P/E of 6.5x at the current share count of 45 million. The buyback is active and authorized for just under 10% of the shares.

Peers trade at bifurcated EBITDA multiples. Gerdau, Vallourec, and Mittal at 4x while Reliance, Steel Dynamics, and Nucor trade at 8-9x. The 10K names direct competitors from both sets. Timken’s products are specialized, and we think they could trade towards the higher valued peer set if they can sustain good results. 6.5x our estimate of $190 million in EBITDA leads to our PT of $29 per share.

Furthermore, running a single electric arc furnace is suboptimal at best. Customers, employees, and shareholders would all be better served with this asset as part of a large organization that can better absorb necessary maintenance downtime and serve customers from something other than safety stock when accidents happen as they regrettably do in heavy industry. This is speculative, but our interviews suggest they have been considered an acquisition target by bigger industry players at some point. This is more likely to happen without resistance from the well-paid Timken family.

Downside is protected by book value at 19% below market value. If EBITDA ends up way below our estimate at $150 million due to cyclical end markets, it could still trade 4.5x EBITDA and only result in a loss to shareholders in the teens. The free cash yield should also be in the mid-teens this year.

Conclusion

TimkenSteel should earn meaningfully more than market expectations in 2023. Management targets for volume were recently achieved with weaker end markets than exist today. Every  component of their gross profit bridge is pointing positive. Using our higher estimates of profits, the stock is dirt cheap. There is big upside if the market starts moving them toward the higher end of steel industry multiple ranges, and limited downside in both estimates and multiples should end markets turn sideways or down.

Catalyst

Catalysts

1. Receiving an additional insurance payment
2. Earning more than consensus throughout 2023
3. Actively buying back shares
4. Buyout target (speculative)

Risks

• Cyclical industry
• Needs safety improvements
• Dependent on a single facility to operate

Protection

• Low valuation
• Low expectations
• Specialty capabilities
• Worst case scenario is in the very recent past