Description

Note: I wrote this up but can't post from office (no Google docs access).  In the interim, fl3474a posted a long on AerCap Holdings.  It's well worth a read.  For what it's worth, I moved capital from AER to AL after the ILFC transaction for the reasons outlined below.  However, I'm only coincidentally posting this right after AER was posted, and I don't disagree with the argument for AER as a long…

Air Lease (AL) was written up a few years ago by UCB1868, and I refer you there for good background on the company.  Since 2014, the stock is flat while the underlying business has significantly increased in value.  I view the current entry point as quite compelling.

Basic structure of their leasing business: order new planes from OEMs 5-7 years in advance, sign up and lock in the lessees a couple years in advance of delivery.  Lever up the planes 2:1 debt to equity and lease them out on triple net terms for 10-12 years, sell the planes after about 8 years with lease terms remaining (and prior to the first big maintenance capital outlay).  Current pre-tax ROE's are in the high teens, and the company will pay no taxes for the next decade at least.  You can buy this currently for a 15% premium to book.

So why is this a high ROE business when it seems like it should be closer to commodity financing?

Earnings are driven by the spread between the lease rate and your costs: depreciation, interest, lease costs and corporate G&A.  Airlines are willing to pay the spread due to some combination of the following: reduced capital intensity, lessor tax benefits, residual risk reduction, bulk purchase discounts (at least relative to smaller airlines), fleet flexibility, availability of hard to procure slots on popular planes, and added value from fleet management consulting.  Sale leaseback transactions solve most factors, but Air Lease gains some extra spread due to its order book, data, industry expertise (good practical examples of benefits to airlines in their last few investor days), and willingness to make big bets on future plane utilization.

On the cost side of the spread, Air Lease is run quite efficiently.  How many other companies can do $800M in EBIT with 76 employees?  And from there they can double the fleet with ~20% increase in workforce.  Udvar-Hazy learned a lot of best practices over his lifetime in the business, and he's been able to build his current iteration from the ground up applying all of them.  Some keys: limit fleet to the most in-demand and liquid models, don't own old planes where more stuff can go wrong, keep all assets unencumbered to maximize flexibility, set incentives to maximize productivity from a lean staff.  Leasing expenses plus G&A total an industry low 7.5% of revenue (for comparison, peer AerCap (AER) even ex- one time charges has run 18-19%.  This inflated when they bought the aged ILFC portfolio from AIG, but even before that their costs ran 10%+.)

It is difficult to compare the lessors apples to apples, due to factors like fleet age (AL youngest portfolio has higher margins and lower ROE because of depreciation relative to book value), leverage (AL runs the lowest, which obviously lowers ROE, tax rates (not big differences on cash tax, but AL much higher on GAAP taxes), and strategy (AYR for example is "value investing", trying to buy low and sell high on individual assets.)

A few other highlights of the business:

Tax.  This is meaningful for all aircraft leasing companies, as they can take advantage of Irish law designed to help them.

Key provisions: Book depreciation at 3.4% vs tax depreciation at 12% is a huge difference; 1031 aircarft exchanges for tax deferrals, and 12.5% Irish tax rate on capital gains.  AL parent is a US company, with subsidiaries able to capture all of the tax advantages.  Despite reporting GAAP financials with US statutory tax rate, cash taxes are zero, and will remain so for at least several years longer than they keep growing.  It's worth noting for comparison purposes that AER reports a tax rate of 14-15% and AYR <10%.  (Note: if a US corporate tax rate somehow gets done, it immediately increases book equity.  That's sufficient to hold a half dozen more planes while keeping the leverage constant…)

Order book.  They are big buyer of planes.  They're a top 10 global buyer of both next generation narrow body planes, as well as the new A330, A350, and B787s.  Large or concentrated airlines buy more of some models, and AerCap and GECAS also have very sizable order books.  But as a % of their current book value, AL is huge.  Looking at deliveries 2019 and beyond (before which they already have lessors signed up, so the lease pricing risk is off the table), I estimate AL is committed to 1.7x current book.  AER is committed to about 0.4x.  It's interesting that AER CEO Gus Kelly pretty consistently talked about not needing a big "speculative" order book and doing well in the sale leaseback market, until he bought ILFC.  With the huge backlogs currently in place, there is embedded value here that is disproportionate compared with other lessors.  I don't have any way I find reliable to put a $ value on this, and anyway AL is not going to monetize this with a meaningful level of sales of order slots.  I view it more as a positive for demand generation, and some level of pricing power.

Asset Management fees.  Over the past couple years, AL has begun to build out two distinct asset management opportunities.  Blackbird relates to the deals they see above their capital capacity constraints; they can offload extra purchases into a JV where they have an equity stake.  Thunderbolt allows for sales of aircraft at the end of AL's ownership period around 8 years, into a managed vehicle where AL continues to service the leases.  In both cases they earn an asset management fee and have additional upside potential.  These currently generate $10M/yr of high margin revenue without incremental capital, and future growth is supported by high demand from credit investors.  Note that Thunderbolt has an earnout structure based on residual values; AL believes the industry standard Ascend benchmarks are wrong.  If AL is correct, upside is $59M seven years from now and the structure is likely repeatable.

Valuation.

When UCB wrote this up, part of the case was a justified book multiple expansion, from then 1.4x to 1.8x based on pre-tax ROE expanding to 20%.  The ROE expansion has proceeded on target.  However, the multiple hasn't followed, and AL's stock spent significant time trading under book.  So why is it different now?

The one thing I can point to is earnings.  The P/E was mid-teens 3 years ago, and <11 now, going to 8-8.5 in a couple years, with a fully loaded GAAP tax rate.  On pre-tax earnings it's <7x with earnings growing ~10%.  I can't prove it, but unless something blows up, I think this will re-rate.  Until it does, with these returns I don't mind waiting.

Risks

What could make this go off the rails?  A few thoughts on what I see as the key areas of concern.  A number of these overlap, but I'll attempt to isolate factors where possible.

1.       Competition.  The Chinese are coming, and pockets are deep.  The risk is that spreads compress as a result.

New entrants into this business generally have to start with sale-leaseback transactions, which have always been highly competitive based on price.  In demand aircraft get some level of lease premium.  It takes time to build up an order book, and with Boeing's backlog you couldn't do this in a meaningful way at least until the middle of next decade.  The larger pockets of Chinese capital are in the very early stages of building up this order book inside China (almost 25% of AL's book value, so this is a real threat.)  Outside China, tax and other issues make this more difficult.  Still, some level of spread compression eventually is a reasonable outcome.

2.       Order book risk

The flip side of the order book is the risk taken.  AL is exposed in the 2019-2023 period where they've committed to orders and don't have a lease agreement in place.  As discussed above, AL has a very high amount of exposure here relative to other lessors.

Right now, airlines are as or more profitable than they've ever been.  Production schedules are supported by growth in passenger miles traveled of around 5%, which comes substantially by global entry into the middle class (pretty much no one flies with annual income < USD$10,000, but incremental income above that has strong correlation with flights per year.)

There are various factors that could disrupt this picture, and I can't dismiss any: global recession, increased terrorist activity, oil price spike are three big ones.  Still, the annual growth in passenger miles is one of the more stable economic data sets I've ever seen.  If it was to significantly inflect, AL has some buffer from highly liquid  models.  Exposure from 2019 and beyond as % of list price: A320 family 27%, 737 Max 27%, 787 9/10 22%, A350 13%, A330 12%.  Total narrow is 54%, wide 46%.  (For comparison, AER order book in the same time period is 91% narrow body.)  

If you think a wide body glut is coming and prices on new production will crash, AL would be a tough place to be.

3.       Residual risk

Lessors including AL generally depreciate straight line at 3.4% (25 year assumed useful life to a 15% residual.)  Certain consultancies generally report declines in current market values (CMV) of planes higher than 3.4%, and substantially higher for some models in some years.  Ascend has argued that the depreciation period should be 20 years rather than 25.  If that matches the actual economics, that would hit current book value by ~3%, and pretax margins would be a little over 2% lower.  On some older planes, losses would be more significant, and reporting such transactions could hit the stock price.

AL has argued for some time that Ascend CMVs in particular are wrong.  Their claim is credible to date, as they have consistently realized gains on sale, contrary to the Ascend CMVs.  The recent Thunderbolt transaction provides some interesting data, which appears to square with AL assumptions.   They sold 19 of their old planes, averaging 12.5 yrs old, with leases expiring on average in 1Q 2020, at a gain.  A snapshot of what they sold:

Note: My book estimates throughout are that AL purchased at 55% of list.  The discount paid of course varies somewhat by plane, and potentially the discount has increased as AL has scaled, but looking at either annual fleet data or transactions imply that number is pretty close.

But maybe they're cherry picking dispositions… This is hard to disprove so far, due to not enough data.  In the current market, if you were doing that, you'd want to hide problems with widebodies (see the whole fallout when Delta CEO Richard Anderson said he bought a used 777 for  <$10M…)  Since 2014, they've only sold 4 widebodies (not counting intra-quarter trades).  In the Thunderbolt transaction, they only sold one A330, and it was only 10% of the transaction.  They could hide a meaningful mark to market loss in here.

Capping the risk: wide body lease expirations in the next 5 years by model, with estimate of % of book value.

After 5 years, they obviously do have other widebodies.  For reference, 777-300ERs are 25% of book but are younger than the average fleet age.  A330s are 15% and about 5 years of age, inclusive of the above expirations.  They have no 767s or 777-200's outside of the two in the above chart.  Their exposures are liquid planes, still in production.

4.       Airline bankruptcies

In general, owners of aircraft have very high ability to repossess collateral in bankruptcy (e.g. section 1110 of US  bankruptcy code, and other jurisdictions have equivalents).  AL, AER, and GECAS all have examples of specific recovery of assets from recent idiosyncratic BKs.  Importantly, AL has high ability to understand whether a particular airplane will make money for a particular airline, and so they have a number of examples of uninterrupted cash flows in BK.  Combined with proactive management, and concentrated fleets in popular models, the risk from one off bankruptcies is pretty manageable.

What about something more widespread?  2016 airline EBIT margins were ~11%, so industry is at peak health helped by rationalization and high load factors, lower oil prices, and healthy demand growth.  How worried do we need to be about an industry downturn?

It would depend I think on the reasons.  BKs resulting from cost structure problems due to labor issues aren't issues with the assets.  If there was another wave of legacy carrier BKs (say in Europe), due to pressures from low cost carriers, that also is not unmanageable for lessors.  The concern would be from a big and lasting oversupply of planes.  The Boeing/Airbus duopoly seems to be acting basically rationally assuming that the passenger miles growth stays somewhere near 4-5% (deliveries next five years will be ~7% of installed base, offset by fleet retirement  running maybe 3-4%) .  The list of concerns above (in the order book section) could potentially take the demand number down.  If you had a sudden and prolonged downward inflection in passenger miles, bankruptcies basically accelerate the problems for a lessor: you get the planes back but have to re-lease them into a saturated market.

The longest interruption example we have historically was post 9/11.  At ILFC at that time, 10% of the fleet was subject to BK proceedings.  Some of that had to get remarketed, and some was subject to repricing.  The press at the time reported peak to trough drops in lease rates ranging from 20-40%, bottoming out around May 2003.  ILFC briefly hit an unutilized rate of just over 2%.  So upper limit on impact would be 10% of fleet at 40% discount plus 2% idle, or a 6% drop in revenue (falling straight to pretax income, given that costs are fixed relative to price drops/idle assets.)  

ILFC ROA's did drop from 3.5% in the 3 years to 2001 to 2.6% in 2003.  Adjusting for interest rate rise, if you kept that 3.5% in 2003, revenues would have been a little over 7.5% higher.  Udvar-Hazy said on an AIG call at that time that the biggest impact on their returns was not from the BKs but from having to place order book into a depressed market post 9/11.  To further isolate, they would have some level of impact from just ordinary course lease expirations.  If you accept that, impact from the BKs has to be less than half of 7.5% to revenue, and perhaps significantly less.

If you used 3% of revenue as a reasonable worst case from BKs in a 9/11 type event, that would hit pretax profit by ~7.5%.

5.       Liquidity/Credit

The thing that really hurt ILFC wasn't declines in lease rates, but having AIG as parent during the financial crisis.  Udvar-Hazy sold to AIG back in 1990 because of the benefit of having that credit rating and accessing low cost debt.  He then got pissed about being starved for capital, and basically vowed that wouldn't happen again.  Key elements of strategy to make sure they don't have to make forced sales:

o    Industry low 2.5:1 D/E level, and they manage the balance sheet to that level

o    Interest rate exposure low: 80% fixed rate debt, forward lease contracts have interest rate adjustors

o    Flexibility: over 90% of debt is unsecured, and well balanced maturity scehdule.

o    Liquidity of ~$3B, against about $2B in annual credit needs (half from maturity schedule, and half from business growth).

o    Expanded ABS market access.  They have added to credit protections and transparency, which has increased size and number of bidders on their offerings and leads industry.

In an environment where you have credit markets freeze up combined with falling asset prices/lease rates, if it lasts long enough, you eventually have to sell stuff you don’t want to sell.  If push came to shove, I think it's clear management would prioritize maintaining the credit rating above near or medium term maximizing of returns.  If a crisis happened such that you couldn't get acceptable funding even with an I-grade rating, after say 9 months I'd expect them to be selling into whatever the market is at that time.

6.       Capital Allocation and Incentives

One thing I don't like is that AL did not buy back stock, even when their P/B got as low as 0.8x in early 2016.  I don't see how anything else was a better use of capital than buying their own fleet at a discount like that.

Udvar-Hazy owns $200M of AL shares, so it's difficult to argue that he doesn't have skin in the game.  But he has a lot more money than that, and so you could argue that maximizing per share returns is not his 100% focus.  I would venture that he's likely motivated more by building the largest and most profitable leasing company, and share buybacks fall in the category of slowing down that progress.

You do get predictability and competence from management, and those two aren't worth nothing

Catalyst

None, just cheapness, and earnings