Monday, October 19, 2015

Rolls Royce: Aircraft and Engine retirements

Roll's recent profit warning was due to the transition from old engines to new one.  Old planes are being retired, reducing maintenance income, while Rolls is still waiting to ramp up production in new engines.

I need to get some idea of the numbers of aircraft being retired in the future.  Is there more to come?

First, lets look out the different aircraft models based on range and payload:

Ignore Narrow body aircraft at the bottom left, which are irrelevant to Rolls.

Jumbos, in the top right, are large 4-engined planes.  They are now a niche product: for polar flights, freight, or busy slot-constrained airports.  They are only profitable when full, which is risky for airlines as they can't scale down.  Passengers also prefer greater frequency with smaller planes.  B747s are being phased out for passenger use, and the is A380 only successfully used by mid-east carriers.

Soonest to be replaced will be the B747s and A340s, due to their fuel-guzzling 4 engines.  Followed by the B757, now also out of production.  They will be replaced by the B787, A350, future B777X and future A330neo.

Its not a 1-1 replacement.  Larger aircraft are more profitable (when full), but more risky to operate.  For example: an A350 gives better per seat fuel milage, but may be harder to fill, so could be substituted by 1.x B787's, allowing the airline to scale down the route when necessary.

How long do planes last for?  The lifespan of a Wide-body averages 25 years.  It depends on the individual plane's model and milage - some last 35 years.  Freighters tend to last longer.

(Source: Avolon white papers 2015 and 2012.

To estimate the number of aircraft retiring, I searched the production list for all aircraft in the "Wide Body" and "Jumbo" categories above.  I only took currently 'Active' aircraft, ignoring 'stored' ones and did not distinguish between passenger or freight (or the occasional military).  Then I added 25 years to the each plane's first flight year, to estimate when it would retire.  The resulting graph gives the number of Rolls Royce engines due to retire in red and non-Rolls engines in green:

Again, it is number of engines, not planes.  And its only a rough estimate, as each plane could retire up to 10 years earlier or later.  So we can't predict the timing of any more earnings surprises in the future.

What we can tell is:
  • The numbers expected to retire are around 50 engines from 2014-2016, 150 engines from 2018-2022, and 150-200 engines from 2023-2026.
  • The predicted retirements match favourably with engines coming on line in the next few years.  Around 780 A350's are on order, but their rollout is constrained by Airbus' production rate.  Its currently 3 A350's per month (i.e.: 72 engines per year), expected to raise to 10 per month (240 engines per year) in 2018.  And maybe 13 planes/month (or 312 engines per year) after that.
  • Plus another 350+ B787 planes (700+ engines) from the current order book, lets say over the next 9 years.
Based on the order book and the projected number of engines retiring, Roll's problems of retiring engines are temporary.  In the long run, there is no evidence that the number of engines maintained will drop, and it will probably increase.

Bought Rolls Royce

Bought another 2073 RYCEY ADRs, at $10.4255.  Total cost USD 21,621.

Thats priced at roughly 13X earnings.  Rare to find inexpensive blue chips in year 6 of a bull market.  Now I have a full position in this counter.

Now around 85% in cash.

Monday, October 5, 2015

Bought USD

In late 2011, I sold my Perth property and converted the money to SGD, because:
  • Commodity booms are cyclical, and we were 9 years into one.
  • The property's price had risen 4 times in 14 years, (non-mining) wages had only risen 50-100%.
  • I was expecting a recession.
The last didn't happen, and I was 2 years early, but the decision proved correct:

Today I converted a significant amount of SGD to USD.  This decision is a lot less certain.

Holding SGD is a little bit of a risk (not a lot), as its lumped with Emerging markets.  The risk is if China devalues the Yuan significantly, due to capital flight and the unwinding of the carry trade puffed up since 2008 by endless rounds of QE, and they start to lose control (1) (2) (3).  But holding USD is a risk if the US economy falls or if we get QE4....QE for-ever....

I don't know what will happen - Nobody knows nuthin'.  But its too much of a risk to hold all my money in SGD, better to balance a little.  Maybe I'm right.  Or maybe I'm the last sucker to make this trade.

Now I'm holding:

In the big picture, I'm still holding cash waiting for the stocks to fall.  And still avoiding commodity currencies like the plague.  Not trying to make money yet, just trying not avoid too much risk.

Friday, August 28, 2015

Rolls Royce: Cashflows

A look at Rolls earnings and cashflows.  Same as for any company I want to buy. 

Earnings vs Cashflows

First, ignore the Net Financing.  These contain large profits or losses from their currency hedges (mark to market), irrelevant to their underlying business.  All the earnings below exclude this.

Lets see how past earnings and cashflows compare:

The stated earnings “smooth-out” the Cash Flow from Operations (CFO).  Sometimes they are higher, sometimes lower.  For the next two years, CFO is expected to be below earnings as large numbers of new Trent 1000 and XWB engines are built at a loss.  When this happens, revenue and profits from future maintenance and long term contracts are recognised at the initial point of sale and held on the balance sheet as assets under Accounts Receivables.  In detail:

All these assets are held under "Amounts Recoverable from Contract" under Receivables.  TotalCare Assets (in blue below) have risen to form the majority of these:

This is a risk, given given the uncertainty in estimating the revenue and costs of long term contracts.  Rolls gave an model example for a single contract (slides/transcript):

The resulting profit (and difference from cashflows) is modelled as:

The modelled profit/cashflows over the product's lifecycle (building many engines over the decades):

In Roll's 1H2015 presentation (p24), the cashflows for their newest engine show that they expect it may be a drain on cashflow for a few more years:

I don't think there is anything funny going on, but earnings will be below cashflows for the next few years, and I need to understand this if I'm going to hold the shares. As investors, we have no way of knowing how aggressive their accounting is.  Their accounting practices were reviewed by the FRC in 2014, where they agreed on the treatment of TotalCare.

The expected free cashflow for 2015 is between -150m to +150m (p17).  Essentially zero.  I'm guessing it may be negative for a few years after that.

There may be write offs risk if flight hours decrease suddenly (e.g.: SARS, financial crisis) or planes are grounded.  If that happened, I'd regret buying the shares before the write-off, instead of after.

Cashflow Generation

A look at Cash Flow from Operations and Cash Flow from Investments.  Major acquisitions/disposals and one-off events for CFI are annotated:

Both annotated acquisitions are for non-aviation businesses.  Without them, they would have generated cash every year.  Hopefully the new CEO will stop.

In the long run, ten years or so, if Rolls wants to develop a narrow-body engine, that would take considerable investment.


I like the company.  If the A350 is a success, earnings should trough in 2015, though free cashflow may be negative for a few years after.  Biggest risk is a recession, financial crisis or SARS/911 type of event, which could affect their revenues and may lead to write-offs.

Saturday, July 11, 2015

US Natural Gas (part 2) - valuing Advantage Oil and Gas (AAV)

Continuing my previous Nat Gas post.   Steps showing how I value Advantage Oil and Gas, a low-cost Canadian natural gas producer.  Since Natural gas prices are highly cyclical, I am trying to normalise them based on the futures market (see Market Based Forecasts here in a paper by Aswath Damodaran).  The futures market is best predictor of market prices as they know more than anyone else.

Heres the details:
  • What are AAV's realised prices compared to the benchmark?  The company gives their realised price and the AECO price (in CAD/mcf) in their quarterly results:

          Showing just the discount:

          Around 0-30 cents.  The discount to the Benchmark is pretty small, unlike in the Marcellus.  Sometime even negative!  Lets take 20c.  Though theres a risk that the discount may widen if too many companies start producing in the low cost Montney area.
  • Look at the Natural gas futures prices.
         Source - (as of 6-Jul-2016).  Great source for simple AECO and NYMEX prices.
         AECO price is in CAD/GJ, which is approximates very roughly to CAD/mcf (100 GJ = 94 mcf).  So multiply by (100/94) to get mcf price.
  • Look at AAV's projected production.  In 2017 they project 245 thousand mcf/day, which is an 1.83 times their 2015Q1 daily production.  From the above futures chart, CAD 3.07 (per GJ) is about CAD 3.26 per mcf.
  • Based on their 1Q2015 profits, project what their profits would be at different gas prices in 2017.  I am using a discount of 20c.  Usually I would go through the income statement and guess which of their costs are fixed (e.g.: General & Admin) and which are depend on production volumes (e.g.: operation expense, DD&A).  But for AAV it didn't make much difference (mostly variable costs), so I just multiplied their profits proportional to the expected production increase.  I get:

Highest I can get is CAD 3.90 at the 2019 price.  At USD 6 plus (~ CAD 7.80) , the stock is overvalued. My ballpark figures can't even get close.

[Edit: 4th Aug:
   Producing gas cheaply is not enough, you also have to deliver it to customers.  Canada has a small population of 35m.  Monty, in West Canada, has historically exported gas to East Canada and the NorthEast US.  Both are under threat by prolific US production there, especially in the Marcellus.  I think that Canada's Montney gas production will rely on LNG exports to Asia.
   Thats a big uncertainty. I don't know enough about the industry to buy gas producers in a new emerging area.]

Price/EBIDTA or Price/Cashflow

Most analysts I've seen use price/EBIDTA or Price to Cashflow (not sure which cashflow...) to value Oil and Gas E&P companies.  You get much lower multiples, because these denominators exclude Depletion, Depreciation and Amortisation (DD&A), which is the largest component of costs for most Oil and Gas companies.  For AAV's 1Q 2015 results for example, DD&A was around 2/3rd of their expenses:

It doesn't make sense to exclude DD&A from the valuation, and also exclude the Cash Flows for Investment needed to drill new wells to make up for the depletion.  Especially when shale wells deplete so rapidly.  For example, AAV's 2015 June investor presentation (p5) shows hyperbolic decline curves for wells drilled in different years:

I cant see why analysts use price EBIDTA or price/cashflow metrics.  This applies to all oil and gas E&Ps, not just AAV.  I just don't get it.

Thursday, July 9, 2015

Sold Kweichow Moutai

Sold my holdings, 900 shares @ CNY 244.  Profit ~ SGD 15.5K.

Why did I sell?  Its a toss up between:

  • Hold, because its a good company at reasonable valuations and good long term growth potential, and I've got nothing else to buy.
I don't think any bursting bubble ever has ever been successfully averted up by a government anywhere ever.  The only way they could do it - and its just possible in China - is to order the banks to print money and buy whatever price the market offers.  They will do whatever it takes.  See the nice chart here.

Hope that, after the carnage, I can buy back at a cheaper price later.


I'm now 91% in cash.  Buffet said it best - like an oversexed guy on a desert island.  Still nothing to buy.

Wednesday, July 1, 2015

US Natural Gas

US Natural gas prices have been low since 2008, and are now below the cost of production.  When this happens to a commodity, eventually either its price must go up, or the the cost of production must drop.  The usual value-investment strategy is to buy a lost cost producer which can outlast the competition and will benefit when the commodity price rises later.

In short: I think US Natural Gas prices will rise soon due to falls in production.  But I could not find a stock reasonably priced enough to bet on this.

Natural gas pricing

Since natural gas is difficult to transport and impossible to store, it price is set regionally, not nationally or globally.  US Nat Gas prices reference the Henry Hub (Nymex) price, and gas sold by companies in gas producing areas is sold at a regional discount, reflecting transport costs and the area's supply/demand (gas produced vs outgoing pipeline capacity).   The discount changes over time, for example, in the Marcellus, Cabot had a discount of 10c to 30c in 2013, and 89c in 2Q14.  The Marcellus differential for the last year seems to be around 80c to $1.40 - see the 2 NGI charts here.

One note for calculations: the Nymex price is in btu, but most companies provide their average realised price in mcf.  There's no way to convert between the two.

Production Costs

What is the cost of production, and how do we find the lowest cost producer?

Commodity companies always provide their own version of production costs (e.g.: C1 costs, cash costs, extraction costs, half-cycle costs) in presentations.  These exclude Depreciation, Deletion and Amortisation (DD&A),  General Administrative (SG&A), interest, and sometimes royalties.   Ignore all this rubbish and just use the expenses from the income statements instead.  Use the latest quarter, because the gas industry costs change so fast. Strip out hedging (derivitives), one off costs (e.g.: litigation) or irrelevant costs (transport of 3rd party gas).  Subtract Oil and liquids revenue from the gas production expenses based on whatever price they realised that quarter.  After this, we can get a breakeven cost of gas per mcf.

I checked the 2015Q1 income statements of 7 US/Canadian companies.  Only three were were making money unhedged:

This was period included the Q1 seasonal price spike - so I'm sure that overall production costs are higher than the market price.

Demand vs Supply

US production is around 70-75 bcf/day.

Only one area, the Marcellus has rising - almost exponential - production.  All other areas are flat or declining - See this nice 2013 map and graphs.  Although the Marcellus has the lowest cost of production, it also has the widest differential to Nymex pricing.  There are some indications that national production will peak or decline this year:
  • 2 articles (1) (2) by Bill Powers: the states' Department of Natural Resources and Texas Railroad commission data currently shows flat or falling output.  EIA projections of output increases are wrong.
  • Several companies have indicated flat or lower production in their 1Q 2015 earnings calls. Cabot stated they will reduce their Q2 production by ~10% sequentially, and "continue to monitor the price environment before we make any decisions on selling more gas into the local market".  Chesapeake stated that they shut in some production from December onwards, Q1 Marcellus production growth was "pretty well flat", and they intend to maintain this (albeit with the ability to quickly grow if prices rise).
Couldn't find a cost curve for natural gas.  There's too many players, and the production costs keep changing due to technological advances.

Shale gas wells have a high initial decline rate (See first graph in the first result here), like shale oil.  So we can't ignore DD&A - it represents money that needs to be ploughed back into drilling new wells as current ones decline.

On the demand side, natural gas is in a secular uptrend, due to the fact that its been cheap for so long, and as a lower-carbon replacement for coal.  Everyone knows it will will always be cheap.

Company valuations

The cheapest producers are Advantage, Cabot and Chesapeake.   Southwestern is also profitable, as it gets a higher realised price as most production is outside the Marcellus.

To value a company: I pick a what I think the long term average commodity price should be, project their earnings for that price, then apply a PE ratio to that.  I'm picking a NYMEX price of $4 (roughly $4/mcf on average) - just a guess since I have no cost curve.  I also pick a PE of 12.  Theres nothing magical about that number, but thats where I'd be comfortable buying and holding long-term through hell or high water.


Targets "close to 3 Bcf per day by end of '17".  At a realised price of $3.80/mc ($4 NYMEX and an optimistic 20c discount), I get an annual EPS of $1.80, giving a target price of $21.6.  The stock is now too expensive.


Targets 245 million mcf/day by end 2017.  At a realised price of CAD 4.50 (approximately USD 3.60), I get an annual EPS of CAD 0.75, giving a target price of CAD 9.00 (or USD 7.18).  May be worth investigating further.


Although they have a low cost of production, they also had a terribly low realised price ($1.61 vs Cabot's $2.23).  They gave no reason why: "This was primarily the result of weaker Marcellus Shale basis differentials in the Current Quarter compared to the Prior Quarter and increased gathering and transportation costs. "  Cabot also operates in the Marcellus.  The difference does not seem to be permanent: in 1Q 2014, their realised price was  $3.86, better than Cabot's $3.74.  In between then, they sold off a lot of gas producing assets.  Lets see if their realised prices improve first.