How comparable are aircraft ownership and maintenance costs across airlines?
Accounting for aircraft ownership and maintenance costs
I’ve been wanting to write an article on the subject for some time, in part because I wanted to try to understand what was driving some of the differences we see in maintenance and aircraft ownership costs when comparing airlines.
Depreciation for an airline is dominated by fleet assets and to a lesser extent by major maintenance costs, which are capitalised and depreciated due to their “lumpiness”. Minor maintenance is expensed as incurred. Exactly where to draw the line between minor and major maintenance might differ slightly between airlines, but the sum of these two cost lines should be “comparable”. But there are some big differences in the reported costs, as you can see from the following chart for the twelve months to March 2022.
Ryanair in particular always comes out as super low cost in these areas and I wanted to see if I could understand and explain what is driving this. Are their accounting policies extra aggressive? Or do they get their aircraft at lower prices than everybody else? Or is something else going on?
The reason it has taken me so long to write this article is that accounting for aircraft ownership and major maintenance is one of the more complex areas in airline accounting. My previous attempts to understand the figures have all ended in me drowning in complexity and losing the will to live somewhere in the middle of the analysis, but I am determined to stick with it to the end this time.
Let’s get started.
Why understanding accounting policy matters
In the end all cash outflows related to owning or renting aircraft and carrying out major maintenance will need to flow through the profit and loss account. But with aircraft being long-lived assets and with major maintenance being very "lumpy", exactly when these costs are recognised is quite dependent on the choice of accounting policy and on the assumptions that underpin it. Whilst all the major airlines apply the same international accounting standards, there is quite a bit of “wiggle room” in those to accommodate different approaches.
Also important is how the aircraft is financed, especially whether it is owned or leased. Over time, the accountants have done their best to try to reduce the difference this makes by bringing leased assets "on balance sheet". However, in doing that they have added another layer of complexity and another set of assumptions that need to be made. As well as affecting the timing of how aircraft ownership and major maintenance costs appear in the profit and loss account, those assumptions can also affect in which expense line of the profit and loss account these costs appear.
The timing of expense recognition won't matter that much when it comes to airlines with large and stable fleets, with aircraft evenly spread across all ages. If younger aircraft are "favoured" by a particular policy, that will be balanced out by the older, "disfavoured" aircraft in the overall numbers. However, for fast growing companies, a policy which pushes expense recognition later in the life of an aircraft can significantly flatter overall results, since there will be many more young aircraft in the fleet than old ones.
Also troublesome for the analyst or investor is that different accounting choices can result in costs appearing in different lines of the profit and loss account. Analysts make a lot of use of profit metrics like "EBITDA", "EBIT" and "Recurring EBIT" in comparing airlines. If costs move from maintenance to depreciation, EBIT (Earnings Before Interest and Tax) is unaffected but we get an improvement in EBITDA (Earnings Before Interest, Tax, Depreciation and Amortisation). If costs move from depreciation to interest costs, overall profits are the same but EBIT improves. If costs move from depreciation to profit/loss on disposal, EBIT doesn't change but Recurring EBIT does, since profits and losses on disposal are usually treated as non-recurring.
The accounting and financing choices made by an airline also make a lot of difference to the balance sheet, despite the accountants’ attempts to reduce the impact of leasing. The "Right of Use (ROU)" assets that accountants create when they bring operating leases onto the balance sheet only bring about half the value of a new aircraft onto the balance sheet for typical lease lengths. For an eight year old aircraft on a ten year lease with only two years remaining, the ROU asset and related debt liability would be less than 20% of the equivalent owned aircraft asset at the same age.
Before the invention of ROU assets and liabilities, at least analysts knew that operating leased aircraft were not on the balance sheet and made their own adjustments to get “total debt”, typically by multiplying the annual lease payments by a factor of around seven. For a new aircraft, that would be similar to the ROU approach, bringing about half the total value of the aircraft on balance sheet. But the value brought into the balance sheet would remain stable throughout the life of the lease, rather than dropping away as the remaining lease term shortened, as happens now with ROU accounting. Given that in reality, an airline will have aircraft at varying points in their leases, the ROU approach typically brings far less of the leased aircraft value on balance sheet than the old “multiply the lease payments by seven” approach did.
All of this is before even considering the impact of tax. Many aircraft financing deals are heavily motivated by tax. An airline registered in a country which gives tax advantages to aircraft investment may decide to own its aircraft. The savings it gets will show up in the tax line. An airline which doesn't get such advantages can often still access them by leasing aircraft through jurisdictions which do. Those tax savings will show up in the depreciation or interest cost lines, once again making comparisons between airlines difficult.
None of this would matter much if these costs were small in the scheme of the overall profit and loss accounts and balance sheets, but they are not. They can also vary significantly between airlines. In its most recent quarterly results for September 2022, Wizz reported total costs for maintenance, depreciation and financing equal to 18% of revenues. At the other extreme, Ryanair's costs for the same items totalled only 9% of revenue.
How airlines account for aircraft ownership and maintenance
For the purposes of the numerical illustrations that follow, I’m going to use an example of an aircraft which is acquired for $50m new. Boeing and Airbus will quote list prices for 737s and A320s which are more than double this, but no major airline or leasing company pays anything like list prices, so I think this is a better “round number” to use for illustration.
I will assume that major maintenance is required every 8 years at a cost of $5m. The reality will be messier than that, with some major maintenance required earlier and some later, but we need to keep things as simple as possible if we aren’t to get lost in the weeds. For the same reason, I’m mostly going to ignore inflation and exchange rate movements.
Is $5m every 8 years a reasonable assumption? I’ve seen some maintenance cost estimates for narrow-body aircraft that would suggest a much higher number, given that engine overhauls should I think be included in any definition of major maintenance. However, the work I’ve done reverse engineering low-cost carrier accounts suggest a figure of about that level is what they are using for new aircraft. In any event, that’s the figure I’m going to use for the purposes of this illustration.
I will look at the policies at four airlines, the three big European low cost carriers and my old employer, IAG. I included IAG because they take a somewhat different approach to the other three and I wanted to understand what impact that had in practice.
Let us start with the “simple” case, owned aircraft.
Owned aircraft
At IAG, a short-haul aircraft would be depreciated over 23 years to a residual value of 5% of its original cost. So for a $50m aircraft, annual depreciation of $2.07m ($50m x 0.95 / 23) would be charged to the P&L every year the aircraft remained in the fleet. When major maintenance events occur, those would be capitalised and depreciated over the time to the next check. So in our simplified example, from year 8 onwards, another $0.63m of annual depreciation would be incurred ($5m / 8). At 10 years, the value of the aircraft on the balance sheet would be $29.3m ($50m - $50m x 0.95 x 10/23) and there would be a maintenance asset of $3.75m ($5m x 6/8), adding up to $33.1m.
At Wizz, the initial $50m aircraft value would be “componentised”, which means breaking it down into a $45m value for the aircraft plus a maintenance asset of $5m. The idea here is to reflect the fact that a significant part of the value of a new aircraft is that it can be used for several years without requiring major maintenance. The $5m maintenance asset would be depreciated over the time to the next check, or eight years in our example. That gives an annual depreciation charge for major maintenance of $0.63m ($5m / 8). The charge will cut in immediately, whereas IAG takes a major maintenance cost holiday in the early years.
Actually, the Wizz depreciation charge for maintenance it is a bit more complicated than I’ve suggested. Rather than a flat charge per period, the part of the maintenance asset that relates to the engines is actually written down on a mixture of “per flying hour” and “per takeoff/landing cycle”. As long as aircraft utilisation rates are stable, the depreciation charge will be evenly spread over the eight years. During the year, a bigger proportion of the cost will be allocated to the high utilisation summer months than will be the case for an airline like IAG that depreciates maintenance costs at a flat rate per month. Even on an annual basis, there will have been a divergence between the two approaches during the pandemic, since utilisation rates dropped to very low levels. IAG’s approach won’t have given them any depreciation cost savings, whereas at Wizz and the other low-cost airlines, the way they account means that depreciation costs are partly variable with activity.
The remaining $45m of the initial asset value at Wizz is depreciated to a 50% residual value over 14 years (Wizz is silent about what would happen after that). This gives us an annual depreciation charge for the “non-maintenance” element of the aircraft value of $1.61m ($45m x 0.5 / 14), bringing the total annual depreciation charge to $2.24m. That is an 8% higher charge than would be the case under IAG’s policies in the early years, but 17% lower from year 8 onwards. At year 10, the value of the aircraft would be $28.9m, marginally lower than at IAG. The maintenance asset would be $3.75m, the same figure as we saw at IAG. So for young aircraft, the Wizz approach is more conservative than IAG’s. That’s reassuring given how fast the airline is growing and how young its fleet it. But the two policies converge at about year 11.
Like Wizz, Ryanair’s policy also involves componentising the initial aircraft value. Their policy is to depreciate the non-maintenance part of the aircraft value over 23 years to a residual value of 15% of the “market price of new aircraft, as determined from time to time”. Ignoring inflation and any deviations between what Ryanair paid for aircraft and “market price”, that depreciation policy could be more slightly more conservative than at Wizz. Writing 85% of an asset off over 23 years gives a very similar figure to writing 50% of it off over 14 years (85%/23 = 3.7%, 50%/14 = 3.6%). But whilst Wizz is very clear in the notes to their accounts that the residual they are using is 50% of the initial value of the aircraft, excluding the maintenance component, it is much less clear what Ryanair are doing. Does the “market price of a new aircraft” exclude the maintenance component? If it doesn’t, then with the figures in this example, Ryanair’s depreciation policy is almost identical to Wizz, writing off the aircraft component from $45m to $7.5m (15% of $50m) over 23 years, which works out at a depreciation charge of 3.6% of the initial value, in line with what we saw at Wizz.
Whilst Wizz only writes down engine maintenance assets in line with activity, with the airframe part depreciating on a fixed calendar basis, Ryanair appears to write down airframe maintenance assets in line with activity too. So Ryanair’s accounting policies give it the most variable depreciation costs of all these airlines. I’m sure that helped it to report the lowest losses during the pandemic, as well as making its policy of grounding lots of aircraft every winter less damaging to its profit and loss account.
On the face of it, all three depreciation policies, whilst quite different in detail, end up with very similar annual depreciation charges and asset values for typical mid life aircraft. However, we glossed over a potentially important element of the Ryanair policy. The other two airlines base their residual values on their own acquisition costs, whilst Ryanair uses a percentage of “market prices for new aircraft”. How big a difference could that make? If Ryanair paid $50m and it assesses the market price as $65m, then instead of the $7.5m residual value we calculated before, it would use $10m. That would cut its annual depreciation charge to 3.4%. Or in absolute numbers to $1.52m, giving it a 6% saving in depreciation compared to Wizz’s policy on the aircraft component.
The other area where Ryanair’s policy of using a percentage of market prices to determine its residual values could give it savings relates to inflation. I can’t tell from the notes to the accounts whether the residual values for existing aircraft on its books are reassessed after they have been set. Does the determination of market prices “from time to time” only apply to setting residuals for new aircraft being added, or does it also affect residual values of existing aircraft? If the latter, then the residual values will be constantly increased to reflect inflation in new aircraft market values. Ryanair’s owned aircraft are about 9 years old on average and even historically low 2% inflation would add up to about 20% over 9 years. So by the time our example $50m aircraft is 9 years old, the residual value it is being depreciated to could have gone up to $11.7m (15% of today’s $65m “market price”, inflated by 20%). With only $33.3m needed to be written off over 23 years, you could argue they only need to have provided for $12.9m by year 9 ($33.3m x 9/23), which works out at $1.43m a year. If that is what they are doing, Ryanair’s policy would give it an 11.2% depreciation cost advantage compared to Wizz for an identically priced owned aircraft.
At easyJet, the initial acquisition cost is also componentised. The useful economic life for CEO aircraft (“Classic Engine Option”) was changed from 23 to 18 years in July 2021, which triggered a £50m increase in its annual depreciation bill. In the figures for the 12 months ending March 2022 that I showed at the start of this article, the change of economic life will have been 75% reflected. In theory then, for the CEO aircraft which still dominate both easyJet’s and Ryanair’s fleets, easyJet’s depreciation policy is now more conservative than Ryanair’s. But easyJet is much less transparent about what it assumes for residual values. The notes to the accounts simply say that it uses “market values for equivalently aged assets” to determine residual values. This has many characteristics in common with the Ryanair approach - divorcing residual values from its actual acquisition cost and using undisclosed market values, which would in theory inflate over time. Quite how easyJet’s depreciation charges based on market value residuals for 18 year old aircraft compare to Ryanair’s figures based on depreciating to residuals at 23 years of 15% of the market value for new aircraft compare is not clear. I’d guess that easyJet is taking a more conservative approach, but it is hard to say from the outside.
OK, deep breaths. Owned aircraft was supposed to be the easy case. Perhaps you can now start to see why I’ve struggled for so long to write this article?
Before we move on to the even more complex world of leased aircraft, I think I should summarise the position for owned aircraft. As we’ve seen, the detailed policies at different airlines vary quite a bit. However, the different policies tend to yield quite similar figures for depreciation and aircraft values over time. That shouldn’t come as a surprise, since the accountants and the audit committees of the airlines spend a lot of time cross-checking the figures with market reference points and alternative approaches. When it comes to owned aircraft, I’m reasonably comfortable that the differences between the way these four airlines account for ownership and major maintenance costs don’t generate big distortions.
It does seem to be the case that Ryanair comes out at the more aggressive end of the spectrum in terms of accounting. But I think they can afford to take that approach because their aircraft deals are probably also the best in the industry, and so even a more “aggressive” depreciation policy that writes its aircraft assets down at a slower rate still delivers reasonable aircraft values over time when they are benchmarked against market prices.
If you are still with me, I’m now going to move on to the even more complex world of accounting for leased aircraft.
Leased aircraft
Let me first get out of the way what accountants call “finance leases”. This is where the lease length is comparable to the economic life of the aircraft and the airline retains the whole or the vast majority of the residual value risk. In economic substance, these are almost the same thing as owning the aircraft and financing it with a mortgage and accountants have for a long time put aircraft financed in this way on the balance sheet. For such leases, it is usually straightforward to work out what the interest rate is for the financing element. Interest rates and residual value assumptions may well be explicitly stated in the lease terms, but even if not, the residual value will be relatively small and a long way into the future for such long leases. That means that the lease rate is very closely linked to the interest rate inherent in the lease. In almost all important respects, aircraft on finance leases are accounted for in the same way as owned assets, which we have already covered.
Where things get more complicated is when it comes to what used to be called “operating leases”. These are leases which cover only a part of the aircraft’s economic life. For new aircraft, they usually have a lease term of 8 to 10 years. The present value of the lease payments is substantially less than the value of the aircraft at the start, meaning that the aircraft is still expected to have significant value at the end of the lease. Since the airline can walk away from the aircraft at that time, and the lessor can place the aircraft with a different airline, the “ownership risk” is shared between lessor and lessee.
Such aircraft used to be accounted for by just expensing the lease payments, but a few years ago the accountants decided they should bring them onto the balance sheet. But since the ownership risk is genuinely shared, putting the aircraft into the airline’s assets at full value, as would be done with a finance lease, isn’t really an option. As we’ve already seen, the present value of committed lease payments is a significantly smaller number, so if you put an asset of $50m on the balance sheet but have only $30m of lease liabilities to recognise, your balance sheet won’t balance. They could have created another $20m of liabilities to reflect the fact that the airline had the obligation to return the aircraft to the lessor at the end of the lease, but the accountants went another way. They decided to stick with $30m of liabilities and create an asset of equal value, which they called a “Right of Use (ROU)” asset.
Let’s see how ROU assets are accounted for. As was the case for owned aircraft, major maintenance costs are going to be an important topic. In fact, they are going to be an even bigger issue, since the the lessor cares a lot about the maintenance condition of the aircraft at the end of the lease. An aircraft fresh out of a major maintenance check is worth a lot more than one which is about to need one. The “return conditions” which are written into the lease are therefore critically important. These can vary from lease to lease, depending on the airline and the lessor’s priorities and negotiating position. For the purposes of the illustrations here, I will assume that “as is” conditions have been used and assume an eight year lease term. That means that the aircraft can be handed back just before its major maintenance checks are due and without requiring any return condition work or payments.
Before we can look at the airline accounting, we need to work what would be a reasonable lease rate from the perspective of a lessor for a $50m aircraft with such “airline favourable” return conditions. I’m going to assume that the lessor has the same view on aircraft values as is assumed in the Wizz accounting policies. So that means that at the end of the lease, the aircraft is going to be worth $32.1m ($50m less $5m maintenance element, less eight years of depreciation of $1.61m p.a.). Given the return condition assumptions, there will be no maintenance asset. I’ll assume that the leasing company has a cost of debt of 3.0%, gearing of 80% and a cost of equity of 12.0%, which works out to an average cost of funds of 4.8%. I’ve made no allowance for expenses, but I’ve used a fairly high cost of equity and also ignored the tax benefits that I’m sure any well run leasing company will get. That gives me a lease rate of $357k per month, or $4.3m p.a.
How would that get accounted for by the airline? First of all, you will need to work out the initial lease liability, which is the present value of the lease payments. What discount rate should be used? The accountants say that it should be the interest rate “inherent in the lease contract”, if known, or the airline’s “incremental borrowing rate (IBR)”. You and I know that the interest rate inherent in the contract is 4.8%, but all the airline knows is the lease rate of $357 per month. They don’t know what the leasing company thinks the residual value will be at the end of the lease, or the precise details of how it will be funded, or the leasing company’s costs and profit margin. So in practice, for leases like this, airlines usually use their incremental borrowing rate. In the case of Wizz, this is disclosed in the accounts as 3.4%. So we can now work out the initial lease liability by discounting the lease payments at that rate, which gives $29.6m. Glossing over a few minor details to do with up-front costs and such like, a matching ROU asset of $29.6m is created, so that the balance sheet still balances. That asset will be written down over the life of the lease through an annual depreciation charge of $3.70m ($29.6m / 8), which will remain stable throughout the lease. Interest starting at $1.0m per annum ($29.6m x 3.4%) will gradually decline over the course of the lease as the lease liability is “paid off”.
Let us pause for a moment to consider how the depreciation charge compares for an operating leased aircraft, relative to an owned one. The ROU depreciation charge we’ve just calculated is $3.70m, whilst the owned aircraft charge we derived earlier was $2.24m. It is 65% “more expensive” to lease rather than buy your aircraft, at least using these assumptions.
What is going on? Part of the answer is the difference between the actual financing cost embedded in the lease and the incremental borrowing rate assumed. If we’d used 4.8% instead of 3.4% to discount the lease payments, we have a smaller asset and lower depreciation. So mismatches between the “real” cost of the financing for leases and the IBR ends up in depreciation. But even if we’d used 4.8%, we’d have calculated a depreciation charge of $3.49m, still 56% higher than for an owned asset.
The source of the remaining gap is the way that the residual value of the aircraft finds its way into the depreciation charge. For an owned aircraft, $1m of extra residual value in eight years time decreases the depreciation charge over the eight years by $1m. For a leased aircraft, that extra $1m of residual value gets discounted at the lessor’s cost of capital. Eight years at 4.8%, gives a discount factor of 1.455, reducing the value by 31%.
I’ve only worked out the leasing figures based on Wizz policies. The policies of other airlines are very similar for the core accounting. Ryanair uses an IBR of 2.39%. Neither easyJet nor IAG disclose the precise figure for their IBR, but it looks to be around 4% for both based on the interest expense. These differences mean that for the same lease rate, Ryanair will put more of the cost into depreciation and less into interest costs than Wizz. IAG and easyJet will put more of the cost into interest. The difference a small adjustment in IBR can make to the depreciation charge can be big - just going from 3.4% to 4% cuts the depreciation charge by 15%.
There are also differences in the way that the airlines account for major maintenance when it comes to leased aircraft. They largely mirror the differences in approach that we saw for owned aircraft, such as Wizz componentising the ROU asset into a maintenance asset and an aircraft asset, which IAG doesn’t do. But since both elements are largely being written off over the life of the lease, it makes even less of a difference than it did for owned assets. This article is already far too long, so I’m not going to go into the details.
So why would any airline lease rather than buy?
If leasing aircraft results in much higher costs in the profit and loss account, why would any airline ever lease?
Sometimes, leasing companies can buy aircraft more cheaply than airlines, or get better access to delivery positions. Leasing companies can borrow more cheaply than some airlines. However, I don’t think any of these difference apply in relation to any of the airlines we are discussing here.
More important is tax. As I pointed out earlier, leasing can often be a route to access advantageous tax benefits that are not available in an airline’s country of registration. In my calculations earlier, I ignored tax and tax benefits can bring down the effective funding rate and resulting lease cost significantly. It is perhaps no surprise that the airline which makes the least use of leasing is Ryanair. It is registered in low tax Ireland, the same country where many of the biggest aircraft leasing companies are based.
Another reason is risk management and flexibility. Being able to flex capacity downwards or switch to different aircraft types without facing the challenge of selling or subleasing aircraft can be a valuable thing. Once again, you can understand why Ryanair doesn’t much value this aspect of leasing - its whole business model is based on operating a single aircraft type and it generally maintains or even grows capacity in a downturn to take share from weaker rivals.
The final reason is that leasing aircraft reduces capital requirements. There is a reason that the leasing companies need to earn a return on the residual value at the end of the lease - they need to finance it in the meantime, something the airline no longer needs to do if they lease. That shows up elsewhere in the profit and loss account - in the interest costs. For a leased aircraft, we already know the interest costs that will be charged. It is given by the incremental borrowing rate times the outstanding ROU lease liability. For an owned asset, the interest cost will depend on how it is financed. I am going to assume that it is financed with borrowings at the same IBR rate. In theory, that’s what the IBR should mean. I’ll also assume that the loan to finance the aircraft is repaid over the aircraft life in line with the annual depreciation charge. So I can calculate a notional interest cost by multiplying the IBR by the book value of the aircraft.
That gives me the following comparison for the P&L cost of owning aircraft versus leasing them over an 8 year term, under both IAG’s and Wizz’s accounting policies. I restricted myself to those two because they are the only two which give full disclosure about their residual value assumptions. I will also remind you that the lease contract we modelled had “as is” return conditions. If the aircraft had to be returned in full or half life condition, there would be additional maintenance costs charged, but that would be offset by a lower lease rate. That would show up in lower interest and depreciation charges.
Leasing aircraft still comes out as more expensive than owning them, under these assumptions. There is a cost for the flexibility and risk sharing that a lease provides. But you can see the “total P&L” difference is much lower than the operating cost impact.
I hope I have also been able to demonstrate that, even for identically priced aircraft, where the ownership costs end up between depreciation, interest and maintenance depends a lot on accounting policies and how the aircraft is financed. So take care when comparing airline results where these things are not aligned.
Final thoughts
I had expected to find that some of Ryanair’s advantages in reported depreciation and maintenance costs would be driven by more aggressive accounting. I still think that might be the case, but given the level of scrutiny given to the subject by the auditors, I don’t think the differences are that material. Overall, I think their low costs in this area are much more likely driven by having better terms from aircraft and engine suppliers.
Of more importance are differences in reported costs due to how the different airlines finance their fleets. A big part of why Ryanair comes out best is that almost all their aircraft are owned, not leased. That comes at the cost of consuming more capital, but the airline’s consistent profitability over the years means that its balance sheet is one of the strongest in the business and it can raise funds very cheaply. In fact, it doesn’t even incur higher interest costs than other airlines - quite the opposite in fact. That’s because it mainly buys its aircraft from free cashflow. And I’m sure that being based in low-tax Ireland helps too.
Part of the reason for writing this article was to improve and check my own understanding. I know there will be people who understand this area much better than I do, so if I have got any of this wrong, or I’ve missed important differences between the airlines, please let me know in the comments or via email.
If you’ve made it all the way to the end without falling asleep or giving up, well done and thank you for your patience.
