Saturday, February 7, 2015

Hawaiian Holdings

This may be the last time I write something about Hawaiian Holdings. It reported full year earnings a few days ago and provided guidance for the year ahead. Its guidance provided for fuel costs of $1.90 to $2.00 per gallon. 

I am instead going to plug in a fuel cost of $2.30 per gallon (corresponding with $75/bbl for Brent) so that 2015 looks something like this:

and, consequently, the end of year picture would look like this absent a return of capital to shareholders:

The company has begun to buy back the convertibles. The diluted share count at the end of 2015 should therefore be in the neighborhood of 55 million.

$2.30/ gallon in 2015 and $2.80/gallon on average in the years beyond (corresponding to $90/bbl Brent) implies the following range of values for Hawaiian's shares:

I think excess cash plus 10x FCF is fair and I therefore wouldn't be surprised if the shares traded at ~$37 within a year. 

In the medium term, we should also see some improvement in HA's operating performance as the international business matures and its load factor rises from the low 70s to the mid 80s.

This is what HA has looked like on a segment basis:

Note: DCASM = cost of operating the aircraft divided by available seat miles and expressed in cents.

and I'm suggesting that its future will likely look something like this:

A $35 to ~$50 range implies that Hawaiian is an attractive investment in its own right -- that is, apart from its usefulness as a hedge against the oily names in my portfolio. 


Supplemental info:

HA's cost competitiveness:

Note: CASM figures are adjusted in order to allow for like-for-like comparisons of competitiveness. Short haul flights are more expensive than long-haul flights -- more labor, less flying time, more maintenance, etc.
 Untapped pricing power?

Capacity discipline by domestic airlines?

 Room for improvement in the international business:

edited Feb 9

Disclosure: I am long HA


  1. Hey red,

    Thank you for the update.

    Would you please explain the difference between the "stage-length adjusted to 1000 miles" CASM/RASM and the unadjusted figures?



  2. Hi Michael,

    Sure. The idea is to arrive at a like-for-like comparison of unit costs. Its a lot less expensive to fly long distance routes than to fly short haul routes: less time at the gate /more time in the air; less passenger and aircraft servicing costs per hour; loading and unloading; etc - you get the idea.

    So, for example, unit costs in HA's international segment are going to be less than unit costs in its interisland segment but it would be nonsense to say that international was more cost competitive than interisland.

    When adjusted for a particular stage length -- sometimes 1,000 miles is used and sometimes 1500 miles -- one can see who's who and what's what much more clearly.

  3. Hello Red,

    dont you think, they are more in position of excess debt than excess cash?

  4. Debt is excess if it isn't manageable in a downturn or if it exceeds a reasonable assessment of the value of the enterprise.

    The company has laid out, in its Dec shareholders presentation, what it thinks its liquidity and debt/EBITDA requirements are, given the nature of its business. (It's at their website and you can listen to it and judge for yourself).

    The major domestic carriers -- Delta,American and so on -- are cyclical. Maybe they survive a recession because of the new capacity disciple in place but maybe they don't. I don't have an opinion either way -- they are complicated businesses and beyond my capacity to fathom.

    But I'm comfortable with the idea that HA's interisland monopoly -- not so cyclical, in fact probably anti-cyclical -- would be able to sustain it.

    So I would be horrified if HA carried no or little debt -- it would be quite wasteful and wouldn't speak well of management's concern for value creation.

    If I were looking at HA with fresh eyes I'd try to model a worst case EBITDA minus maintenance capex number and see how many times it covers interest expense. I'd judge a number below 2x in terest as bad and a number above above 4x as good enough.

  5. Hey red,

    Thank you for the valuable comments.

    I'm trying now to get a sense of HA's normalized maintenance capex.

    So far I've taken HA's net sales, gross PP&E, net PP&E, D&A charges and cash flow capex - and tried to get a sensical maintenance capex out of it.

    HA's average PP&E/revenue ratio is about 0.5 for 2006-2014.
    Its real revenue growth (2014 figure less 2006 figure) is about $1270m for this period.
    That means HA's *growth* capex is about $635m for 2006-2014 (0.5*$1270m).
    HA's sum of capex between those years is about $1800m.
    So, HA's *maintenance* capex is $1165m for 2006-2014 ($1800m-$635m).
    Comparing it to the sum of HA's D&A charges between 2006-2014 of about $500m, gives a "maintenance capex to D&A charge" ratio of 2.33 (233%).

    So that means HA's 2014 maintenance capex is $225m (2.33*$96.4m), which is quite a big number relative to HA's operating income, and is of course different than your numbers.

    I'd much appreciate if you could tell how you came up with your maintenance capex figure.

    Kind regards,


  6. Well MCX is basically replacing aircraft as they age so I'd count how many aircraft they have, how much they cost each, how often they'd have to be replaced, and what their resale value is.

    You'll notice that all of the preexisting fleet has been replaced in the last 3 years and several additional aircraft have been added.

    Aircraft are generally depreciated over 20 years and their resale value after that is 10% to 15% of sticker price when they were new. Roughly speaking; it depends on the popularity of the aircraft and what shape it is in and so on.

    So: How much did they spend per aircraft, How long are they to be amortized for, What is their residual value -- that should give you a fairly accurate MCX estimate and it's not going to be materially different from the depreciation figure in the financial statements. So I would (and do) use that.

    (Plus, since they've recently refreshed their entire fleet they're not likely to actually spend cash on aircraft MCX for many years. They do have plans for growth capex though)

  7. red,

    That's great. Thank you.


  8. Red,

    You obtained 244 FCF by subtracting 50 of interest. But I´m used to calculate FCF withouth subtracting interest. Am I doing it in a wrong way?

    In your four tabla maybe there is a mistake in 2014 EBIT:
    Gross profit(816)less overhead (682) is 134, and you have 245.


  9. Ah right -- I've fixed the errors.

    FCF: We've had this conversation before. There's FCF to the enterprise, which is how you're doing it, and there's FCF to equity, which is how I've done it here.

    If you want to know how much FCF will flow to the equity then it makes sense to subtract the interest payments.

  10. Any concerns about the strong dollar reducing international travel?

  11. Hey red,

    Regarding my attempt at getting to a sensible maintenance capex for HA, I've taken 31.12.2013 number of operating aircraft, which is the relevant number of planes for 2014 results, and matched those planes with their list prices which I found on the web.

    I've decided to go on list prices and not to try and compute some 2nd market prices for used planes, which of course would be cheaper, but than again also the residual value would be lower and likewise the depreciation period which be much shorter.

    So these are the results -

    14 * A330-200 * $221.7m
    12 * 767-300 * $182.8m
    18 * 717-200 * $40m

    and in total = $6,017.4m
    Residual value (10%) = $601.7m
    Amount for depreciation along 20 yrs = $5,415.7m
    Annual depreciation charge = $270.8m

    So, I've got annual maintenance capex = $270.8m.

    Now, we do need to keep in mind that the reported D&A, $96.4m for 2014, isn't conclusive for our needs, as it doesn't take into account the operating-lease-underlying depreciation charge which is about $27m according to the company.

    So all in all, we have a reported adjusted D&A annual charge of $123.4, and maintenance capex (calculated by me) of $270.8m, which means of course that if while we compute HA's normalized earnings we replace the former with the latter, we would get quite a low earnings figure.

    I guess this happens mainly because capital and operating leases' computed depreciation charge derives from a relatively low present value figure for the leased aircraft, as the accounting rules make the company calculate the present value of only the minimum amount of lease payments under contract. Which means we get an artificially low adjusted D&A charge because the accounting rules treat leased aircraft not as a indispensable,necessary and permanent PP&E, which they effectively are, but as a temporary asset, depending on the length of the leasing contract.

    red, what do you think? Have I got it completely wrong..?

    Kind regards,


  12. Well, let's keep it simple. What would owner earnings be if HA leased all of its planes instead of owning some them?

  13. Also Michael, I'd use this to price the aircraft

  14. Ok, thanks. I'll look to it.

    Regarding your question, if you mean what would happen if HA had been leasing all of its planes by operating lease, I would say that the annual rent would have been about $782m (which implies a 13% yield for the lessors).

    This is much more expensive of course than buying planes with regular debt, considering the fact that regular debt costs about 4-8% I guess.

    Anyway, this $782m would have replaced HA's reported D&A charge + (let's say) all of the "Maintenance materials and repairs" costs + Aircraft rent + interest expense, which all together equals to $493m for 2014. So we would have an additional charge of $289m.

    This of course would have made PBT negative.

  15. I don't think we can derive the price of a A330-200 plane from this press release.

    The press release says that HA borrowed some $444.5m in order to buy 6 A330-200 planes, which presumably means, if I understand you correctly, that the cost of one new A330-200 for HA is $74.1m.

    1. This alleged price of $74.1m is way too low than the published list price of $221.7m, and keeping in mind that Airbus and Boeing are a duopoly, I don't see a reason for this too wide a gap.

    2. We don't know that HA didn't paid some of the price from its own funds.

    3. We don't know that HA didn't paid some of the price from another debt instrument.

    All in all, as I said, I think its much more conservative to take the list price (and maybe, just maybe give it a little haircut, but definitely not much) than trying to compute our own prices from half reports.

    What do you think?

  16. $74 million plus the %15m each that was already prepaid.

    Go here

    2nd presentation down, page 34
    $90m per A330


    Go to the latest 10-K and look at the Gross PP&E line item. That's what they've paid. Divide it by # of aircraft.

    I think your method was fine but the aircraft cost estimate threw everything else off.

  17. red,

    It seems you're right.
    I took your advice and looked at the PP&E gross value against the number of owned + capital leased planes, and it turns out the company bought them for about 0.48 of the list price.
    I guess it happens this way when Boeing & Airbus get order years in advance.

    Anyway, computing maintenance capex by this scale of price, turns the $6017.4m I have computed before to $2888.4m after scaling it, and to $2599.8m after taking into account 10% of residual value. Depreciating this sum for 20 years gives us an annual depreciation charge of $130m, or maintenance capex of that sum.

    And as you said, it matches quite beautifully to the reported adjusted D&A annual charge of $123.4m.

    Very nice.

    red, I wonder whether you remember any press release or investor Q&A that made an explicit and clear remark about the price paid for any of the planes. I know you've showed me the 6 planes debt's press release and a chart for annual aircraft capex, but I'd really like to learn about the 0.48 list price cost from an unambiguous, no-doubt straightforward source.

    Anyway, thank you for taking the time brushing off my doubts.

    Kind regards,


  18. I don't think there's anything like that, actually. I know that there was a release in 2009 or 2010 announcing the agreement with Airbus for 24 aircraft but it was for a mix of A330-200s, A350-XWBs and included purchase rights so not as clean as you would like. I think the indicated value of the package was in the neighborhood of $4400m.

    But you can always look at the cash flow statements in 2009-2010. That's when A330s started arriving and the evolution of the capex figures in the time period should tell you what you want to know.

  19. Hey.

    thanks for a great write up! I have question that it would be great to get you input on.

    Won't lowered fuel costs for the industry lead to lower prices for the consumer and RASM in a highly competitive industry like this?

    A back of the envelope would indicate that if the airlines would like to keep the same level of profitability RASM would decline c. 8.5% (assuming 35% fuel and 58% fixed costs, pre fuel price decrease).

    One might argue that that doesn't apply to the Interisland routes but I would not feel comfortable assuming the same for the NA and International routes. What's your view on this?

  20. ...and assuming fuel costs decrease with 24%

  21. Good question,thanks.

    I've uploaded some numbers here:

    In 2009 -- weak demand, more players, fuel at $1.77/gallon on average, the spread between RASM and Fuel cost per available seat mile was as you see there.

    Using that as a benchmark suggests that the spread should be 7.40 or so at $2.30/gallon.

    This assumes, of course, that capacity discipline is not worse than in 2009 -- and I recognize that that's at the heart of your question. I don't know the answer to that but one way to protect against it would be to buy puts on the most likely miscreants and losers in that scenario -- AAL, Delta etc.

  22. What do you think about consolidating airline industry? When the weak players are shaken out, and a few large players remain, it could turn into a more rational oligopoly? Which would be good for HA, as that means less pressure on their west coast fares?

  23. Airline industry in US has already consolidated quite enough and there are few (if any) weak players plying their trade between the West Coast and Hawaii because the competition along those routes is just too fierce -- they can't afford it.

    I think there are a number of good reasons why that segment will remain perennially under pressure. For one thing, it's attractive to have a frequent flyer program that offers Hawaiian as a possible destination. So, in that way, Hawaiian routes can act as a loss leader for airlines such as LUV and DAL.

    Also, I think HA itself would be quite willing to only break even along those routes as long as it feed traffic into its high margin interisland network. In the last year, therefore, HA itself has been the main driver of capacity additions along those routes.

  24. Hi Red - Appreciate the write-up. Was wondering why you think gross margins for the international business will be closer to inter-island vs NA? Unlike inter-island (where HA has dominates market share), what would prevent competitors driving down international margins by adding capacity? Is it that there's less of a need for APAC airlines to hold routes to Hawaii to make it more attractive for passengers? Thanks!

  25. Hi,

    I've uploaded a picture of what I'm suggesting

    Segment profitability is more apparent once overhead burden is allocated by capacity.

    If you compare this with the last graphic in the main body of my blog post, the profitability that I'm imagining for the international segment is not out of line with industry benchmarks.

    Bear in mind that HA is a full service airline (with associated pricing) with an LCC's cost profile and it could be argued that I've understated that segment's potential a little.

    On the other hand, as long as the dollar/yen and dollar/AUD are where they are, pricing power will remain disguised.

    All in all, I think a 3.5 cent spread is a fair target though, of course, this is ultimately a matter of judgment and supposition.

  26. Hi Red,

    Great blog. I have truly enjoyed reading it.

    I was curious what is your reasoning for RASM to stay at 2014 level when company's guidance for Q1/2015 is significantly lower (13.56 vs 12.34)? Is there something I am missing?


  27. Hi

    Good question,thx. Q1 revenue shortfall is a one-off so the full year impact is 25m meaning that RASM would come in at 13.41 for 2015, other things being equal.

    My bias is to believe that HA will charge 5% more on its interisland flights to make that up and keep system-wide RASM steady. A $60 ticket becomes a $63 ticket.

    Of course, if you find that implausible, the thing to do is to use 13.41 as full year RASM.

    I don't think it makes too much of a difference.

  28. Hey Red,

    Two things.

    First, on the Q4 earnings call, management stated that maintenance capex was just $60MM in 2014 and will be lower going forward. Can you make any sense of this?

    Second, how did you pull together the segmented financials? Did you have to use a manual process using the data on the RITA site? Did you literally go destination by destination and total the results?


  29. Hi Punch,

    They've just finished overhauling their fleet so there'll be no need to replace aging planes for another decade +. That's the sense in which they're using the term maintenance capex -- cash outlflows. The ordinary meaning -- a reserve against future capex would have MCX equal depreciation.

    The cash MCX guidance they've given reflects engine parts and such rather than aircraft.

    2) I did it both ways and then found and used this:

    And the segment revenue figures are in the annual reports.

  30. Thanks, Red. Did you assume that "Pacific" = Interisland?

  31. Ah no. I should have mentioned instead that "narrowbodies" = interisland

  32. Sorry about this. But how did you calculate RASM for narrowbodies (inter-island)? I can't find revenue broken out by aircraft type.

  33. See my reply before last: the segment revenues are in the ARs (and sometimes but not always in the 10-K filings too).

    e.g. in the latest 10-k

    "During fiscal year 2014, approximately 24% of our passenger revenue was generated from our Neighbor Island routes"

  34. I saw that and zipped right by it. My apologies. Thank you for your patience.

  35. Hi Red,

    How are you getting to your 2014 CASM/DCASM/RASM numbers, given that full year 2014 data hasn't been released yet?


  36. Extrapolating from the 2014 SEC filings and from what we know about the cost narrowbody/widebody cost structure. Won't be perfect but it'll get you close.

  37. Thanks - in terms of wide body/narrow body cost structure, I assume you mean what we've seen historically with HA's margins in interisland vs everything else?

  38. He red based on latest release I get
    4441648 * (12.96) = 576,063,980 revenue for Q2
    4441648 * 8.21 = 364,659,300 in costs
    and 2.25 * 58m = 129m$ in fuel costs

    So that means about 83m of operating income.

    With 15m interest, and a 35% tax rate that is about 44m$ of net income. Not much off the analyst estimates of 35-39m. It seems Ebit will be more like 320m$ this year?

    Or is it Q3 and Q4 that will suprise on the upside?

  39. summer vacation period falls in Q3 so it looks like this at the low end:

  40. How do they get to 61c of adjusted EPS? Are they assuming average fuel hedge costs? I was suprised to see this is actually a earnings miss. It seems HA should trade at least to 2bn$ with these estimates? Does the market think they are more vulnerable to a strong dollar?

  41. red, I am interested in your opinion on HA now. Results were better than expected. I am still long and would sell at $50. Looks like next year will have increased results due to lower fuel cost. I just fear this is not sustainable over the next 5-10 yrs.

  42. Hi Martin,

    At $2.50/gallon jet fuel HA earnings $2.80/share growing at 5% per year. So this is an opportune time to think seriously about exiting while there is some incentive for someone to take the shares off your hands.

  43. Thank you Red. Long term I am also not that bullish. As a hedge for oil exposure in the portfolio HA is also not that great because oil price cannot fall that much again. For the short term their cost should be low though, which makes me believe there could be a higher price (based on TTM EPS) to unload shares. I tend to sell too early most of the times. But that would mainly be the greater fools' short term orientation to hold on.
    "Based on the fuel curve as of January 25, our economic fuel cost per gallon for the first quarter is expected to be in the range of $1.50 to $1.60 and for the full year $1.35 to $1.45."

    If they would grow earnings 5% into perpetuity HA would be relatively cheap.


  45. Worth watching HA's share price over the next few months.


    Maybe you have more followers...


  48. Basic model for anyone interested in puzzling through it. LUV is hoping to eat into the first column.



Note: Only a member of this blog may post a comment.