Wednesday, November 4, 2015

Flybe plc - Regional Airline


Flybe is a UK-based regional airline. Its core business is flying short, thin, domestic routes – 280 miles on average; 40,000 to 60,000 passengers per year – that can support only one or two departures per day, if that. 

These routes are expensive to operate and expensively priced. But they compete against coaches, trains, ferries, and self-driving, and therefore appeal to time-sensitive travelers – the businessperson who needs to get from Aberdeen to Belfast and back again before supper; the Durham University student who needs to get to a grandparent’s funeral in Cardiff or to a sister’s wedding in the New Forest; the working couple from Salford off for a B&B weekend in Devon. (In the summer season, of course, there’s plenty of trade and everyone’s load factors go up). 

Profitability in a business like this requires a number of basic elements:

  • Stick to your knitting: short, thin routes such that viable competition is unlikely;
  • Use aircraft appropriate to these short, thin routes;
  • Configure the hubs and flight schedules so that each aircraft (and pilot) flies as many hours a day as possible;
  • Use young, inexpensive crews;
  • Make do without layers of middle management;
  • Use an unbundled, ancillary fare structure to drive down headline ticket costs and thereby increase load factors and total revenue per flight; and
  • Outsource as many non-core functions as possible
  • Maintain a healthy balance sheet

To cut a long story short, Flybe fell afoul of each of these principles. Half of its fleet consisted of regional jets that it couldn’t fill, that burned fuel at an alarming rate, that cost a great deal to own, and that tempted it into flying the longer, fatter routes flown by Easyjet and Ryanair. It employed 800 or so staff more than necessary, which is to say that two thirds of its non-crew personnel were excess. Its cost structure was so bloated that it felt the need to push its fares up even as its competitors’ fares, already much lower, were steadily falling. It outsourced nothing to speak of. 

Its own house in disorder it entered into a joint venture with Finnair to run a Scandanavian regional airline operation that turned out – no surprise – to be both unprofitable and a cash drain. Its liabilities – debts, lease and purchase obligations, and so on -- amounted to £800 million. By 2013 Flybe was functionally insolvent.

The Turnaround

Flybe’s then management attempted a first round of restructuring. At the end – not the beginning – of that exercise 89 of Flybe’s 158 routes, or 56%, did not cover their direct operating costs, i.e. the cost of the aircraft, crew, fuel and navigation. We have subsequently discovered that it cost less to ground the majority of its E195 regional jets at an annual cost of £26 million a year than to fly them. 

Those are grim facts that contain within them the insight that the remaining 44% of routes, corresponding to the 78-seat turboprop aircraft flying lightly contested airport-pairs – the short thin routes I spoke of earlier – were profitable enough and steady enough to year in, year out bear most of the weight of both the distended fixed costs and the gross margin losses attributable to the unprofitable jets. 

In any case, previous management had to go and they did. Equity needed to be raised and it was. Saad Hammad, a former Chief Commercial Office of easyJet and subsequently Managing Director of a private equity firm, was hired as chief executive.  He was promised a base salary in the £400,000 to £500,000 range and 3% to 4% of any improvement in Flybe’s market capitalization. There is no upper limit on that compensation package. 

He, in turn, hired a new management team – some from easyJet, some not – and they have since set about stripping away the unprofitable routes, restructuring the fixed cost base, optimizing the route network, and investing in yields in order to reveal the underlying profitability and stability of the core business. 

So far, so good . . .

. . . although it may not be immediately obvious from a quick glance at the company’s filings or presentations. The reconciliations to GAAP earnings look like this:

The implications for the financial statements are as follows:

I'll pause here to note that, given Flybe's principal competition on its core routes is intermodal transportation, 2015 was a particularly challenging year. Its fuel hedging program is such that it is still pumping $1,000/MT fuel into its aircraft whereas coaches, ferries and private cars are using fuel that costs half as much. Flybe's cost disadvantage relative to these modes of transportation is/was therefore at its widest. The fuel hedges roll off in FY 2016/7.

The new management team have also substantially eliminated the £800 million in liabilities mentioned earlier. £40 million of the remainder are attributable to the rent and maintenance costs of the excess E195 jets: £20 million this fiscal year, £10 million next year, £6 million in 2016/7, £4 million after that, and nothing thereafter. 

Beyond that, Flybe will take over the lease on 48 78-seat Q400 aircraft  from Republic Airlines. these airlines are ideal for Flybe's core routes and uneconomical for the distances that Republic typically flies. In any case, Hammad and his team have assessed that there are suitable airport-pairs (short, thin routes within the UK and also between such cross-channel city pairs as Birmingham-Brest) sufficient to absorb the 1.6 or so million seats that these extra aircraft imply.


The viability of Flybe's business model depends on the viability of the UK's regional airports. These airports exist only because of the financial support of individual local authorities and the central government. If that support withers the investment thesis fails.

The turnaround has not been all sweetness and light. The decision to enter the fray at London City Airport is, in my view, ill-judged. The hoped for potential of white label flying for European flag carriers has not and will probably never come to fruition. Individually these are immaterial matters. Together, however, they betray some potential signs of the weakness that has traditionally afflicted the Airlines' executive teams: the desire to strut on a grander stage.


Ultimately, I suspect that Flybe will be taken private by a private equity firm (or by a business like Torghatten that also owns Broderoe). The nature of its business means that, once the map of short, thin routes has been infilled there is no further profitable growth to be had and therefore no reason to be publicly listed. 

In any case, net cash plus 12% FCF yield is, in my judgment, an attractive purchase price for the whole business once fully fixed. 

Why does this opprtunity exist?

Two or three reasons, I think, although one can never be sure what other people are thinking. 

First, the analysts at the larger sell-side shops are  capitalizing the rental & maintenance expense of the excess E195 aircraft. So, when they last updated their views after the publication of the 2015 results, they calculated a liability of 7 x £26 = £182 million (or 8 x £26 = £208) instead of simply capping the liability of £80 million. As mentioned above the liability is, as of this week, capped at £40 million.

Second, Flybe's costs translated into the airline industry's standard measure of unit cost, cost per available seat kilometer, or "CASK", can cause observers -- even those who ought to know better -- to compare Flybe unfavourably to, say, easyJet. Something like this: Flybe's ex-fuel CASK is ~16p, easyJet's is ~3.6p and therefore Flybe is not "fit to compete".  Unmentioned is the fact that CASKs (and RASKs for that matter) rise exponentially, ski-slope like, at stage lengths less than 500 miles. Flybe's average stage length is 280 miles and there is no way yet devised to fly these stage lengths using European crews and paying European airport and navigation fees at CASKs below ~13 or 14p. 

Third, it's both an airline and a turnaround.There a slew of adjustments in the 2015 results. Etc.

Disclosure: I own shares in Flybe

Addendum November 9th, 2015


  1. Hi red,

    Thanks for the post. Regarding the fuel costs, the current hedges roll off in 2016/ is the management going to lock in the current lower prices for the anticipated fuel use beyond 2017? What if oil spikes in 2016/2017?


    1. That's right,whatever the fuel price is then is locked in for two years or so. If it reverts back to $100/barrel in the interim then Fuel/seat costs remain in the 10/seat range as in the exhibit above and the profit/seat = ~7, implying the ability to earn 70 to 75 million of so per year.

  2. Hi Red, thanks for the write up. Shouldn't you capitalise the leases and add pension deficit though to net cash? i.e. Flybe actually has net debt in 400-500m area?

    1. Thanks for the question. You could do it that way if you want to value it on an EV/EBITDAR basis. I have added a table to the body of the post above. Note that the excess, parked E195 aircraft ought not be capitalized -- only the operating aircraft ought to be capitalized in the usual manner. Target multiple = 4.5x EBITDAR and gets us to roughly the same price objective as with the alternative, FCF-based approach I used .

  3. Hi Red,

    For 2017: Do you estimate the revenues with the aircraft increases for 2016-17, and applying more or less the same load factor of the last quarter? And how do you do for the net (cash)/debt?

    Thanks for your writtes up.

  4. Hi David

    Regarding revenue per seat, here's what the last few years (including the recession have looked like):

    2008: seats = 10.7 million; Passenger revenue per seat = 47.12
    2009: 11.1; 48.22
    2010: 11.3; 46.06
    2011: 11.6; 46.98
    2012: 11.6; 48.72
    2013: 11.3; 48.82
    2014: 11.1: 49.73
    2015: 10.3; 51.35

    2010-1 was affected by the icelandic storm.

    In the years 2009-2013 the 118-seat aircraft constituted a significant share of the flying fleet. They were too large for the routes operated (and expensive to own and run). Load factors in that period = ~65%.
    Since 2014 these have been increasingly taken out of service and parked. This has allowed Flybe to reduce fares. The combined effect of smaller planes on average and lower fares has pushed load factors up so that revenue per seat has gone up by 2.50.

    The aircraft being delivered over the next few years are 78 seat Q400s so this trend of increased load factors and higher revenue/seat should continue. Based on their fleet plan (the last slide of their 2015 full year results presentation) I am estimating that Flybe's seat capacity goes back up to 11.9 million by 2017.

    Building it up from the information that Flybe regularly presents:

    "Seats" = # of flights x average # of seats per flight
    "Flights" = block hours per aircraft per day x flights per block hour x # of aircraft x 365

    In 2017 I expect
    Seats per aircraft = 80
    Block hours per aircraft per day = 7.7
    Flights per block hour = 0.88
    average # of aircraft during the year = 65

    In English: 65 80-seat aircraft flyng 2,745 flights a year each translates into seat capacity of 11.9 million
    If the average flight does at least as well as the average performance in 2015 then 2017 revenue ==> ~620 million (i.e 11.9*52.15)

    FCF: I think I show how I get to net cash in the 3rd table of the post. Let me know if it's still not clear

    1. Hello Red. Thanks for such a useful write up.

      This is not very material, but is .88 above the 2013 value by any chance? If so, then it appears to be "block hours per flight" instead of "flights per block hour".

      Number of flights = 11,298,200 seats / 85.60 average seats per aircraft = 131,988.32
      Flights per aircraft per day = 131,988.32 flights / 59.90 aircraft / 365 days = 6.04
      Flights per block hour = 6.04 flights per aircraft per day / 5.30 utilization, block hours per aircraft per day = 1.14
      Block hours per flight = 5.30 utilization, block hours per aircraft per day / 6.04 flights per aircraft per day = .88

      I think "flights per block hour" is correct in your "Flights" formula and not "block hours per flight".

  5. Red,

    Net cash in 2017 (228) shouldn´t be equal to net cash in 2015 (77) plus FCF 2017 (101) ?

    Thanks for your time.

    1. and some FCF from 2016 also .. I think maybe 30 to 50 in 2016 FCF

  6. And there you go. From today's earnings related presentation / conference call:

    Not content with doing things the easy way -- get rid of bad routes, stick to knitting, find zero cost option for excess jets -- and earning 80 to 100 million a year, the management team has decided that they want to become a pan-European regional airline. Edinburgh-Paris-Dusseldorf etc. If they have any compatitive advantage on these kinds of routes it is likely to be slight and fleeting.

    So, it seems to me that an element of unnecessary risk has been introduced into the situation and it is important for one to size one's position accordingly. They have modeled these routes and they look good but they also modeled Londin City to Dublin before they started flying and were forced to beat a hasty retreat when lower cost incumbents started a fare war.

    1. The question is what are they going to do with those 48 new aircrafts. Aren't UK lines saturated? In france there are a few opportunities involving cities like Bordeaux/Lyon/Marseille/Strasbourg but they would face competition from Air France (no worry) and Easyjet (more annoying). In germany, low-costs like Eurowings and ICE offer a good coverage alreadty. In Spain, there might be some opportunities. Vueling is not that cheap and the fast train does not go everywhere and is expensive.I don't really see where they can grow.

    2. Well I'll frame my answer in this way. There are concentric circles of profitability. In the center are the routes that are highly profitable and Flybe's principal challenge is to lose as little as possible in flying the the routes on the outer edges. Flying 118 seat jets on these routes is disastrous; flying 78-seaters is less so.

      So, from an investing perspective you'd say:

      - they cut out 36 from the cost base before the excess aircraft
      - they will have cut >26 from the disastrous aircraft-route combination
      - that's net +60 to profit before any fuel benefits
      - it doesn't matter so much whether fuel stays low or goes back up, it's the temporal mismatch between old/new fuel prices that matters
      - so plus ~35 from fuel ==> +95 to profit

      Flybe lost ~30 million in 2013 so baseline profit after absorbing 48 new aircraft (so that seat capacity is back at 2013 levels) = 95-30 = 60.

      That's simply from assuming that flying 118-seat jets on bad routes is worse than flying 78-seat turboprops.

      As for the remainder 40 or so here are the options and safety outlets:
      a) More intra-UK routes now within reach b/c of lower guage aircraft and better scheduling technology/skills (so # of routes goes from ~154 to, say, 200)

      b) if there are still excess aircraft (esp outside the summer season) there are low-destiny route possibilities between UK regions and smaller continental towns: Durham, Exeter, Preston, Chichester, Inverness etc to Barcelonnette, Porto Santo, Bullingen, Brest, Dijon, etc and vice versa

      c) if that's not enough to keep them busy all year round then use them for charters (charters with 118-seat jets is a much tougher sell)

      The only thing is to not compete on longer, well-traveled routes. What it looks like is happening is that they're trying popular airports and routes (e.g. London City, Bournemouth) losing money on them and hightailing it out of there. It's undignified and baffling but better than the alternative, Jim French way of doing things.

  7. Hi Red,

    Would be interested to hear your thoughts on the Q3 trading update?

    Decline in passenger revenue per seat is quite in line with the update they gave at the Interim result. If I recall correctly Saad was questioned on it and answered along the lines of immature routes and business passengers booking later. The "immature routes" part would usually be a feasible explanation but no longer looks so great given they were able to increase capacity with minimal dilution to revenue per seat in H1...


  8. Hi,

    This half and next are the points of maximum pain for Flybe. With spot jet fuel where it is compared to where Flybe's hedged costs are, the competitive/substitution risk is tremendous.

    Tt's off-peak so they've chosen to focus on the segment that is price insensitive -- i.e business passengers for whom time and same day RTs are more important than whether a ticket is + ot - a fiver. It was explained awkwardly by them as "defending yield". So, I get it.

    A lot of their international routes are France and Benelux, too, so I understand how the load factor could have come under pressure in 3Q.

    It was a profitable Q, though, and also a YoY improvement, despite the LF weakness. Sometimes the result gets lost as the commentary takes the foreground so I'll emphasize that here.

    This idea really depends on just two things: (1) their operating decisions are not worse than those made by the prior mob in 2012/3; and (2) the relationship gap their fuel hedges and spot price normalizes or reverse over the next 18 months. If these two points hold up then Flybe earns 40p/share and and has >100p in excess cash per share by 4Q 2016/7.

    For a while, this thing is going to trade as the HSBC analyst wills it to. And then the EPS / FCF / excess cash numbers will be unbearable enough that it won't trade at these multiples.

    1. Thanks.

      On the Q being profitable - my calculations are: 3Q15 cost per seat was 53.48, so working backwards to 3Q16 you get to 50.96 cost per seat (4.7% decrease v pcp.). And then adding on contract flying and MRO revenue you get to a ~3m operating profit. Is this how you've approached it? I see what you mean, the update was poor optically and did take a small amount of digging to work out the underlying profitability.

      Also looks like fuel costs are set to decline 25-30% in FY 17 based on what they've hedged so far, probably more given they've only partially hedged.

      Overall, I think it does take some imagination to see how you lose a lot at the price where FLYB is currently trading at, and it seems you agree too...

      Many thanks again.

    2. That's right. Also, there are enough data points in the H1 2016 presentation to allow one to see this:

      New routes are bleeding cash as new routes tend to do. Since H1 is the money maker the short term investment case turns on the question of (a) whether new routes will continue to perform worse than the discontinued routes, and (b) if so, whether management will continue to fly those routes rather than, for example, simply parking the planes.

      In the medium term, the E195 will be retired, the company is FCF positive eith way, and it has plenty of liquidity in any case. So in the medium term I think it's not easy to see how it doesn't work out. In the short term (i.e. by H1 2017 report) whether it works out or not is a function of management competence in route selection.

    3. Thanks for the heads up - I could only get to implied RPS for the new routes in H1 FY16 from slide 21 and that comes to 37.8 GBP (-31% vs 54.75 RPS in H1 FY15).

      By the way, your per seat loss is inconsistent with total combined loss for the new routes, I think its your new routes revenue line that's causing the problem. Loss from new routes should be far more modest.

    4. Also keen to hear how you get legacy RPS for 1H 2015.

      Thanks and sorry for the harassment!

  9. Good morning. Not at all -- I appreciate the discussion.

    I think we're more or less on the same page. I have uploaded my derivations & corrected sheets here
    and here

    1. Thanks for this, just FYI the H2 2016 numbers still don't add up but anyway the point is made and I can see it for myself, and yes we very much are on the same page. The yield and load factor assumptions are your own?

      The upside seems crazy but the assumptions made to get there are all very reasonable in my view. Would you know what sell-side numbers are at?

  10. Consensus EBITDA forecasts
    2015/6 = 23
    2016/7 = 48
    2017/8 = 63

    Haven't changed all that much since May of last year. Price targets have been quite volatile, though.

  11. Hi,

    Any evidence which suggests that fares for coaches/trains/ferries have fallen along with oil?

    I'm not based in the UK and internet searches have not yielded much so any help would be appreciated.

    Many thanks.

  12. Train pricing is up 1%, coaches are suffering, and I don't know about ferries. The pressure is coming from / has come from self-driving (and from lower pricing from competitor airlines that are not as hedged as Flybe has been)

    There are a number of publicly traded bus/rail companies that have provided color on fuel impact in 1H 2016, including
    Go-Ahead Group
    National Express

  13. This could just be me, but I'm getting ~37 pounds for operating cost per seat (ex fuel and aircraft rental) in 1H16, which is a decent level of inflation compared to 1H15 if you strip out all the one-off investments and one-offs.

    However, from what's been provided in trading updates it seems like it comes back down to 33 pounds per seat in 2H.

    Are you getting similar numbers? I suspect it is the cost associated with launching new routes etc.

  14. Sounds about right. Take out 77p/seat for the reclassification of the cost of sales of on-board items and you have 36.29/seat in 1H 2016 v 35.42/seat in 1H 2015. Fresh routes will do that. In the future, period-to-period, cost per seat will bounce around because of GBP weakness/strength (maintenance costs & rents are in USD), timing of maintenance, etc. But the trend over multiple periods should be down.

  15. Hi Red,

    The excess E-195 don´t have to be capitalized, we just have to take 40 millions of liabilities. So I imagine, that the way to calculate the capitalized rent for operating aircrafts is substracting the rental charges of the E-195 of the total rental charges in 2015 (ANNUAL RENTAL CHARGES 2015 37x2), and then multiplying the result for 7.

    But the problem for me is how I could know the rental charge of the E-195. Could you please explain me how you get the capitalized rent of 382 for the year 2015?


    1. Hi David,

      If, for example, you look here
      and search for "surplus capacity costs" you should see that 6 month cost of the E195s was 14.5M so it annualized to 29M.

      However, the company has now found uses for the E195s. They are no longer parked and they should be flying at or near breakeven profitability soon. In any case, to the extent that some of them are still loss-making they will be returned to the lessor according to timetable laid out in the annual report.

  16. Thanks a lot Red.

    And be carefull on Saturday with Benzema :)

  17. Hi Red,

    You sold Texhong shares expecting a drop in the share price caused by the cotton low prices in Q1 2016?


    1. Ah no, just trying to reduce the total amount I have invested in the market in the most taxefficient and balanced way possible. here will be a few transactions like that in Q2.

  18. Thoughts on the trading update?

    Quarter was pretty in-line but it seems like market didn't like it. I don't really see why though.

    Based on the cash balance they've put out, it looks like they did ~£15m of FCF if you don't include the aircraft purchases. This is assuming they didn't borrow any more cash and I don't see why they would.

    1. They finally properly hedged fuel and dollar so we have some visibility on unit costs as we head into the next 18 months. That's good. The rest of it was in line with what one would expect and on trend with Q3. They have implied in the past that time to (income statement) breakeven for new routes = 18 months. If that turns out to be so and the rate of capacity growth slows as it should, FY 2017 will see cash flowing. So the Interims in November are significant.

      Airline, UK market, small cap, consumer-facing, turnaround: I don't think we can count on the market to anticipate. So the Interims in November are significant but I expect a random walk in the meantime.

      I've linked to this in my tracking portfolio page:

      I doubt the investment case is a great deal more complicated than that.

    2. Agreed, thanks!

    3. Hi Red,

      Just revisiting this. Do you think the upcoming full year results could be meaningful?

      The stock's sold off based largely on the last 2 trading updates that on the surface didn't leave much to cheer about. However, working through the numbers I have the second half as just under break-even. I'm not quite sure what the market is expecting here but with earnings in the region of 15-20m + re-emphasising reduced fuel costs and lower surplus capacity costs going into FY17 I can imagine this could move the stock up a bit. Regardless, I don't see how Flybe can disappoint in a big way either.

      Am I missing anything?

    4. This is what brokerage expectations look like as of january

      Flybe earns its profits in the first half of its fiscal year -- i.e. the period covering the summer season and reported in November.
      If the fundamentals are what I think they are then earnings/FCF will equal market cap -- and net cash will exceed market cap -- soon enough that the return on one's investment will be quite attractive.

      Factos that could affacte sentiment and share price in the interim:
      Jet fuel price
      Forward bookings
      Hot money inflows/outflows
      Brexit vote
      Easyjet's results and/or outlook
      etc etc

      For some perspective, the market felt better about Flybe when it was insolvent and in a liquidity crisis than it does now. That's not what one would have predicted as being a reasonably likely scenario. So I don't know that it's worth the bother of trying to anticipate the market's reactions to Flybe's results, outlook or environment in the immediate term -- I don't think it can be done and i don't think it matters all that much.

    5. hi red,

      couple of questions...

      1) would it be correct to assume that the 'tracking portfolio' roughly reflects your actual portfolio?
      2) how do you think about sizing your positions?
      3) why is flybe not a bigger position? :) is it to do with tail risks like terrorism, air plane crashes etc?

    6. Hi

      1) Yes

      2) Probability of involuntary loss over a 2 to 3 year time frame. (This is a judgment call based on the interplay between business model, my understanding of what's going on, indebtedness, management quality, macro exposure etc).

      3) I think Future Bright, Paradise, Rentech, & LSB Industries are as undervalued as Flybe and since they are not all exposed to the same stimuli I hope/expect that they will pay off at different times. If so, I will recycle gains from winners into laggards when the winners approach the outer edges of a reasonable valuation. So if, say, LXB goes to $25 this year and Flybe stays put I'll be happy enough. That's more or less why it isn't larger.

      Plus, you know, just one decision from the UK government to eliminate subsidies to regional airports would place the value (let alone the price) of the shares under serious pressure. Brexit, too.

  19. In your opinion, what can Flybe do to increase their profitability and turn their business around better?

    1. My view is that they've more or less already done what they need to do and that those actions will bear fruit over the next 18 months so so.

  20. red, you seem to assume the losses on unprofitable international routes will end soon. Why do you think FlyBe will then not try some new other routes? By the way sold all my HA recently.

    1. Hi Martin,
      Remember, the unprofitable international routes exist as a way of using the excess E195 aircraft that have been the source of Flybe's problems all along.
      Stage 1: they were losing so much money flying routes unsuited to them that they led Flyybe into a liquidity crisis;
      Stage 2: New management, new strategy -- park them and eat a smaller loss of 26MM per year
      Stage 3: "Okay, we've looked at more suitable routes where we thing we can breakeven at least"

      Now the leases on the E195s expire -- in a staged manner -- over the next three years. (You can find the expiration/handback timetable in one oftheir presentations). If stage 3 doesn't work then the aircraft are returned to the lessors and the problem goes away.

    2. hello red, as always - thanks for posting.

      i am not sure i understand stage 3 "breakeven" comment in context of the following.
      "Completion of Project Blackbird delivers a £40m mitigation versus Flybe's previously indicated obligation of £80m over the remaining term of the aircraft leases. We currently anticipate that the financial cost of resolving the remaining E195s will be £20m this year, reducing to around £10m next year, and then to £6m the year after and £4m in the following year."

      don't we expect a 40m cost over the next 4 years? when you say breakeven - are you implying they could do better?

    3. I should clarify. There were 9 excess E195. Average annual rent 2.88MM and average remaining term of 3 years. Total liability of 80MM.

      5 have been placed under multiyear PSOs and it looks like the subsidy is equal to ~92.5% of rent on those aircraft, so 40MM mitigation (3 x 2.88 x 92.5% x 5).

      The problem of the remaining 4 aircraft has not been resolved at all. That's, I think, the 40MM remaining liability. So the question is still whether to keep these 4 aircraft parked or to fly them.

      It seems to me that they have determined/guessed that 3 of these aircraft can be put to use in such a manner that they lose less money flying than when parked over ther remaining three years of the lease term. (The 4th, "standby" aircraft is effectively still parked).

      Otherwise it would be hard to understand the point of placing them into service since one knows a priori that they'll bleed profusely in the first months/year of service (as they in fact have).

      Hope that makes some sense. (I am agnostic about whether flying them is a good idea.I think it matters at the margins but it is not, I think, central to the thesis).

    4. yes, does make sense. thank you.

      i do admire the your knowledge and understading in each of your investments. do you really just spend 2 hours a day researching? if so, i really wish i had your brains!

  21. red, thank you very much. Now I hope to see what you are seeing. I do not know the management team, but their decisions look good so far. You can really distil such a situation to a simple but powerful model. This would be very difficult for me to have seen.
    Your analysis of airlines was especially convincing to me (HA and now FlyBe).

  22. Saad can be heard preening in this recent podcast

  23. results:
    For flybe I really do not understand the chart. I hope I am more right than the other market participants. I like the conservative tone about capacity expansion. Numbers are a little lower than expected, but no reason for the share price to go down. Cash flow is good already.

    With them owning more aircrafts they will have more residual value risk going forward. Do you know whether they want to fly until the aircrafts are fully depreciated?

  24. Martin,

    I don't know and I doubt that they themselves know. Average fleet age is 7 years so that issue becomes relevant several years from now and at the margins.

    re: the aircraft / capacity expansion issue: In addition to the cost savings associated with owning v leasing, Flybe's implied market valuation will also compress as the numerator in the EV/EBITDAR formula comes down. If you're into that sort of thing.

    That fellow sitting at the front -- "Andrew" -- he's the pace setter wrt sentiment on this stock. You might evaluate the nature of his questions in the Q&A to get a sense of what the market is thinking.

    1. red,
      EV/Ebitdar and those relative pricing stuff is near meaningless to me. In the end I care about sustainably distributable cash flow, which is difficult to arrive at. I have thought about the capital expansion cited as a reason for lower numbers. My thesis was Flybe is not competing directly with the longer range airlines. So where exactly was the capacity expansion? Industry wide capital expansion could even be good for flybe, if this increases demand and flybe could make more feeder flights from small local airports. At the moment I have no high conviction, because I do not understand why this affected flybe negatively. I just want to go long and wait some yrs.

    2. "My thesis was Flybe is not competing directly with the longer range airlines"

      No, but it does compete in the short haul market with Easyjet and some others. Capacity expansion by others at rates above demand growth in, for example, intra-UK flying will reduce load factors for everyone. (If you're looking for particular routes, I suspect that London City - Dublin is the biggest culprit.) Plus, capacity expansion in addition to lower industry-wide air fares caused by lower jet fuel will have hurt Flybe in particular because Flybe jet fuel costs haven't come down as much as they have for others.

      So, looking at the figures in the post above: 2015 illustrative profit + adjustments + say a 30M/year reduction in jet fuel = 36.2 + 30 = 66M in earnings on 10.3 million seats compared to today's market cap of ~110M.

      The company says that 7.7M of these seats have experienced Rev/Seat growth in 2015 and also in 2016. This must have come from load factor improvements rather than from yield (because of the trend in jet fuel prices over that period). Incremental load factors/revenue, as we know, are almost entirely incremental profits so the earning power is probably higher than 66M. Plus aircraft ownership vs leasing will add more than 8M/year so that earnings power is > 70M.

      On the other hand, Flybe's mkt cap is 70M implying 10M in earnings/FCF. So one doesn't need to know what distributable cash flows will exactly be. One just needs to understand how the additional Q400s will lose more than 60M/year on a sustained basis. Is that even mathematically possible? That's the question in my view.

    3. red,
      the distributable cash flow is always important. Earnings can take time to translate into cash if ever. Buying the aircrafts will consume some cash. If they were to distribute the 66m this time next year, the share price would be much higher.
      I have quite some creativity and can envision them loosing money:If there were crazy capacity expansion like in some ship markets, all used aircrafts could be rendered economically useless. Only homogenous and new fleets with very low variable cost could be flown. To be fair: the supply side for aircrafts is way more rational than for ships. Although there are more suppliers for the smaller ones than the two for the longer range ones.

      The tailwind from lower fuel will not be sustainable unless Flybe were the lowest cost provider. Why should all profit flow to flybe with increased competition instead to the customer? I will wait and find out and with the volatility maybe trade a little bit around the position.

    4. "I .. can envision them losing money"

      Ah, okay. No reason to force conviction if you don't have it.

    5. No one can force conviction except if you are some kind of fanatic following a guru. Only education will help. I am long Flybe, but I can always see a path to loose money on every of my investments even German bunds or whatever. Maybe I am more diversified than you. I never had a position above 40% for example.
      Strange things can happen like a new vulcano, more terrorist attacks, rising oil price, reckless capacity expansion by competitors, stop of subsidies for local airports, plane crash, better infrastructure for rail and railroad, mgmt leaving etc.

  25. Hi Red,
    I am very enjoy your blog
    I have several small questions regarning flybe earnings release.
    Flybe posted that revenue per seat declined 12% at H1 till june. If this rate will continue it will bring revenue per seat to 49pound for H1. If we estimate oil price reduction of approximatlly ~35% it will save 3.5pound for seat costs. Even if we will take your assumtion of 41.5pound cost per seat ex fuel we will get 2pound earnings before taxes per seat. At this schenario Flybe will earn only 10-12 Milion pound at H1. I know tha you modeled much higher earning can you please explain more detailed your assumptions?

    1. Thx for the question. There's a long answer and a short answer. The short answer is that

      1) On the COST side, my model above excluded 26M (what would now be 20M) of annual so-called "excess capacity costs" which will wind down over the next two years as the leases on the E195 aircraft expire.

      2) on the REVENUE side, Q1 2017 average load factors and yields are down because
      a) of the Brussels terror events,
      b) the E195s are now flying (and no longer parked as excess capacity),
      c) of industry capacity growth.

      I have attempted a rough model to account for the decision to fly the E195s here.

      You might try and unpack it and see if it makes sense to you: i.e. What are the legacy routes earning (roughly)? Are losses from the E195s capped at a certain amount? What's the rediual value (i.e. legacy routes less E195 libility)? Given that residual value what kind of performance from new Q400s can Flybe's current market value support? Is it close? Arguable? etc

      That's, in my view, the best way of grappling with what appears to be a complicated situation.

    2. Based on Slide 31 of the annual results presentation, I got a revenue per seat of 53.10 for the "legacy" routes that they left alone. Is that what you've got also?

    3. I think that's a bit low. We know that the new routes are -27% network average so 38.9 per seat. That implies the average of the remainder = 570/10.5 = 54.25 per seat. And I think that number is a good enough representation of the "legacy"/"unchanged" routes. (Push comes to shove, I myself think that "unchanged" weighs in in the high 54s and the tweaked legacy routes in the high 49s.

  26. Thanks for posting!

    Q re the rough model. Would you say that new seat kms are largely e195? If yes, wouldn't the cost per seat be very different from q400?

    If the costs per seat are (significantly) lower then its possible that you are overestimating the profit of "unchanged" routes and overestimating the loss of the "new" routes.

    I don't know what the cost numbers should look like for q400 v/s e195. But, I am hoping to get educated about it in your reply to this post. :)

    1. 1) I think you have to think back to why it is that they parked the E195s in the first place. Why park the low cost aircraft and fly the expensive aircraft?

      2) Cost per seat is something of a nonsense metric, honestly. I have use it because the company uses it but it is entirely free of content except when used to compare two or more time periods. Profit/seat is less nonsensical so I'd approach the analysis in the latter terms. So, reframing your question: Am I confident that profit/seat on unchanged routes is higher than on new routes? The answer has to yes.

    2. the cost per trip: e195 > q800. but for the cost per seat: i would have expected the reverse to be true. correct me if wrong.

      i was trying to construct flybe without the e195s and therefore estimate the costs. and hence the q.

      another question, do you know when the leases expire for the e195s? it seems like it happens over the next three years based on the 10m, 6m, 4m. however i cant seem to find anything in the ar or elsewhere to confirm this.

    3. Fleet plan = slide 45 of 2016 results presentation.


      "i was trying to construct flybe without the e195s and therefore estimate the costs. and hence the q"

      Okay, this helps.

      2015 is what Flybe looks like without the E195s. They were parked and the numbers are as in my post above.

      So compare costs before fuel, depreciation and rent
      2015: 325M on 10.3M seats = 31.56/seat
      2016: 379M on 11.3M seats = 33.57/seat
      The difference in costs is 54M on 1M seats = 54.23/seat

      Back of the envelope: 8 additional Q400s x 64 seats each x 2 flights/day x 365 days = ~0.4M seats. If they cost roughly the same to operate as the legacy q400s they will account for ~12M of the additional operating expense in 2016.

      That leaves 42M to be split between the establishment of new bases and the operation of E195s.

      So, taking our cue from incremental ground ops expense per seat we might estimate that 15M to 20M of incremental expense is related to newly established bases leaving 34M to 39M (or 54/seat to 62/seat) as the cost of operating the E195s. Before fuel (the E195s of course burn more fuel) and before the 20M in rent expense associated with them. Anyway that's higher than the 31.75/seat associated with the Q400s.

      That's the sort of process I'd follow to try and estimate the cost of flying the E195s.

      Or you could use the 2015 ex-fuel cost/seat numbers as approximately right, adjust for fuel and thereby get a good enough picture of what Flybe looks like without the E195s.

    4. "2015 is what Flybe looks like without the E195s"

      the following is data off caa. seat km available and used ('000) per aircraft type. in 2015 e195 flew 706M seat km v/s 931M in 2016. and here the year = financial year (apr-mar).

    5. Thanks for posting.

      I should have said, 2015 is what Flybe looks like without the EXCESS E195s. There were 14 E195s in 2015 of which 9 were parked and 6 were in service. Those 6 are no longer with us having been returned to the lessors. Their routes (including charter flights) have been taken over by E175s and Q400s so that their cost structures are better today than they were in 2015 and "Flybe without the previously excess E195s" is a little more profitable than the 2015 numbers above suggest.

    6. Thanks for replying.

      off topic - brexit - i am looking at names that will benefit disproportionately with a remain vote. flybe is probably a candidate. some car retailers look to have had 20-30% drops. i was wondering if you had any names you thought are possible candidates?

    7. Brexit would create an unholy mess in the real economy. Housing & benking would come under serious pressure, capital would flee, interest rates would rise -- that much seems clear enough -- and it would take some years to recover from the resulting recession. So I'm inclined to wait and see how securities prices (on both sides of the Channel) shake out first.

  27. Thanks for posting so much information about this idea. I have a basic question: How did you build your model for the "legacy" routes you included in your Flybe Precis spreadsheet? In particular, how did you determine how many seats to put in this bucket, along with the expected yield and load factor?

    I have been interpreting your reference to "legacy" seats to refer mainly to the profitable UK regional airport to regional airport routes that face little to no direct competition from other airlines. I think this understanding of "legacy" routes is fairly close to the routes Flybe was flying in H1 2014/15 (prior to the London City expansion). The H1 2014/15 reports a yield of about 70/seat, which is what you use, but load factors are in the mid-70's, rather than 83%. (It looks like load factor peaked at 83% in August 2014.) In addition, in H1 2014/15, Flybe was flying at an annual rate of over 10 million "seats," rather than the 9 million that you use for your "legacy" routes. Did you somehow bridge from the reported H1 2014/15 figures to your 9 million seat "legacy" routes, or did you take a different approach altogether?

  28. Nevermind the question above. I now see that you posted two other spreadsheets that show how you got the numbers. Thanks again for posting your thoughts on this company.

  29. happened. UK is out of Eurozone. Does this jeopardize the fate of smaller airports in UK? Flybe in trouble?

  30. The smaller airports are UK funded. Is Flybe in trouble? No, not at twice this price.
    If this Leave vote ends with the UK's expulsion from the free trade zone the dream of contract flying for continental flag carriers is dead as a door nail.

    If it ends with visa requirements for short duration visits then 20% or so of routes will have lower load factors than otherwise. (Assuming that a separate agreement can be negotiated with Ireland.)

    1. I do not know at the moment what this means exactly for Flybe. But I think at this price with Brexit it is a better buy than before without brexit. Maybe I am wrong, but I am purchasing and making this a full position. My base currency is not GBP, but I also do not see that much downside left.

    2. This looks pretty asymmetrical at the current price, but let's entertain the bear case for a minute -

      Brexit drives financial services and other international businesses out of the UK; price inelastic business travelers migrate to another country, or companies start paring down their travel budgets. Volumes and pricing drop for Flybe on their profitable regional routes, to a point where where we start burning cash (need to calculate what breakeven revenue is).


    3. Sounds like a good approach. The answer is in the details, of course:

      Framework questions may include:

      What % of travel is from London City Airport?
      What % of travel is business?
      What % of business travel is (likely to be) overnight?

      Emprical questions:
      Load factors and fares in the 2008-2010 period
      How profitable were Flybe's core routes in 2008-2010
      What is the $USD sensitivity of the cost base?


    4. I should add that I'm not a fan of asymmetrical betting so I am not intending to encourage that.

      It is safe or it is not. Whether it is safe or not is obvious (after the necessary effort in examining the company and the security) or not obvious. That's the way to proceed in my view.

    5. Not sure what you mean by "asymmetrical betting," but I just meant that at current prices there's more upside than downside. Ie it's harder to create a scenario where the company is valued where it is, than it is to create a scenario where the company is worth more. I think every stock we determine as a good investment should have an asymmetrical upside by definition.

      Anyway, you're right about the details. Looks like 50% of passengers are business travelers - will need to do more digging on the other facts.

  31. "Asymmetrical betting": Downside = -x, Upside = 5x. I can imagine both so it's a good deal.

    - given the data available to me;
    - given my understanding of the industry and the dynamics of its environment; and
    - despite a healthy respect for unknown unknowns

    I nevertheless cannot REASONABLY concoct a scenario in which the equity (in this case) is worth less than the current market price (plus a litlle something something for opportunity cost).

    In a rising market the asymmetric bets will work well enough. In a flat or declining market it is generally best to forego those and wait for the securities that you're sure about. (Unless you'e hedging by, say, simultaneously shorting EZJ or RYA or some other biosimilar in which case -x v 5x works well enough).

    That's what I was getting at

  32. Flybe Is An Intriguing Turnaround Story With An EV/EBITDA Ratio Of Less Than 1

  33. Hi Red, just starting to dig into FLYBE. Management, in their 9 June 16 update, before Brexit, indicated a 5% decrease in yields and 12% decrease in passenger per seat revenue. April said to have been grim. they note a definite slowdown of demand growth, oversupply with Easyjet, Ryanair too excited with low fuel costs. In my back of envelope estimate I have it 67% load factor (well below their 70% target)and 66.7 yield. That would translate into a UK revenue of just £503m for Full Year (again, before factoring in any Brexit impact). That compares with last FY total passenger revenue of £572m. Compensating this, fuel expenses should be down £37m..Looks like, as they write, any reduction in fuel costs is "competed away" ! I am afraid a big operating loss for the year is well in sight, and Net cash is all well and good but the off balance sheet obligations are a multiple of it still...Suddenly current price seems not such a bargain or am I missing something? Again just very preliminary work done here but does not look the total no-brainer that you depict!

  34. "I missing something?"

    Maybe. The investment rationale is in the original blog post and in my replies to Qs in the comments section so have a look at those.

    ATR72s and Q400s = monopoly routes, per seat spread of 4 to 6
    E195s = competing with EZJ and RYA, negative spreads of 8 to 10/seat
    Q400s will replace E195 capacity over the next 3 years so 5 million seats will swing from -9 margin to +6 margin = 5x15 =60MM before fuel benefit, lease/buy benefit etc


    In 2015 they parked 9 E195s rather than fly them
    The cost of parking them was 26MM
    which means the the negative CASH direct operating margin on these is > 26M
    9 E195s represent 3.8MM seats so negative cash margin per seat > 6.85
    Overhead allocation/seat = 10/seat
    Therefore negative/seat of flying those nine 195s > 16.85/seat


    If you follow that reasoning you'll get to where I'm at (again, as laid out in the post above and also in the comments).

    If you think this is a turnaround that hasn't turned the thing to do is wait.
    If you think the "analysis" is a bit fancilful and/or effete -- this idea is obviously not going to work for you.

  35. Thanks a lot Red, I really appreciate your analysis, angle, and quick feedback. You say 9 E195s represent 3.8MM seats, but those aircraft carry 118 seats, so 1062 seats/day x rotation (average rotation at FlyBe seems 5x day correct?). So more 1.9MM seats per year right? cheers

    1. Good question, thanks.

      Flybe started FY2015 with 14 E195s and ended it with 9 so average of 11.5 E195s
      9 flown at 4x/day and 2 parked
      Average utilization of Q400s, E175s etc = ~8x/day
      Average utilization of other aircraft is 2x that of E195s because their cost/seat = 1/2 that of E195s (breakeven load factor of 55%-60% on 78 seats vs BELF of >80% on 118 seats)

      Therefore maybe Q400s flying routes previously flown by E195s could manage 8x turns/day: lower BELF = more frequencies and competitive fares v EZJ etc

      If so, 3.8MM to 4MM seats. If not, ~2.25MM seats

      So anywhere between 2.25MM and 4MM is reasonable in my view.

  36. Of interest to those following this comapny, the HSBC analyst -- i.e. the only independent analyst following this company -- yesterday changed his price target from 100p to 30p to account for Brexit.

    Here are the changes to his numbers:

    1. In fact, HSBC's price target now seems to be 20p. Looking through their valuation, they compute with capitalised leases of 297m GBP, which seems to be the key difference. Why do they not cap them? Computing with leases of 40m GBP, you'll end up with about 150p (keeping their other numbers).

      Thanks for the great posts, red. Like many others here I truly value your blog.

    2. Scratch the second part of that message, which is clearly wrong. Apologies for any confusion.

  37. Hi Red - the HSBC analyst changed his model and he is now projecting, for fiscal year 2017, 13 million pounds less of top line PNL growth (compared to his retired model) - so because of operating leverage, is the stock is now worth only a third of his last estimate?
    Does that even make sense? Or maybe the street’s pessimism is taking him over?

  38. "If you want to beat the S&P 500, here's what you do, you buy 500 stocks, and then you sell the airlines. You should do better." - Tom Gayner

    Love your blog red

    1. Hey red, I'd be interested to read what you make of the trading update for Q1.

      In addition, I wonder whether you have more details on the 3-stage DCF model HSBC seems to be using. Often, a lot of the value concentrates in the last (terminal) stage of the business -- the stage which is also the most uncertain. Have you used such multi-stage models? Are they worth it in practice?

    2. Hi Alex, I've uploaded the 100p and the 30p valuations here
      and here

      I thought the Q1 performance was fine. I have a detailed model of what I think is going on with this company's performance that I'll post when I return from vacation.

      re: 3 stage models
      My preference for for airline stocks is to see if one can get to below 2x FCF and good enough capital structure two years out. If one can and the core routes are more or less reliable and noncyclical then it's a bargain. All the valua in a DCF model is then front loaded and one then only has rely on reasonable capital return behavior by management. 2x FCF & 30% payout==> unsustainable dividend yield. I think it more likely than not that Flybe's payout ratio will be 50%+. I think that's as sophisticated as it ought toget from a multiples perspective. From a valuation perspective the hard work ought to be in estmating the earnings power of the 7 or 8 million seats at the core of Flybe's operations.

    3. That seems kind of a strange way of putting a 30p target on there.

      -It seems they do not account for the e195 issue? Leasing costs will drastically come down, wasn't that a 20m drag on 2016 results? And it seems they do not properly account for higher % ownership of planes in the coming year? Even in their old model.
      -It seems strange that they think a market cap of 65m is fair value for a business that will generate 41.8m in earnings in the next 3 years according to their model.
      -Why is there a sharp increase in ground operations cost in their model? Isn't this mainly a GBP cost instead of a USD cost?

      And a question about your model, what happens to their large dollar costs 2-3 years out if the pound stays weak and the hedges roll off? Will they raise prices? That is my main concern. If you throw another few terror attacks in the UK in the mix + weak pound and strong dollar for the next few years, what does this look like?

  39. Just back from vacation; sorry for the delay in responding. I think he's phoning this one in. Small, complicated company. Easiest to cut yield by 6% -- a number apparently picked at random -- and be done with it.

    As to the GBP/USD issue 2 or 3 years out:
    1) Aircraft rent/ownership, fuel, and maintenance costs are USD denominated. So these costs will rise by the amount of the GBP's depreciation.
    2) On its monopoly routes, Flybe has room to raise prices. I think the evidence is that the core + monopoly routes are price inelastic.
    3) by the time 3 years have passed its net cash position ought to rival or exceed its current market cap. Even a very low multiple on depressed future earnings on free cash flow or earnings - 2x say - ought to mean that the future value of Flybe's equity is higher than the market's apparent expectation.

    Nevertheless, airline stocks are subject to volatile changes in market sentiment and are exposed to terrorism, Brexit, volcanic ash, and whatever else. So, in my view, these are not ideal investments for low vol 20% or 50% returns but instead for high vol 3x or 5x or 7x returns.

  40. HI Red,
    Hope you had enjoyed your vacation.
    I've kind of been an ardent (and silent) reader of your blog with everything related to flybe, and it made me want ask, what are your favourite investments blogs?

  41. Well thanks very much for reading.

    Favourite blogs. I'm a flaneur, as you may have suspected, so the "styles"/industries/markets that I'm interested in tend to change according to circumstances. Nevertheless, an abbreviated list of evergreens on my reading list include
    -- Geoff Gannon's blog,
    -- OTC Adventures,
    -- Reminscences of a Stock Blogger,
    -- Punchard Investing,
    links to each of the above here:
    -- Monte Sol here

    Plus I follow "Packer16" at

    Plus I find the market for small and midcap UK companies perennially rich with possibilities and therefore enjoy Richard Beddard's blog and

  42. Thanks. I'm a novice in investing but hope to catch up quickly and contribute back to the conversations here.
    Would it be ok to shoot you an email with general questions about where to start?

    1. Sure, send me Zn email. I'll do my best to be helpful

  43. will do. Are you following Wilhelmsen (H/T OTC Adventures)?

    1. Not really. Normalized dividend of approx 2/share implies ~ 10% yield best case. That's below my hurdle rate.

  44. And what about its spun off company, Treasure ASA?
    it's basically a holding company that holds ~750m USD worth of Hyundai Glovis Co Ltd shares, but trades at 61% to that NAV.

    1. No idea of that sort would really ever make it onto my watchlist.

  45. because there is no catalyst?

  46. Hyundai Glovis dividend yield = 1.8% and growing at 1% in inflation-adjusted terms
    So Treasure's yield at 60% discount = 3% and growing at 1%
    That's your upside if all goes well. Does that sound good? safe?

    Note: 10 year Korean govt bond yields 1.4% today.

    Scenario: Korean pensioners have bid up these shares as govt bonds yields have fallen. When govt bond yields rise again the whole process will unwind. At a 4% 10-year govt bond yield Glovis' mkt cap will fall to US$ 300M. Treasure ASA's market cap at 40% discount will be priced at $180M. An investor today loses 60% of the value of his/her investment plus opportunity cost. [This unwinding process takes 5 years to play out so you've lost 1x(1+10%)^5-1 = $0.60 in opportunity cost for every dollar invested in Treasure -- i.e not trivial].

    Is there a definitive, beyond-reasonable-doubt counter-argument to the above scenario? That's the starting point for an analysis of this investment, in my view, not a discussion of abstractions such as NAV discounts and whatnot. [Value investing is in my view nothing more than a shunning of abstractions].

  47. Hey red,

    any updated thoughts on Flybe here? They are ridiculously cheap here, but do will they be able to hit ~70% load factor in the next few quarters? Appreciate your ongoing thoughts on this, and congrats on a good year so far.


  48. Hi Andre,

    Not sure what there is to say that's new/interesting.

    If we take out working cash equal to 10 weeks of revenue then Flybe's EV is ~90M
    If we add fuel benefit to last year's CFO before working capital movements then adj CFO is 75M
    Maintenance Capex is minimal, say 10M
    so FCF/EV = 72%

    Absent the E195s, pro forma FCF/EV is >100%


    That kind of market valuation implies that the business model is utterly broken. Breakeven load factors for the Q400s are in the low 50%s and for the E175 LFs in the low 60%s. Can that possbly come about given that Flybe can return the E195s and up to 8 Q400s and therefore choose to retrench into its most profitable routes? I doubt it.

    We'll see in November what the summer looked like. Until then I don't know that there's an awful lot to say.

  49. Thanks again for writing your thoughts. If I got it right, you're saying that:
    1. Too much downside in terms of Glovis falling and taking Treasure with it
    2. Since the Majority of Glovis shareholders are Korean Pension funds, they would not buy Treasure instead, despite the NAV discount, since it's out of their investment sphere.

    Is that it?
    In that case it's back to the drawing table.

    1. I'm saying that there are two ways to invest:
      1) figure out what something is worth to you & pay a price that is lower that that value; or
      2) figure out what something is worth to SOMEONE ELSE & pay a price that is lower than that value.

      The 2nd approach -- what hawe here -- requires macro, legal, institutional and anthropological knowledge. I threw out one possible scenario to challenge your perception that an idea structured as
      "X owns Y"
      "X trades at 60% discount to Y"
      "X is cheap".

      makes sense on its face. It doesn't. Y could be overvalued. The institutions that own X may be smarter than the institutions that own Y. Or they may have higher hurdle rates. Or the holders of Y may face legal/regulatory/social obstacles to owning X that they don't face in owning Y. etc. There's a lot to think about.

      If you are starting out in investing I think it is much easier to try to find stocks that you would like to own for your own purposes. You understand where the sales come from and why;, Why the margins are what they are; How much capital expenditure is required for growth; How much cash you'll receive back from the investment, when, and at what risk.

      In the case of this example, you would ask, "What is Glovis worth to me?". If it seems clear to you that you would only pay 3x earnings* for it then a good price to pay for Treasure would be, say, $60 million (i.e. half the intrinsic value of Glovis at a 20% discount). In that way you let trouble come find you rather than actively seeking it out.

      * Earns $1, pays out 20% = $0.20 dividend. 9% yield to accont for risk and 1% real growth of 1%. So $0.20/9% = 2.22 multiple on earnings. You've heard somewhere that Korean companies are maybe going to pay out greater % of earnings in the future and you believe that rumour so maybe you're willing to value it at 3x earnings to capture that potential.

  50. A beautiful explanation, thanks for the primer

  51. I think FLYB maintenance capex is higher than 10m.

    In the last presentation ( last page ) they showed 40 capex related to maintenance. I think we should include this as Mcapex.

    Looking back 7 years average capex is 30-40m. In line with other airlines at 4%-5%/rev

    Thank you

  52. Thanks for the comment/question.

    They changed the way they account for maintenance expense in 2016. If you're modelling aircraft expense I'd total rent, D&A and maintenance expense and then adjust it for the savings that will come from increasing the aircraft ownership %. They detail unit savings in that same presentation. Multiply unit savings by the number of Q400s they would have to own in order to reach their 50% aircraft ownership target. So:

    Depreciation of aircraft + amortization of maintenance reserves + non-capitalized maintenance expense + rent + cost of aircraft finance at 4% = cost of aircraft.

    If you're projecting out beyond 2019 remove the lease and maintenance expenses on the E195 aircraft.

    Historical capex and sector-wide capex/revenue are not useful guides.

  53. What about self driving cars? I mean it is still some time away probably. But usually in valuation you put a 10-15x multiple on earnings. So essentially you predict what will happen 6-8 years from now. It seems to somewhat put a cap on the upside.

    Also how do you treat the leases vs the cash balance? If you capitalize the leases, that is around 200 million or so? If they have to start shutting down bad routes if self driving cars were to be a problem in the future, they might have to deal with leases? And planes that are declining in value. I think they are moving to 40-50% self owned planes, what would their leasing liability look like in let's 7 years from now?

    Probably a longshot risk, but still worth pondering.

  54. Hi Red,

    An article about Saad Hammad (11th September)

  55. Holy cow: Saad resigned or maybe more likely was fired! That is a big surprise. What happened? Any disagreements on the way forward? Any skeleton in the closet? Or was he offered another lucrative position elsewhere? That is so odd, he had such an uncapped incentive package. The thesis looks weaker now IMHO.

  56. Hi Red,

    I were taking a look to godfreys numbers and the guidance for 2017 looks like consistent with the current price.

    The Board provides underlying earnings guidance for FY 2017, as follows:
    Sales $180 m - $182 m
    EBITDA $14 m - $15 m
    EBIT $9.4 m - $10.4 m
    NPAT $6.0 m - $6.6 m

    NPATX12=72 less 22 (net debt) = 50

    I suppose I am missing something, could you please give me your point of view.


    1. Plus receipts from initial franchise fees 60 x 0.5 = 30M
      Plus sell through of excess inventory = 10M less 20% discount = 8M
      Plus add bach restructuring expense of 2MM capitalized at 8x = 16M
      Subtotal: Your 50 plus 54 = 104

      Takes 2 years to work out so 2x annual dividends = 6
      Plus MAYBE they get back on track with stick vacuums = 40
      Total = 144

      But this is a retail operation so it's by definition not a great idea. Others -- Keck Seng, LXU, RTK, FLYB, Future Bright -- are more reliable ideas in my opinion

      But one can really only rely on 90

    2. or CVR Partners which reports tomorrow and seems to me is a gift if it were to dip under $4.50

    3. Why is their cost of debt 9.25%? Couldn't they get cheaper financing if their assets are that attractive? Makes one think the potential 20-25% yield on the equity would be much more risky as well. Seems to me both UAN and LXU are more like attractive long term options at this point. If the pricing situation does not improve significantly within the next 2-3 years you are toast.

    4. Hi Red,
      Note that CVR dropped yesterday to 4.56

    5. Thanks Lior. I own a stake in CVR Partners via Rentech.

  57. management buyout in the offing imo

    1. 33.25 GBX -- efficient markets at work. I'm with Oscar Wilde on this one: "Nowadays people know the price of everything and the value of nothing."

      I have been playing with your model (Flybe precis.xlsx) and find it really hard to bring the price down to 30p on assumptions which are sane/data-driven... am not so sure about the aforementioned MBO hypothesis.

      But we'll see shortly. I'm not counting on sentiment turning around immediately, but am happy to wait.

      Thanks for sharing all your inspiring work, red!

      ps May I ask how you learnt the business model analysis/valuation/accounting? It seems to be a bit of an art at times, particularly when it comes to accounting (what gets added when, how to approach a business/industry). Position sizing/portfolio allocation one can manage, but the true art is valuation (which is the basis of allocation, of course)!

    2. I'll write & post some more detailed notes after it reports. Thanks for reading and for playing around with the model.


      Accounting & business models

      I read a couple of books on how to read financial statements but the ability to read them was mostly earned by just the cumulative impact of having gone through very many of them over the years.

      One can make it less dull by being actively curious about the business itself and relating the accounts to that. How is it possible, for example, that Howden Joinery reports 64% gross margins from selling kitchen cabinets? Or rather, what does it mean to say when it reports 64% gross margins? A question like that draws one into the business model proper. And that in turn informs one's understanding of the accounts, an so on in a virtuous circle. Understanding the concepts (e.g. how the meaning of "revenue" can vary according to context) and how the three statements/accounts are interlinked and must flow in and out of each other is probably the first thing. The most important thing is not to skip some term, item, or relationship that you don't understand but to instead get to the bottom of it.

      Valuation is (a) a good model; and (b) a price that is robust to sensitivities. Therefore one needs to find and map all the major "drivers" and arrange them in the right way.

      e.g. for Flybe:

      ex-fuel cost to operate one Q400 flight, 230 nautical miles
      ditto one E175 flight, 330nm
      ditto one E195 flight, 330nm
      ditto one Boeing 737 flight, 300nm

      absolute amount of fixed costs

      elasticity of demand for business travel
      ditto VFR demand
      ditto summer holiday demand

      % of costs that are dollar denominated

      distribution of margins by # and type of routes


      then one can arrange them in such a way as to estimate break even load factors at various fuel prices, exchange rates, yields.

      (as it happens, in Flybe's case, establishing the direct operating cost of a Q400 flight over 230 and 300nm is the key task)


      If you know virtually nothing about a business/industry it is helpful to read the risk factors the annual reports and also all of the notes to the financial statements. Then the IPO prospectus and/or bond prospectus. Then equivalent documents of similar publicly traded companies. Then some (online) trade journals. You could also maybe try to find investment theses from years ago and look at how the author understood the business and how events unfolded from there. That way you can first map out what you wouldn't have known to ask about.

      After that it is simpler -- you decide what it is that you need to know and either you find answers or, if you can't, you move on to the next potential opportunity.

    3. Thanks, red!! you've made multiple good points that I will take away from this!

      Any books on accounting you would recommend? I've got Damodaran's investment valuation, but it's a bit lengthy/dry and not necessarily about accounting. I feel like there should be a book which explains business models and accounting in lockstep. Not sure this exists. But there's always room to write one! =)

    4. I've seen people recommend this one

      Other useful books:
      'Understanding Michael Porter' by Joan Magretta
      'Of Permanent Value' by Kilpatrick and 'Snowball' by Alice Shroeder
      'The Art of Profitability' by Adrian Slywotzky

      The Kilpatrick and Slywotzky books will seem a little lowbrow to you but my suggestion is to put up with their tone and format and meditate on the content.

      And finally, this book
      lays out the practical challenges in investing in a way that is interesting and rings true to me.

    5. Thanks! Again! If you're still around once I'm better at this, I'll pay you back! In ideas.

    6. This comment has been removed by the author.

    7. Thanks red for your ideas. I thought i will share one of my own. Over the last few days I have been building a position in sunteck realty. A Mumbai based property developer. 74% promoter ownership and increasing. Selling for less than book. Where book doesn't reflect the true value of underlying assets. A reasonable debt position.

    8. and KKR involvement? It looks promising. Thanks for sharing

    9. Apparently a 5 % stake in 2 of Their bandra projects.

      There is a report from indianivesh which is a quick summary of the assets.

      Also would recommend reading about how things started with their partnership with ajay piramal for the bandra land and then kotak for the goregaon to get a sense of the promoter. Ajay piramal and ashish dhawan have stakes in this name.

    10. Right-ho. Thanks. I'll read.

  58. Thoughts after the H1 2016/2017 release?
    -Would have been nice to have more details about Saad Hammad's departure.
    -Is the 47m pension deficit an issue?
    -Using their Stated Adjusted Pre Tax Profit of 15.9m, accounting for H1 being their busy time, accounting for 20% UK taxes, 12x multiple, is 1.28p/share. Looking at pro forma EBITDAR numbers taking into account rent without the E195s, has the return looking better.
    -Without a doubt, they are still operating less efficiently than they will given the 3 new plane arrivals in H1, E 195's, the trend to owning more aircraft than leasing, "challenging market", etc., all contribute to less than an optimal fleet and margins lower than they will be soon. Nice tailwinds here.



  59. Hi Tom,

    I didn't see anything alarming in the results.

    The March 22 terrorist acts in Belgium impacted the results by 5.5 million via lost flights & loads
    + air traffic controllers' strike in France had a similar scale impact.
    + the 15m or so impact from the fall of Sterling.
    + new routes.

    It's best to assume that $1 to $1.25 is sterling's new habitat.

    Anyway, take out the 195s and it looks better.

    Reduce the fleet by 5 or 6 Q400s and it looks better still: the 175s fill in for the non-excess 195s; Q400s fill in for the 175s; and the least attractive Q400 routes are dropped.

    Operationally and strategically they are doing the correct things: going where the codeshare potential is and either not competing or competing with flag carriers that can't compete on price.

    Earnings power = roughly 100 million with passporting rights and something like 40 or 50 million without.

    Back of the envelope:

    Cost of a Q400 flight = £3,100 (assuming $500 fuel + $1.20 USD/GBP)
    Overhead/Fixed costs = £950 per flight for the first 120,000 flights and £100/flight thereafter
    Breakeven load factor at £70 yield = £3,100/(78*£70) = ~58%
    Target 70% load factor on 120,000 flights ==> £675 EBIT per flight x 120K = £80 million EBIT

    Flybe operated 123K flights in FY 2015/6.

    Plus the action envisaged in Slide 34 of their presentation. On the extreme left are the high frequency routes -- London City to Edinburgh, Belfast City to Birmingham, Manchester to Paris, etc -- that they won't be able to reproduce. But they can
    1) cut off the right hand tail by returning some Q400s; and
    2) find routes on the continent, e.g. Dusseldorf to Milan, that can handle a 10x to 20x weekly schedule -- that would be a bonus.

    However, a ban on Europeans or others entering the UK and/or a loss of airline passporting rights would constitute a severe blow to this thesis

    1. Re slide 34, is considering per route profit after taking into account the total cost? If so, excluding the very few routes are actually loosing money. Which doesn't make sense. So wondering if you know or can guess what marginal costs they might have included?

    2. Thx for the question.

      It is a margin contribution slide and therefore includes direct operating costs while excluding items commonsly classified as overhead.

      Direct operating costs are variable costs + aircraft ownership cost:

      Flight & Cabin Crew
      Maintenance & overhaul
      En-route & navigation
      Landing & departure fees
      Fuel & oil
      Handling & parking

      + Aircraft lease expense
      + Aircraft depreciation expense


      Exluded are indirect expenses:

      Flight crew training
      Flight crew others (other)
      Depreciation of ground equipment
      Station costs
      Passenger services
      Advertising & promotion

      Indirect expenses sum to 130 million pounds and are not closely related to volume:
      2011 = 141
      2012 = 130
      2013 = 140
      2014 = 132
      2015 = 128
      2016 = 131

  60. Laffin buying shares...

  61. Red, which broker do you use to buy UK penny stocks? IB seems terrible with penny stocks in general.


    The cost of her salary at year one could reach 1m pound.

  64. I found this article quite interesting:

    1. Here are a couple of companion pieces if you haven't seen them

    2. Hi red,

      There's an interesting article in the FT about landing rights w.r.t. Brexit:

      Taking a bit of a simplistic view, it seems that the main risk of Brexit is UK carriers losing the "8th & 9th freedoms of the air", effectively meaning all UK airlines must have routes leaving or arriving in the UK.

      If Flybe reduces its fleet over the next 2 years, cuts unprofitable routes and does not unnecessarily ramp up extra-UK routes (neither starting nor langing in UK), Brexit may even have a positive effect for Flybe in the short-term since LCCs like easyjet cannot operate the inner-UK routes unless they start leasing smaller airplanes and directly compete with Flybe. Other EU27 carriers will not be able to operate domestic UK flights in this scenario.

      Other than that, I'll go with Munger/Franklin: Patience - the art of waiting without tiring of waiting. Hope the market gains some confidence in June.

      Thanks for your blog!

    3. Well Azerbaijan, for example, is an EASA member so it is not impossible that the UK remains a member.

      If the UK's negotiating stance/behaviour is such that it is expelled from EASA -- entirely possible, in my view -- then we'll need to watch what easyJet is/isn't able to accomplish for itself with respect to EU-compliant airline operating certificates and so on.

      As things stand today, with jet fuel futures and the USD/GBP exchange rate being what they are, Flybe ought to generate 55MM to 60MM in pretax profit in FY 2016/7 and between 105MM and 110MM in FY 2017/8. Given that I think it ought to be able to withstand all but the most dire post-brexit scenarios when 2019 arrives.

    4. Here are a few more articles touching on Brexit impact. In a scenario where Flybe loses the ability to fly intra-EU routes, then it seems Ryanair would likely lose the ability to fly intra-UK routes.

      UK-based airlines told to move to Europe after Brexit or lose major routes (Mar 22)
      "The UK could react to the imposition of EU ownership rules on airlines by developing ownership rules of its own, which could prevent carriers such as the Ireland-based Ryanair from flying UK domestic routes, as it does today."

      Ryanair gives Government one year Brexit DEADLINE – or threatens to CUT UK flights (Feb 25)
      "Ryanair bosses have threatened to cut capacity and domestic UK flights unless the Government has a Brexit plan in place in one year’s time. Kenny Jacobs, chief marketing officer, said: 'Airlines plan a year in advance, so that means we need to know how is this going to work by the end of February/start of March 2018.'"

      Interview w/ new CEO (March 2):
      "Ms Ourmieres-Widener said it was ultimately “too early to comment” on the impact Brexit will have on Flybe. Unlike peers such as easyjet, she noted, Flybe was 'not looking at buying an AOC (Air Operator’s Certificate)' to allow it to fly across Europe should the UK lose access to the European aviation market. 'Our main focus remains much more on the UK,' she said. 'Eight-five per cent of our bookings, our reservations are coming from the UK market. The exposure is a little bit different.'”

    5. Thank you, I hadn't read the interview in the Herald. The risks/uncertainties can be grouped as follows:
      1) AOC for intra-EU operations
      2) Visa and landing rights requirements between individual EU countries and the UK
      3) the impact of Brexit on the UK economy and therefore on the demand for business travel
      4) the GBP/UK and GBP/EUR exchange rates

      Hard Bexit is not net neutral to Flybe's prospects. The issue is whether, in two years' time, Flybe will be in good enough shape operationally and financially, to withstand the likely severe economic consequences of leaving the EU without a deal.

    6. Hi red,

      How do you get to 55-60MM pretax profit in FY 2016/7? Do you assume half of the new routes have matured?

      From this year's trading statements, I estimate FY 2016/7 as follows:

      Seats: 12.7m (8.7m on legacy and 4m on new routes)
      Pax: 8.9m
      Pax revenue: £667m

      Pax yield: £71.8
      Load factor: 70.1%

      Pax revenue / seat: £48.5
      Charter/contract flying/seat: £3.95
      Revenue / seat: £52.3

      Ex-fuel cost / seat: £45
      Fuel cost / seat: £6.67
      Cost / seat: £51.75

      Airline profit would thus come in at about £7m. For the moment, I think the forced capacity growth (and the resulting load factor decline) is roughly cancelled out by lower fuel cost. I think this is likely to continue into H2 2017/8, at which point routes will have matured and 6 planes will be returned. In short, I believe there is substantial value (barring a catastrophic Brexit scenario); just it won't shine through in the next year.

    7. Sorry for the confusion, I wrote FY 16/17 when I meant to say FY 17/8, i.e. the results in the autumn of this calendar year.

      Here's a precis

      I have used spot jet fuel of $1.80/gal and spot GBP/USD of $1.25. I have assumed no major terrorist events. For ease of use, I have also set aside any prognostications on route maturity.

    8. Ah, I see! Thanks, looks reasonable but still need to go through FY 17/8 in detail.

      Given the mentioned desire for network/route optimization, I think it's likely that load factors and yields will eventually improve after FY 16/7, and that may add another £20m-40m in FY 17/8, I think.

      The biggest risk to this thesis now seems to be the position of the UK government, sadly. I will now have to find out what alternatives to AOCs there are.

    9. Excess cash* at the end of FY 2018/9 ought to be about 200 million.
      Breakeven LFs ought to be:
      Q400 41% (vs current LF = 73.5%)
      E175 63$ (vs current LF = 70%)
      That's the thesis in a nutshell: can a business with these properties withstand hard Brexit?

      * "Excess cash" = Net cash less working cash equivalent to 10 weeks of revenue

    10. Another article from today's FT:
      Airlines warned on post-Brexit flight restrictions

    11. Hi Red - In your calculation how did you get the figures of - 1. Benefit of elimination of loss on disposals of 7M, 2. Contract flying benefit of 12.7M, and 3. Operating leverage benefit of 23.2? Also when comparing with 2015/16 figures you havent accounted for the fact that capacity went up after that and thus the corresponding earnings as reported figure is much lower. Just trying to understand the thinking behind the process

    12. I suggest that you link to a simple model that you have sketched out so that I can comment on it


    CFO stepping down...

  66. Flybe hopes to make call on airline's fleet later this year, says chief executive

    "We are in contact with all the OEMs (original equipment manufacturers), because we need to decide where we want to go moving forward," she said. "Our core market is point-to-point regional, so short distances. Within the group we fly Bombardier Q400, Embraer 195s and 175, and ATRs. So we know all these types of aircraft quite well." But Ms Ourmieres-Widener, who ran Dublin-based CityJet between 2010 and the end of 2014, insisted it's "too early to say" what the composition of the future Flybe fleet will be.

    "We are exploring all kinds of scenarios," she said. "It's a complex process." However, she said that the board could make a decision regarding the future fleet this year.

  67. any update on flybe would be appreciate Red.

  68. A couple of takes on the current situation and outlook:

    Legacy issues drag on Flybe
    "A turnaround plan may be under way, but we'd like to see some proof of progress under the new chief executive before we become more bullish on Flybe."

    Flybe plans post peak fleet profit progress
    "After strong growth since the end of the financial year to Mar-2015, Flybe's fleet numbers reached a peak in May-2017 and are now set to fall by 18% over the next three years. The airline's new CEO, Christine Ourmieres-Widener should reap the benefits of actions taken by her predecessor to limit the impact of surplus capacity ordered by his predecessor."

    Christine O-W has now confirmed that Flybe is in consolidation mode, with plans to reduce fleet size and stick to their knitting of flying UK-based routes, rather than pursue continental European expansion. It sounds like Saad Hammand was still too growth-focused for the board's liking, which is what precipitated the change.

    Hope everything's okay with you, Red!

  69. Sounds like the UK gov't is coming under increasing pressure to reduce or eliminate the Air Passenger Duty. This would disproportionately benefit Flybe, as its primary competition on most routes is rail and road. Scotland just passed legislation to phase the APD out, and Northern Ireland's DUP party is planning to push the Tories on the issue in exchange for their support in propping up the minority gov't (albeit 2 years from now, so who knows).

    Airlines welcome plans for Scotland’s new Air Departure Tax
    Scotland has pledged to cut the tax levied by 50% by the end of the current session of the Scottish Parliament and then eventually abolish the tax altogether.

    Newcastle airport urges government to cut air passenger duty taxes
    Newcastle airport is urging the government to cut airport passenger duty taxes in reaction to Scotland's plan to introduce reduced rates in the form of a new tax.
    Newcastle airport say Holyrood's decision is a particular worry for airports in the north of England. They're calling on the government to announce specific measures to support English regional airports from the impacts of Air Passenger Duty (APD) devolution to Scotland.

    DUP to seek corporation tax and air passenger duty cuts to back Tories
    The Democratic Unionist party is to seek further concessions on air passenger duty and special corporation tax status within two years of propping up a minority Conservative government.
    The abolition or radical cut to the air passenger duty for Northern Ireland’s three airports, which is not included in the initial deal, would be a “post-Brexit ask” by the DUP, party sources said. The DUP and the Tories both agreed that they would “review” APD, the abolition of which the airline industry in Northern Ireland claims would create thousands of new jobs and enable the region to compete with airports in the Irish republic, where the tax has been ditched. DUP sources said it would be opportune to demand APD’s abolition in the region once Brexit had happened and the UK was no longer bound by EU imposed rules on airline taxation.

  70. Stobart's CEO won't rule out approach for Flybe

    The boss of the UK's Stobart Group, ... has refused to be drawn on whether the company would be interested in buying UK regional carrier Flybe, but insisted that it is "looking at everything" to grow its business.
    Aviation sources claim that a takeover of Flybe by Stobart Group could make sense as Stobart looks to grow passenger traffic at London Southend Airport, which it owns. Among the routes Dublin-based Stobart Air is launching is a Flybe service between Dublin and Southend.
    We've got Stobart Capital as a separate business and his job is to look at entrepreneurial opportunities to bring to the group to support aviation and renewable energy. The question around Flybe, maybe you should ask us in two or three months," said Mr Brady, who took over as CEO last week.
    Stock exchange-listed Flybe, headed by former CityJet chief executive Christine Ourmieres-Widener, has a market capitalisation of just £77.5m (€88.3m), with its shares trundling along near historic lows.
    But while industry speculation has been growing that Stobart Group could be eyeing Flybe, Mr Brady pointed to the airline's challenged history. "Flybe is a well-chewed transformation story that's not been successful," said Mr Brady. "So, you really have to truly believe that you know better. "Jim French (a previous Flybe CEO) tried it. Saad (Hammad - who was Flybe CEO from 2014 to last October) has tried it. Both unsuccessful."

    1. I hope they won't -- I don't like their business and the way they run it. Does anyone know what the average prices for major shareholders (Aberforth, Hosking, UBS) are? (Before I go and try to come up with some.)

  71. Q1 2017/18 trading statement:

    I have them earning about £50m-£55m in H1 2017/18 based on the mentioned numbers (profit/seat Q1: £4.8, Q2: £9.3-£11.3), compared to £12.7m in H1 2016/17. And that's purely on load factor/yield improvements...i.e. excluding lease/buy, contract flying, E195 mitigation, etc. Is this too rosy?

    If they manage this well, we may well see more than the ~£60m profit previously mentioned for FY2017/18, I think.

    Also, one would hope the UK is now going for a 2-3 year transition deal with the EU, which would give this enough time to play out...


    1. 'm more comfortable with a figure of 40 million for H1 and 20 million net for 2017/8. By 2019/20 substituting Dash-8s for E195s will supply another ~25 million in cost savings and route optimizations another ~20 million or so in profit==> 65 MM total.

      However, this investment will not work unless Open Skies or a facsimile is in place by D-Day -- at least. Also, Bexit/depreciation has already cost 46 milliion in FY 2017 so the effect of the broader Brexit process on Sterling is going to be important.


    Memo from Howard Marks, very interesting,for me at least.

    1. Hi David, the link doesn't seem to work. Could you try to copy it again - or tell what the name of the memo you mention is (since I have found oaktrees memo page) :-)

      Regards Casper

    2. Googling "oaktree" "there they go again" should work

    3. Thanks red. However I thought it related to Flybe, but it appears merely to be a normal (probably warrented) warning from Howard which doesn't really relate to Flybe... :o.

      Regards Casper

  73. It would be odd for the founder of a multi billion dollar debt manager to write a memo on a sub £100mln market cap airline with no public debt.

    1. Laughing out loud at the thought of Howard Marks discussing Flybe at length.

  74. hey red, would you know what a typical code share agreement look like?

    does it mean a fee or certain assured num of seats bought? or would it just mean a potential of increased demand for seats given the tighter integration with the other airlines networks?

  75. My apologies for not having seen this until now. There are several types of codeshare structures but the most material to Flybe's business if the "hard block" variety. In a hard block arrangement, the marketing carrier purchases the block of seats at an agreed price. Flybe, as the marketing carrier, may buy a block of seats on Air France flights from Paris CDG to, say, Fiji at a discounted price. Flybe then markets flights from, say, Birmingham to Fiji. Flybe is then credited for the Paris-Fiji leg in its load factor numbers and reaps the difference between price paid for the block of seats and the revenues generated by those seats. Currently, these hard block seats account for ~550K of Flybe's annual passenger count, or ~5%of the load factor.

    Reciprocally, AF as the marketing carrier buys a discounted block of Flybe operated seats from Paris CDG to Birmingham, etc.

    Benefits = more stable demand for domestic flights (because certain number of seats are purchased by the codeshare partner and cannot be returned), better intergration with international destinations (e.g connecting Leeds Bradford to Air India).
    Risks = fixed purchase obligation on partners' flights

  76. Flybe eyes Ryanair cancellations to push its flights

    Ourmieres-Widener said Flybe was looking at chartering additional aircraft this winter in light of the Monarch collapse, but that any major capacity changes would be seen more next summer. “Airlines publish their network one year in advance, so your flexibility during one season is not so big,” she said.

    1. Thanks. Do keep an eye on the Brexit process, though. The UKG's posture and apparent obliviousness is alarming given the late hour. For example:

  77. Profit warning:

  78. Red,

    Supposedly the maintenance costs will going down as fleet is diminishing. This kind of extra cost is misguiding, but in H2 17-18 they could perfectly make 15M pretax income, and present very good results in load factor and benefit/seat. It isn´t a fantastic opportunity to buy or to much risks associatte with brexit and news extra cost coming?


    1. I think the chaos that would arise in the event of a "hard" Brexit would be too much for this company to cope with, David.

      In the absence of Brexit chaos this stock will return 10x in three years or so and it's (therefore) tempting to come up with reasons why chaotic brexit is an unlikely scenario -- bad for Spain, bad for London, etc etc.

      But the important thing in investing is to live to fight another day and so I think it important to wait and see what happens on the political front rather than to add more at this price. (Shorting Dart Group as a hedge is also an option)

      I'll write up some other interesting stocks over the next couple of months as well as a status report on some of the ones that I own

  79. Very insightfull, thanks very much. Hoping EU and England find the way to understand each other and think about citizens globally.

  80. Red - I have to say your Hard Brexit comment befuddled me slightly. Yes, hard brevet would lead to a sever recession which would be very harmful to Flybe but I don't see the issue as existential. Being a very UK concentrated company it won't be affected as much as Easyjet and Ryanair by disruptions to its routes - surely the only threat is to demand? Or are you referring to a potential further devaluation in GBP?

    That aside, a reasonable set of results, I thought - with things going in the right direction. If you think this is a 10x without hard brexit, mathematics would dictate it's worth some money even if you think it could go to zero

  81. Tim,

    I've uploaded a simple fundamental model here for ease of discussion

    I'm using "Hard Brexit" to mean a situation where Flybe's aircraft operating certificates are not recognized in the EU. That is the default setting and as it stands today Flybe would lose the right to operate flights between the UK and the EU. In that circumstance, it would lose at least 40% of its revenue base. That's not a situation that it could conceivably manage its way out of.

    Let me know if you need my help in deciphering the contents of the link

  82. The results were ok. As jet fuel goes up so do industry yields and Flybe's cost advantage is better accentuated. Absent a hard brexit scenario -- or if there is a three or more year transition period -- the company should do well.

    My prior comments were directed at anyone thinking of putting a substantial share -- or adding to an already substantial share -- of their funds into this security. Pre-referendum this was a true value investment and I wanted to underline that that is no longer the case (although it may become one again at some future date).

  83. Thanks very much for the detailed model, red. I may post some questions once I have had a better look. May I ask where you assembled the rather detailed data from?

    In the spirit of keeping everybody up-to-date on Brexit and its (possible) impliciations for UK aviation, here's another article from the FT from today which discusses a leaked EU document:

    I think the chances of the scenario you defined above as 'hard Brexit', i.e. no flights between the UK and EU, is, of course, possible; but I equally think it is very remote (I'd guesstimate ~1% and note that such a scenario would have negative consequences for both sides -- going far beyond aviation).

    I do acknowledge that capital/portfolio allocation must be computed taking this remote chance of a 'hard Brexit' scenario into account (which is fun given the high return in its absence! =).

    Thanks again for sharing, red. Greatly appreciated!

  84. Always happy to help with valuation 7 business model questions

    Airline data here:
    Airport data:
    Route frequencies derived from:
    Detailed financials (discontinued after 2015):

    E-jet performance:

    Aircraft ownership/lease values:

    1. Forgot to upload this which may save you time if you're willing to trust but verify my data entry skills

    2. This will days, if not weeks, to process, red... thank you!

      One short question:

      Why assume (?) 5% improvement in daily utilization for FY2021? Just curious...

      Finally, a video (! =)

    3. Well I suppose one intuitive way of approaching it is to trace the 6200 annual E195 flights that are reallocated elsewhere. 6200 flights currently flown by the 195s are inherited by the 175s. 6200 flights currently flown by the 175s are taken on by the Dash-8s. That gets you there more or less.

      Also, you can imagine that there are some routes -- Manchester to Amsterdam, for example -- that could benefit from having 2x 88 seats in the breakfast and tea time hours rather than 1x 118 seats.

      In the dead hours -- say 10AM to 4PM -- there are probably more opportunities for the Dash-8s that than there are for the E-jets.

      And there's more flexibility in the choice of holiday destinations in the summer and the dead of winter. E.g. Rather than forcing Cardiff-Milan frequencies at low-ish loads as currently, smaller aircraft allow for a greater # of choices wrt to airport pairs.


      So anyway, I don't think 5% improvement is a challenging target in a halfway normal operating environment. Probably a little low,actually, but at the current valuation I think 5% is enough to work with.

    4. ps I'm based in the US so iPlayer doesn't work for me. But it should help any UK-based visitors to this blog so thanks for posting the link.



    7. The Transport Committee transcript reflects surprisingly high confidence from airline execs that agreements will be made in time to allow continuity of flights between the EU and UK after Brexit. Honestly, they seem more worked up about the APD and passenger compensation rules than about Brexit.

      Red, thanks so much for sharing your work. It took some time to go through it. I hope you don't mind, but I have a few questions.

      RAW DATA
      1.) Route Frequencies tab - Where did you find the raw data for Seats and PAX for each route? I can see where you found the Flights data in the CAA punctuality statistics, but I can't see where you found the Seats or PAX numbers that show 40% of seats are on UK-EU flights.

      2.) Why do you assume the avg yield for Codeshare flights will jump from GBP 80 to 100?
      3.) Why do you assume the avg Hours/Flight for Q400s will increase by 3%? Is that because they'll be taking on some of the longer flights currently operated by E175s?
      4.) Why do you assume that the Daily Utilization of the 1 remaining E195 will more than double (from 4 to 9 hrs)?
      5.) How did you land on those particular uplifted PAX/Flight / Load Factor numbers? The model is obviously quite sensitive to the load factor for the Q400, so I'm curious about the +4 on that PAX/flight number in particular.

      Also, I think I may have found a small typo in cell C11 of FY 2021E... instead of basing the new E195 Ticket Yield on the historical value of 65, the model is using a value of 85 instead. Cell C13 has a similar error (using 75). It has a very small impact on the overall model results, but in the spirit of helpfulness I thought I'd let you know nonetheless.

      The model seems most sensitive to Q400 load factors and to the GBP/USD. Changing the currency rate from 1.35 to 1.20 drops Flybe's per share value from 3.20 to 2.0. Then decreasing Q400 load factors by ~4% takes share value down to 1.0. (Still a triple.) Looking at this sensitivity after reading the testimony of UK airline executives, I'm left feeling that the more probable Brexit risk to Flybe is a further devaluation of the pound accompanied by lower load factors, not a cessation of UK-EU flights. Would you agree with this?

    8. RAW DATA

      The domestic/international split is in the "airline data" link to the CAA source that I posted earlier. There's no good reason for you to have to reproduce it from scratch so I have uploaded it here


      2) The older codeshare arrangements were heavily weighted to Air France & Aer Lingus whereas the newer code share agreements include Air India, Virgiin Atlantic, Emirates, Etihad etc which I suppose to be more heavily weighted to longer-haul flights.

      3) Yes. And the aggregate data suggests that their strategy since 2014 has been to increase the average daily utilization of the entire fleet by flying further and therefore longer.

      4) The E195s are currently mostly engaged in subsidized flying that would otherwise be uneconomical -- e.g. the arrangement with Cardiff airport. When all but one of the E195s are gone it makes sense to me that the remaining aircraft will be used on the busiest/thickest routes and hours.

      5) Reallocate the PAX from E194s to E175s and from E175s to Q400s, take away a fudge factor, eliminate the worst routes, etc and one gets to +3 to +5 extra PAX per Q400 flight. The key question today is of course not whether it is +2 or +4 but whether it is better or worse than today or FY 2015. Setting aside the Bexit-EASA issue any argument that it could be worse in the future than it has been thus far in the post Jim French era would, I think, have to be an interesting and innovative one.

      re: C11 Not a typo. These are the thick business PAX heavy routes supplemented at the weekends with flights to sun destinations

      "I'm left feeling that the more probable Brexit risk to Flybe is a further devaluation of the pound accompanied by lower load factors, not a cessation of UK-EU flights."

      It is a trivial matter to hedge out the impact of a devaluation of the GBP so that is an expense rather than a risk.

      The risk is the EASA/certification issue. There's nothing that can be done about that but to size one's position accordingly.

      The transcript of the Transport Committee hearning is indeed cheery. As with everything Brexit there is an implicit assumption that there is a grown-up somewhere -- in the Mandarinate, in Lufthansa's executive suites, in Spain's Ministry of Tourism -- who will step in and somehow sort it all out. Maybe, but there's as yet no evidence that that is in fact so. The Treeza-Arlene farce today, for example, should give all right thinking people pause.



  86. This is a good primer

    1. Thanks again for sharing. Here's the page on aviation:

      Have you heard/know of the Chicago Convention mentioned in the linked page? Specifically, it says "Under these rules, UK airlines would only enjoy the first five freedoms at most." -- which would probably enough for Flybe to survive.

      Let's see what the next 3 months of transition/implementation period negotiations will bring... Fingers crossed! =)

  87. "Under these rules, UK airlines would only enjoy the first five freedoms at most."

    "AT MOST" is the only part of that sentence that matters.


    (The headline is misleading. It should read " Govt TO SEEK to stay in" EASA.)

  89. Seems like investors are indeed on the fence until future aviation rules are clarified. The bottom of the linked page shows that trading volume during DEC17 is ~10% the volume in JUN17 and 30% of the average 2017 monthly trades:

    Thanks again for all your posts. This blog is better than business school.
    Merry Christmas,

  90. Here's the sectoral report for aviation, published today, but (surprise, surprise!) without the interesting sector views (of the UK goverment):

    On capital allocation:

    I think that there is no investment without risk. In fact, it is even worse: not investing is a risk, too!

    Let's simply assume the risk of no planes flying between the UK and the EU27 is 5% (too high, from my point of view) and that Flybe shares are worth NIL in such a case (which is not clear, either). Otherwise, let's simply say a FLYBE share may quintuple from here (95% chance).

    How much of my total capital C0 should be allocated to this opportunity?

    How much should be allocated to it if we consider UAN, too, and that there may be a 10% chance nitrogen prices do not firm up over the next 3 years, but if they do, UAN share price may quadruple?

    I'd be happy to have your thoughts on this. Thanks!

  91. Have you heard of the Kelly Criterion - it is a formula that exists to decide the optimum amount to bet in just this sort of situation. Given your numbers, you should theoretically allocate 94% of your portfolio to Flybe :)
    In the real world - no one knows the pay off and probabilities, so I would just use it as a loose guide. I would personally put no more than 15-20% into any investment and that would have to have downside protection. Even if you are convinced Flybe is a fantastic investment, I would put no more than 10% in it.

  92. Anon: This is not the first 10x idea I've had and it won't be the last. Similarly, there's no reason for you to believe that it will be your first and last. Invest to survive rather than to make the most money in the short amount of time possible. There is no reason why you shouldn't wait until after EASA participation has been agreed before you make Flybe a full sized position. Will it 5x overnight? Not likely. So a safer 7x or 4x or 3x is likely a better risk-reward proposition than 10x potential today.

    You peg the lack of an aviation framework between the UK and EU as a 5% probablility event. Yes, it would be catastrophic for this government's electoral prospects when tens of thousands of jobs are lost and joe public won't be able to make it to Magaluf. And, yes, the fact is that any of the off-the-shelf options available to the UK -- from "Norway" to "Turkey" to "Canada" -- are consistent with continued EASA participation.

    However, there is a material risk that this government, in trying to bluff its way to a better overall deal than it can recieve, will continue to (1) miscalculate the timetable and/or (2) behave dishonourably with respect to its Phase I commitment on the north-South Irish cooperation.

    Therefore you are implicitly saying that it is 95% probable that the UKG's behavior will CHANGE and SOON ENOUGH. As I say, I think the cleverest thing to do is to build your position as milestones are achieved.

    1. Thanks Tim & red.

      I had heard of the Kelly criterion before, but am sceptical w.r.t. its widespread use as it only applies in very specific conditions. In addition, it does not really help to compare/weigh a number of different opportunities. In practice, I have thus never used/trusted it. Interestingly, Mohnish now sort of warns against it in his most recent Boston college talk. Personally, I am very tempted to try some value-at-risk variant. I know most concentrated (value) investors typically employ simpler processes (5%, 10%, 20%) and do not necessarily trust the maths, often rightly so.

      I do appreciate your comments, warnings perhaps, on position sizing and am happy to be reminded of a) how useful inversion can be, and b) that it is our long-term compounding rate that matters.

      I really hope HMG get their act together so that we all have a good start into the New Year.

      In great gratitude,

      red, the returns of your individual positions seem to sometimes be off. For FLYB, for instance, which was trading at 44p at 1 January 2017 and now at 31p and which you do not seem to have traded during the year, should show a return of 31/44 - 1 = 0.295, not -0.235, no?


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