Description

Following a perfect storm of negative technical dynamics, recent spin Garrett Motion (“GTX”) is a coiled spring ready to double over the next year as investors learn more about the company and sell-side analysts get more constructive. With good visibility to $3+ in near-term earnings power, the stock should be higher a year from now in all but the most punitive recession scenarios, and in an upside case could be a triple.

The thesis in brief:

1. “Mini” spins like GTX have a very good historical track record – the odds are on your side
2. GTX has faced a near “perfect storm” of negative technical dynamics, resulting in a tremendous entry point
3. Insiders clearly see an attractive opportunity and are heavily incented to make this work
4. Sell-side apathy has resulted in gross mismodeling, and has fed investor confusion and pessimism
5. While a large inherited asbestos liability has kept investors away, it isn’t as bad as it appears
6. Business quality is underappreciated and operating sensitivity to a negative macro environment is much smaller than for other auto companies

Business Description:

GTX is the global leader in manufacturing turbochargers for light and commercial vehicles, with a ~33% global share, slightly ahead of BorgWarner @ 29% and IHI @ 15%. They have been producing turbochargers since the 1950s, have extremely high brand recognition and are perceived as the highest quality producer – end customers know the brand and want it (in their investor day deck they highlighted a customer who had tattooed the GTX logo on his arm). They are a global producer, with facilities around the world that mirror their customer base. In addition to turbochargers, they have a smaller position in electric-boosting and connected vehicle technologies.

Turbochargers are currently utilized on ~50% of vehicles sold and are highly-engineered devices that utilize an engine’s exhaust gas to drive a turbine that forces extra compressed air in the combustion chamber, improving an engine’s power output. In practice, turbos are often paired with a smaller engine profile to generate equivalent power to a larger engine but with lower emissions and lower fuel consumption. For a gasoline engine, this means that instead of running a V6 3.0L engine, you could instead get similar power output with a variable nozzle turbocharger and a 4 cylinder 2.3L engine, reducing emissions by 25%. The growth in hybrids is also a driver of turbocharger demand, as battery power can be used to drive the turbine, creating even greater efficiencies.

As a result of the significant benefits that turbos provide, along with increasingly stringent fuel efficiency and emissions standards, penetration levels are increasing rapidly, with gasoline engine penetration having increased from 14% in 2013 to 33% in 2017 (+475bps p.a.), with a further increase expected to 52% by 2022 (+380bps p.a.). Total engine penetration is expected to increase from 47% in 2017 to 59% by 2022.

Current Situation:

The following is a brief timeline of events of the last year and are helpful in understanding the entry point GTX is presenting today:

• In October 2017, Honeywell (“HON”) announced plans to spin-off GTX, which had operated as a subsidiary of HON since 1999
• Leading into the spin-off, sell-side analysts were ascribing $30-40/sh of value for GTX in their HON sum-of-the-parts valuations
• On October 1st 2018, GTX began trading regular-way – with a market cap only ~0.6% of HON’s, the company was an automatic dump for many HON investors and underwent heavy selling, with 70% of the outstanding shares trading in its first five days and the stock going from $18.65 at the open to as low as $15.73 (-16%)
• GTX proceeded to trade in-line with a weak broader market for the remainder of October and into November, reaching $14.81 the day before its Q3 earnings release
• On November 6th GTX reported their Q3 earnings – the release was confusing as it represented the company pre-spin as part of HON, and the implied guidance range for Q4 was overly broad, and directly conflicted with commentary made during the call
• GTX continued to sell-off following the call, bottoming at $11.43 in early December
• Since December, GTX has increased ~30% off of the lows, helped in part by a Jan 15th conference presentation where they revised FY18 guidance toward the top end of the range, assuaging fears that had emerged following their Q3 release

1. “Mini” spins like GTX have a very good historical track record – the odds are on your side

While it is well known that spin-offs are a traditional hunting ground for value investors, it is my belief that in the last ~5 years, this has been, in aggregate, a fairly poor pool of opportunity, with many parent companies spinning out critically impaired divisions to capitalize on investor interest, and with reduced forced selling dynamics as shareholders get smarter about trading in these stocks. However, one niche that still generates tremendous returns post-spin are what I call “mini” spins, where the spun-out entity is less than 5% of the market cap of the parent.

These situations are particularly likely to generate forced selling from both shareholders for whom the new entity is not large enough to justify doing work, as well as from indices who sell as the spun entity no longer meets market cap thresholds. I have counted 15 of these “mini” spins going back to 2010 – they typically bottom a week or two after going regular-way, and median returns from the bottom are extremely attractive, +23% over a 3-month period, and +34% over a 6-month period. Average returns are higher.

The only spin in recent years with a comparable disparity between parent and spinco market caps was ASIX, spun in October 2016, also from HON, with a market cap of < 1% HON’s. That stock bottomed at ~$15/sh and within a year had hit $45/sh. I believe this could be a template for GTX’s performance over the next year.

2. GTX has faced a near “perfect storm” of negative technical dynamics, resulting in a tremendous entry point

While many spins experience some degree of negative technical dynamics, GTX truly experienced a perfect storm. In addition to checking off all the items on a classic spin-off forced-selling checklist (tiny market cap vs. parent, different industry, perceived inferior industry, high leverage), GTX managed to:

• Start trading right as the market peaked and entered a sharp sell-off/panic
• Be a globally-diversified business right as investors started worrying about anything with emerging markets exposure (especially China)
• Be an auto company right as auto sentiment cratered
• Be a cyclical with leverage just as people started worrying about a recession
• Be a special situation just as people fled to safety and away from anything with complexity

Essentially, while most spin-offs stabilize post-spin as investors dig into the company and formulate a view on the company’s prospects and valuation, no one in their right mind was spending time on GTX as the world melted down around them and as they fought fires in their existing portfolio. There were multiple days in November when GTX was down 5-6% on no news at all in a flat market. These dynamics have resulted in GTX being a completely forgotten stock and generating the very attractive entry point seen today.

3. Insiders clearly see an attractive opportunity and are heavily incented to make this work

The two key components of a successful spin-off trade are that “institutions don’t want it” and that “insiders want it”. I’ve established why the former is true above. As for the latter, there are very strong indications that insiders see GTX as an attractive opportunity.

First, while most of the management simply ported over from HON, there was one meaningful addition to the senior leadership team and that was Alessandro Gili, GTX’s CFO, who joined the company from Ferrari, where he was CFO for 3 years prior. At a high level, I can’t understand why he would make this move unless he saw very meaningful potential at GTX. He left a comparable role at a marquee company that is meaningfully more profitable and relocated from Italy to Switzerland. While I couldn’t find his compensation information from Ferrari, >50% of his annual compensation at GTX is in the form of equity and he was given an additional sign-on grant of RSUs valued at ~$3M and vesting over 4 years – he clearly wouldn’t have taken this opportunity if he didn’t see promise.

Second, the remainder of the executive team have levered into GTX stock in a big way. Any HON options or RSUs granted previously were converted into GTX RSUs that struck based on the last when-issued closing price ($18.50). Notably, this gave GTX execs a strong incentive to low-ball expectations for the business in their investor day on September 6th. It also means that senior management own a large chunk of GTX from the get-go – the CEO alone had over 115,000 shares of HON options and awards that converted into 523,482 GTX RSUs. GTX’s top seven managers collectively own 1.6M RSUs, equivalent to $25M, or 2.2% of the company.

4. Sell-side apathy has resulted in gross mismodeling, and has fed investor confusion and pessimism

While sell-side initiations are often viewed as a positive catalyst for spin-offs, in this case they were the opposite, with 2 brokers initiating at sell ratings and 2 others initiating at holds. There were no buy ratings. I believe this tepid response was largely due to sell-side apathy (imagine being asked to do an initiation report in the midst of your entire coverage universe blowing up) as well as due to a lazy thematic short case involving limited exposure to electric vehicles.

As a result, the quality of the initiations and the level of intellectual honesty have been extremely poor. I was able to confirm with GTX IR that no sell-side analyst even bothered to speak with the company before putting out their initiations, and as a result there were material misunderstandings regarding the business, as well as gross modeling errors. I will pick on UBS as they are currently the only analyst to retain their sell rating. In their initiation, UBS:

1. Decided to assume an interest rate of ~5.5% on GTX’s debt despite guidance of ~4% in the form 10 and it actually ending up in the 3s
2. Incorporated $100M of buybacks per year in their model which didn’t end up reducing the share count, resulting in a scenario where GTX is essentially assumed to light $100M on fire each and every year (notably, one of their “key questions” for the stock was the pace of delevering – missing a $100M p.a. leak in your model doesn’t help you answer this question)
3. Arrived at a $13 PT by double counting $175M of asbestos liability payments every single year (in order for them to backsolve into their $13 PT again they would need to bump up their discount rate from 8.2% to 15.5%...)

I brought up all these issues to the UBS team – I’m still waiting for a retraction and for them to increase their DCF-derived PT up to $45…

Furthermore, UBS, along with every other analyst model I’ve seen, projects out $175M of asbestos payments to be made in perpetuity. While GTX’s inherited asbestos liability is large, and has likely kept investors away, treating it as a perpetual expense is extremely punitive. Given that it’s an important piece of the valuation, it’s worth digging into this in more detail.

5. While a large inherited asbestos liability has kept investors away, it isn’t as bad as it appears

As part of the spin, HON has saddled GTX with a large asbestos liability resulting from their legacy ownership of Bendix brakes. There are a couple of important characteristics to note about this arrangement:

• The liability is unrelated to GTX’s business – it arises from an unrelated business that was owned by HON
• The liability actually remains at HON – rather than own the actual liability, GTX has signed an agreement that states that it will pay 90% of the annual costs of settling HON’s asbestos litigation, up to $175M per year. Note that this makes it in effect an unsecured claim
• The $175M payment figure is what every analyst puts in their model, but it is actually the absolute worst-case scenario as it is a cap
• GTX has the ability to defer payments to HON if making payments would cause it to run afoul of covenants on its debt – so in effect this liability is an unsecured claim that can be deferred like a PIK note

What is also unappreciated by the sell-side community is that new asbestos claims have been declining meaningfully over the past few years – while $175M is comparable to the level of payments made over the last few years, this has contributed to resolving 4,000-5,000 cases per year. Meanwhile new cases filed per year have declined from ~4,000 from 2012-14 to 2,645 in 2017. As a result, total unresolved cases have dropped significantly, and I believe they will continue to do so. Note that in the first chart below, the large drops in 2009 and 2013 are mostly associated with the settlement of much lower severity non-malignant cases.

As a result, while $175M may be in the ballpark for payments for the next couple of years, I believe we should start to see declines of ~$5M per year in payments owed as newly filed and unresolved cases continue to decline. This sounds small, but an incremental $5M in profit every year (asbestos payments are not tax-deductible) adds 2% to EPS growth every year for the next 35 years.

6. Business quality is underappreciated and operating sensitivity to a negative macro environment is much smaller than for other auto companies

While many investors’ first reaction on hearing “auto parts” is to run away, I would make the case that GTX is actually a very high-quality business within the sector and should be much less exposed to the cyclical trends that are causing multiples to derate across the entire auto complex.

• Leader in a niche segment with structural tailwinds
  • #1 Turbocharger brand with 33% global share and 50% platform win rate over the last 3 years
  • Penetration tailwinds should drive 5% CAGR in units sold; historical organic growth has been ~5.5% CAGR going back to 2000
• Highly engineered products that are difficult to “knock-off” or disrupt from the low-end
  • Turbochargers spin at 350,000rpm, generate 25Gs, and reach temperatures of 1900F – these are not commodity components/systems that are easily copied
  • New product iterations like “variable geometry” (ability for the turbine geometry to be dynamically altered to maximize performance at different engine speeds) require even more advanced IP
• Highly visible demand for products
  • Speced into vehicle platforms years in advance and co-develop products with their OEM partners
  • Turbochargers are almost exclusively single-source relationships for a given vehicle platform
  • 87% revenue visibility out to 2022 from awarded + replacement business
• Very high margin profile, driven by high IP content
  • 18-20% EBITDA margins with only ~3-4% capex requirements – highest margin profile amongst all major auto parts companies other than GNTX
• Highly stable margin structure driven by outsourcing of production to (largely) captive supplier base
  • 80% of cost base is variable, suppliers bear the brunt of fixed capital investment as well as absorption risk in a downturn

The impact of having a higher margin profile and higher variable costs in a downturn is underappreciated, in my view. If we compare GTX (~15-16% UFCF margins, 90% variable costs) vs. a typical auto parts supplier (6% UFCF margins, 80% variable costs) in a hypothetical 10% SAAR decline environment, the results are striking – while a typical auto parts supplier would see a 25% decline in EBITDA and nearly 50% decline in UFCF, GTX escapes relatively unscathed, with a 14% EBITDA decline and 17% UFCF decline.

Valuation & Estimates:

While we have limited historical information for GTX given that it is a recent spin-off, the past four years (2015-18E) have shown fairly stable performance, with organic growth averaging +4% and EBITDA margins ranging from 17.2-20.1%. Management also put out financial goals through 2022 which, because they were released pre-spin (and before management’s RSUs struck), if anything are likely to be conservative.

If I take 2018’s revised guidance, grow the top-line by 5%, and apply an 18% EBITDA margin (at the bottom-end of the range as shifting mix from diesel to gas turbos will temporarily reduce margins), I end up with 2019 EPS of $3.17/sh – if I project out continued 5% top-line growth, deleveraging, and margin improvement up to 19% by 2022, I end up with a 14-15% EPS growth algorithm, reaching $4.71 by 2022 (with net leverage now down to 0.9x). I appreciate that GTX will trade at a discount to the market multiple given the cyclicality of the auto sector and its leverage, but it really does not seem like a stretch to put a 10x multiple on 2019 EPS for a $32 stock and a 2+ bagger.

Alternatively, we can look at EV multiples. Auto parts comps at the higher-quality end of the scale (proxied by having EBITDA margins > 15%) trade at 10.7x EV/EBITDA-Capex, with EV adjusted for pensions, NOLs, etc. With 2019 EBITDA of ~$640M and capex at $115M, this would justify a $5.6B EV. Subtract from that $1.5B of net debt, $1.4B of PV asbestos liabilities, and $0.3B of misc. tax liabilities gets you to a target market cap of $2.4B, over ~77M diluted shares gets you to a target price of $31 and again a 2+ bagger.

In terms of a bear case, you would need a very meaningful auto recession to justify the current price. I can haircut 2019 revenues by 10% and take down margins by 200bps, and I still end up with EPS of $1.80. At the current price that would be an 8.6x P/E multiple on below mid-cycle levels of demand. On an EV basis, that would still reflect a 10.8x EV/EBITDA-Capex multiple.

In terms of an upside case, if I grow the top-line at 6% p.a., roll-forward another year and assume GTX gets back to 20% EBITDA margins by 2020, they generate $4.40-4.50 in 2020 EPS and have delevered meaningfully. In this scenario I don’t see why this couldn’t be a $45 stock a year from now when the 2020 guide is given.

Risks:

GTX obviously isn’t without risks – below are what I view as the most meaningful ones to be aware of:

• Cyclicality – auto parts are a notoriously cyclical industry; auto production has fully recovered from 2008-09 and is likely somewhere between peak and mid-cycle
  • Mitigant: GTX’s high margin structure, variable cost base, and secular tailwinds should help them outperform the rest of the auto parts space in a downturn scenario
• Leverage – HON spun out GTX with $1.5B of net debt and a $1.4B asbestos liability, depending on how you treat the asbestos liability, it works out to either 3.2x net debt (net debt / post-asbestos EBITDA) of 4.5x net debt (net debt + asbestos liab / pre-asbestos EBITDA)
  • Mitigant: Asbestos liability is only quasi-debt, as payments can be deferred to maintain liquidity and debt covenants
  • Mitigant: Strong FCF conversion and growth in underlying business should reduce leverage by ~0.5x per year
• Limited history as independent company – hard to verify management’s targets are reasonable
  • Mitigant: Management was incentivized to give conservative guidance as a result of their RSUs striking at the last when-issued price
• Asbestos liability – opaque valuation of liability, hard to determine how long the $175M payments will continue for or what the ultimate magnitude of the liability is
  • Mitigant: $175M is a capped payment that is embedded in sell-side (and my own) estimates – the actual result can be no worse than this from a cash flow perspective
  • Mitigant: Recent history of claims trends suggests this is finally beginning to wind down and is a finite-life obligation not a permanent one
• EV penetration – turbochargers are not present on fully electric vehicles, resulting in content loss if EVs continue to take meaningful share
  • Mitigant: EV is unlikely to be the only solution in the car fleet of the future, GTX content on hybrid cars is likely to be even greater than on gas/diesel which will mitigate some of this impact
  • Mitigant: GTX is in the process of developing and piloting solutions for electric vehicles and autonomous vehicles, with a 5-year investment in cybersecurity and vehicle health management software capabilities, and a team of > 100 software engineers

Catalyst

• FY19 guide
• Continued operational execution
• Clarity on asbestos liability
• Sell-side upgrades
• Delevering