Garrett Motion splits 60% turbo manufacturing market share with one other dominant player: BorgWarner.
Turbo penetration will increase as automakers struggle to meet stringent CO2 standards with smaller engines and hybrid power trains.
Battery Electric Vehicles (BEVs) are the major long-term threat to turbos, and uptake estimates of 20 to 40 million by 2030 indicate -0.3% to +1.1% unit growth in non-BEV autos.
High tech turbos seem safe technologically and competitively despite short- and long-term unit sales concerns.
This article was selected to be shared with PRO+ subscribers - find out more here.
Garrett Motion (GTX) is a big player in making turbos for diesel and gasoline passenger cars as well as commercial vehicles and aftermarket sales. Investors are worried about both a cyclical downturn in auto sales in the short term and secular decline in turbos as automakers shift to BEVs over the long term. I think these concerns are overstated because turbos are part of the path for Original Equipment Manufacturers (OEMs) to meet CO2 standards, and BEV uptake doesn't indicate a rapid or significant decline in turbos for at least a decade, and perhaps not even then. Garrett's price reflects most of the bad news and little potential upside, setting it up as a potentially good investment.
Garrett Motion is a dominant designer and manufacturer of high-tech turbos (along with BorgWarner (BWA)). It was spun out of Honeywell (HON) along with a lot of debt, and significant tax and asbestos liabilities (more on this below). Technological complexity and volume production protect Garrett both from competitors and OEM backwards integration. Turbos allow smaller engines to roughly match the performance of larger displacement engines, allowing car manufacturers to lower CO2 emissions, meeting higher regulatory standards around the world, especially in Europe and China. Turbos also allow hybrids, both plug in and not, to approximately match larger displacement performance with much lower CO2 emissions.
A bit over 30% of Garrett's turbos go into light vehicle diesel engines, while another 30% goes to gasoline, but diesel turbos in light vehicles are rapidly declining (thanks to VW and others' diesel-gate scandals) and gasoline turbos are growing just as fast, allowing Garrett to maintain market share and unit volumes. 20% of Garrett's turbos go to commercial vehicles, which are likely to transition to BEV much later or perhaps not at all due to inadequate performance for some time (hydrogen fuel cells and compressed and liquid natural gas are potential competitors to BEV for commercial uses). 12% of Garrett's turbos are aftermarket, which won't go away any time soon as it takes decades for existing autos with turbos to drop out of the on-road inventory.
Ford (F) (13% of revenue) and Volkswagen (OTCPK:VWAGY) (8%) are its two largest customers, and Garrett serves over 60 OEMs in total; its top 10 customers account of 60% of sales, so there is certainly some buyer power and customer concentration issues. Design wins with auto manufacturers tend to start around three years before production begins, runs through the product's life-cycle, which can last years, then remain in production as replacement is needed over the remaining life of those vehicles. This means Garrett has a lot of visibility into unit volumes in the short to intermediate future. The downside to this setup is that it also allows OEMs to drive a hard bargain by requiring pricing declines over vehicle lifespans, thus compelling suppliers to decrease costs to maintain margins, which Garrett has successfully done historically. Garrett claims 80% of their costs are variable, and that's because they've shifted production to suppliers, meaning it is hit less when inevitable auto downturns strike.
This is a good business, but it is cyclical, and the BEV threat is real even if perhaps overstated. In addition, Garrett is getting ready for life after gas and diesel turbos with: e-turbos, two-stage electric compressors for fuel cell electric vehicles, electric compressors for mild hybrid vehicles, e-boosting technology, and a push into connected vehicles with cyber security, OEM diagnostic and prognostic solutions, and fleet early warning systems. These nascent efforts account for a mere smidgen of revenue and span the range from technology closely related with turbos to having little or nothing to do with Garrett's core competencies. Garrett has a lot to prove in this space before it will be taken seriously. Too, its demonstrated expertise in turbos and accompanying electronics won't easily or automatically translate into equivalent unit volumes, pricing or margins in the same way, so it's a real risk if/as BEVs become the platform of choice. But Garrett's position is not a lost cause, either.
Honeywell ended up with asbestos liabilities after merging with Allied Signal in 1999, coming from its Allied Chemical, NARCO and Bendix subsidiaries. When Honeywell decided to spin-out Garrett, it also decided to offload part of those liabilities onto Garrett. The company must pay a max of $175 million a year but can pay less if Honeywell's liabilities end up lower, as they have been in the last 2 years, partly due to lower payouts and partly due to insurance coverage. Garrett is only responsible for 90% of those liabilities after insurance deductions and is limited to the $175 million per year specified above. These liabilities go away after 2048 or if the payments required decline below $25 million a year for 3 years. This liability can also be delayed if debt covenants are at risk of being violated, which reduces Garrett's liquidity risk, but that liability would accumulate and still must be paid out of future profits. This is a limited liability, but a painful one, and hits the P&L every quarter. Garrett's accounting for this liability, to make matters worse, is considered a control issue because of weak disclosure from Honeywell, which is now under investigation for its liability reporting with the SEC (SEC Opens Investigation Into Honeywell's Asbestos Accounting). Ouch, but not the end of the world.
Honeywell also saddled Garrett with a bunch of debt, which is a bad position for a cyclical business going into unit declines (more likely to get worse before better, especially globally). The debt principal Garrett owes is $23 million in 2019, $28 million in 2020, $47 million in 2021, $65 million in 2022, $231 million in 2023, and $1,234 million thereafter. Interest payments are projected to be $51 million in 2019, $50 million in 2020, $50 million in 2021, $47 million in 2022, $46 million in 2023 and $93 million thereafter.
As if that weren't enough, the corporate tax cut of 2017 burdened future tax repatriation liabilities to Honeywell, and Garrett will have to pay its portion. That's $19 million a year from 2019 to 2022, $36 million in 2023, $48 million in 2024, and $60 million in 2025.
Put all that together, and you've got the potential for Garrett to owe $268 million in 2019, $272 million in 2020, $291 million in 2021, $306 million in 2022, $448 million in 2023 and $1,550 thereafter. That's no problem if Garrett has operating income of $450 million every 12 months, but you can see there is little margin for error if sales come in lower, which seems likely with global auto units slumping. It would be wise for Garrett to term out its debt maturities to reduce the possibility of owing lump sums at disastrous times. With bigger principal payments not starting until 2023, it has 5 years to potentially work this issue.
Garrett's CEO, Olivier Rabiller, led the spin-off from Honeywell, was the CEO of Honeywell Transportation Systems before, has been with Honeywell since 2002 and was with Renault for seven years prior to that. Rabiller has a European background, which is important because that is where Garrett sells most of its turbos.
Garrett's former CFO, Alessandro Gili, had come from Ferrari having worked its spin-off from Fiat Chrysler, but abruptly decided to leave the company as of September 30, 2019 "to pursue other interests". This doesn't look good no matter how you cut it. The most generous interpretation I could come up with was that he was hired to do the spin-off and that's all he intended to do. More unflattering interpretations are that he sees something fundamentally wrong with the business, was frustrated by its financial complexity, or even saw roaches and didn't want to stay around to see how many there were.
Executive pay is high, from $2 million to over $7 million for the top five executives. Short term compensation is based 75% on company performance (50% organic growth, 50% adjusted EBITDA), and 25% individual performance (discretionary). Long term incentives are stock options/restricted stock units/performance plan/performance restricted stock units. Executives own $420,000 to $750,000 of stock, probably all equity awards, which is but a small fraction of what they are paid each year. In other words, they're hired hands, not owner/operators. Oh, and they haven't been buying the stock as it has hit recent lows, either.
I'm using a sales per share figure of $40.80, which is trailing twelve-month sales of $3.2 billion divided by 79 million shares (as usual, I'm counting all outstanding share equivalents). I only have 4 ½ years of data, but sales per share have been as high as $43.34 and as low as $38.05, so my assumption seems in the right zip code. I looked at BorgWarner's (Garrett's closest comparable) sales trend over the last 27 years and used this as a proxy to see how far off trend Garrett's sales could be low or high, and Garrett's sales seemed on trend relative to that standard.
I'm using an 8% net margin as my mid-cycle estimate, which is a little higher than the 4 ½ year average of 7.6% and median of 7.7%, but a slightly higher number makes sense considering the lower tax rate post-2018 and potential of paying down debt and lowering interest cost over time.
Putting together my sales and net margin estimates give a mid-cycle estimate of earnings per share of $3.26. I'm estimating sales growth of 2.6% to 8.1%, which depends on unit growth of turbos being counteracted by unit declines due to BEV uptake, which I think will be slow over the next five years and accelerate after that. This range also assumes gasoline and hybrid with internal combustion engines continue to grow faster than overall auto unit growth over the short to intermediate term.
I'm estimating a margin growth range of -4.5% to +1.8%, just allowing for the fact that margins have more room to decline than climb in a cyclical industry facing unit growth headwinds, both cyclical and secular, while also allowing for the possibility that turbo pricing and value-add in addition to price efficiencies can allow for possible margin expansion. Debt paydown is a significant swing factor here.
I'm estimating growth in share count of 0.7% to 1.4% because I don't think Garrett is or will be able to buy back shares any time soon (it would be better to pay down debt and other spin-off liabilities anyway). Garrett has been growing share count, in fact, and I expect that to continue going forward due to share awards primarily and short-term financing needs potentially.
Putting all those assumptions and estimates together and I see my $3.26 earnings per share showing a range of 3.3% declines per year to growing 9.2% per year. Hanging multiples of 4.1x to 14.5x on that estimate ($13 to $47 price target range) and the current price around $10 looks like 3.0x or lower, which is below my value range expectation and implies a very negative assessment. Garrett has plenty of warts from liabilities to secular headwinds, but that price seems to factor in all the bad news and none of the upside potential.
$10 per share times 78.829 million shares is $799 million in equity value, plus $1.566 billion in debt yields an enterprise value of $2.365 billion. I considered including the asbestos liabilities in that debt sum but decided against it due to Garrett's ability to delay those costs if debt covenants become an issue. Trailing twelve-month EBITDA is $558 million, so an EV/EBITDA multiple of 4.2x. There's too little stock price history to get a good read on multiples, but 4.2x to 5.6x is the range it's been trading around over the last year, so the low end of last year's range. I think that's cheap and would consider 5x to 8x ($16 to $37 price target range) a more reasonable range given its potential growth and financial characteristics relative to EBITDA. Just for reference, BorgWarner's EV/EBITDA multiples over the last decade have been 7x to 12x; granted, BorgWarner has fewer liabilities, broader product offerings and a better history, but it gives you a flavor for how cheap Garrett is to its closest comparable.
Investor sentiment is low, I think, because unit production of autos is down, and because of the threat of BEVs in the future. What could change that sentiment? First, higher unit production of autos and turbos, in particular, could sway investor sentiment, and over time the realization that turbo units probably won't decline as much as many fear over the short to intermediate-term. In other words, I don't think investor sentiment will change until they see it in the numbers, which could happen fast if there is a cyclical rebound, or more likely will take time as quarterly numbers come out. Also, Garrett could term out its debt, find ways to lay off its asbestos liabilities, broaden its product portfolio to make it clearer it can compete more broadly in hybrid and BEV options. Any of those things could lead to sentiment change.
Looking at BorgWarner, you can see that investors could, perhaps, react favorably to better than expected results. BorgWarner has a broader product portfolio, including more products that will go into hybrids and BEVs, giving them a sentiment uplift that reduces investor fears. Also, BorgWarner isn't saddled with as much debt or tax liabilities and just sold off their asbestos liabilities. As far as turbos are concerned, I don't think Garrett and BorgWarner's businesses are performing much differently, but BorgWarner is more diversified, less leveraged, and has fewer liabilities.
Risks abound for Garrett, starting with the cyclical downturn it currently faces. The light auto market had been doing very well up until 2018, but that uptrend is now pointing down. Most industry growth in light autos comes from Asia and, at that, mostly from China, so the heavy downturn there starting in 2018 and Europe's slump - where most of Garrett's business comes from - is putting pressure on short term unit growth and margins. Added to this, its significant business in North America is also under pressure, but less so than Asia or Europe. Given all that, a cyclical downturn could be hard on Garrett, especially if long or steep, and considering Garrett's liabilities.
Which brings up the second biggest risk for Garrett: tax and asbestos liabilities as well as a high debt load - all topped off with the cyclical downturn. It would be one thing to see revenues and profits turn down during a cyclical crash, it's quite another if you have a raft of liabilities to pay during that downturn. I think the odds are against this dragging Garrett into bankruptcy (low odds doesn't mean zero), but it could turn into negative earnings and an even lower stock price.
Unit growth and turbo penetration are key to Garrett's economics. If units grow as more gasoline and hybrid cars install turbos to meet higher CO2 standards, it will be good to great for Garrett. If units decline due to cyclical or secular reasons, or because alternative technologies to turbos are used to meet CO2 requirements, then that would be bad for Garrett. Because of long lead times in auto production, Garrett will see this coming for years in advance, so they could potentially adapt by making cost cuts or adopting new technologies, but that will not remove the risk or sting of unit declines or decreasing auto penetration.ventional turbos. Even optimistic BEV uptake projections would not cause big declines for Garrett over the short to intermediate time frame, but it's possible that BEV uptake takes off due to faster consumer adoption, government regulation, or lower BEV production costs. This would have a big impact on Garrett's multiple and fundamentals over time, so it bears careful watching.
Pricing pressure would be another threat. This could come from 1) BorgWarner, new entrants or other competitors, 2) greater pressure from OEMs that are themselves feeling a strong need to lower costs, or 3) the success of competing technologies. All the above could lead to lower turbo pricing that would, in turn, hurt profits and lower growth rates.
A final threat to consider is that margins could be squeezed due to the pricing pressures mentioned from OEMs. Too, cost burdens from below could surface in terms of raw materials, but also suppliers that Garrett outsources to in order to keep variable costs at 80% of revenue. Lower margins could easily trash my model and make Garrett anything but cheap.
Garrett Motion has plenty of warts: a cyclical downturn with multiple liability headwinds, the potential for lower unit growth both cyclical and secular, and potential threats to pricing and margins. Despite these unpleasantries, Garrett is well positioned with technical expertise, a duopoly with strong market share, long lead time relationships with OEMs, and very respectable profit margins (at least, for an industrial business). Added to this, the company is priced very cheaply on the cyclical and secular risks highlighted above. It's entirely possible the company is priced too cheaply, and thus an interesting investment opportunity.
Disclosure: I am/we are long GTX, BWA.
Additional disclosure: © Copyright 2019 Athena Capital Management Corp. All rights reserved. Permission is hereby granted to electronically link, forward, or store this document in its entirety or to quote passages as long as source is attributed to "Michael Rivers, CFA, Athena Capital Management." Nothing in this letter should be considered investment, financial, tax, or legal advice. The opinions, estimates, and projections contained herein are subject to change without notice. Information throughout this letter has been obtained from sources believed to be accurate and reliable, but such accuracy cannot be guaranteed.