Wednesday, April 8, 2020


Defli is an Indonesian chocolate manufacturer and distributor listed on SGX.  75% of their revenue is from Indonesia, the rest from the Philippines.

The Industry and Competitors

An Oct 2019 DBS report covers this well:

  • Delfi's distribution is strong in rural areas ("General Trade"):
  • Distributors in General Trade are largely exclusive to Delfi, creating a moat.
  • The Modern Trade channel has been growing faster than General Trade, but has recently been slowed due to unhappiness.  New modern outlets can only be set up a certain distance from existing traditional outlets.
  • Delfi has the largest market share in Indonesia.  They mainly compete with Mayora in the value segment, while the global brands (Nestle, Cadbury, Mars) compete in the premium segment.
  • Indonesia has laws protecting local manufacturers: "Under the Ministry of Trade (MoT) Regulation amendment No. 56/2014 for traditional markets, shopping malls, and modern stores and outlets, at least 80% of sales must come from domestic products, and private labels may account for a maximum of 15% of SKUs."
  • Delfi's market share is around 10% in the Philippines.
This article says that Petra and Mayora have a combined market share of 80%.


Before starting, when looking at historical numbers, note that the company sold its cocoa ingredients division in 2013, but some effects can still be seen on the income statement up to 2015.  Where the results of both divisions - two very different businesses - are merged together.

Around 60-65% of revenue is COGS.  Selling and Administrative costs make up another 20%.  Leaving pre-tax operating margins at 8-9%.

So COGS has the largest impact on profits.  DBS Vickers estimates that cocoa/sugar/milk make up 30%/15%/15% of total costs respectively (p6).  These commodity prices were very low in 2015/16.  Gross margin has varied from 32 to 38.5%:

The other determinant of gross profit margin is the proportion of own-brand products sold:

Clean balance sheet with zero net debt and minimal operating leases (p6).

CFO swings around wildly, due to working capital.

CFO is usually more than CFI.

Inventories and receivables have been steadily rising, which is a bit of a worry:

  • In the 2018 and 2019, management stated that increased receivables are because of a greater proportion of sales to Modern Trade customers, who have longer settlement terms.
  • In 2018, the group changed to the direct shipment model, where they took over the distribution function form their regional distributors, to their Modern Trade customer.  This caused an increase in inventories
  • The spike in 2017 inventories was caused by "a Government-imposed transportation disruption which necessitated the deferment of some deliveries to January 2018."


If I buy this, I am paying SGD (or USD) to buy an asset in Rupiah (IDR).  And the company is operationally leveraged to IDR/USD exchange rate (product prices in IDR, costs in USD).

In the current part of the economic cycle, where there is a USD shortage, EMs like Indonesia are particularly vulnerable.  May be better to wait for a crisis first.


Good company starting to turnaround.  Good financials, has growth potential after the crisis is over, and seems to have a moat.  Valuation is reasonable, but not cheap yet.

With so many moving parts affecting their profitability (Indonesian consumer demand, exchange rates, raw material prices, proportion of own-brand products), its hard to know how to value it.  The stock price has been falling for 6 long years.  Would I be willing to catch a falling knife?  Maybe at a single digit PE ratio.  6 to 8 times earnings would be SGD 40 to 50c.  Without currency fluctuations.

A better way is to wait till this crisis is over, and EMs start to be back in favour again.  Indonesia is a dirt poor, politically unstable country and we don't know how bad things can get.  You need a different strategy when buying in a place like that. Wait till things are getting less worse - there will still be a lot of time to make money and ride the trend up.  Don't catch a falling knife in a rioting country.


Starting this year, they no longer report quarterly results.  Next results are for June 2020.

They paid a dividend of SGD 4.66c in 2019 (3.76c in 2018), around 50% and 75% payout ratios respectively.  This is important, as a way of rewarding shareholders so that the share does not become a "value trap".  I could not find a dividend policy.

Monday, April 6, 2020

As an asset light online travel agency (OTA), is a good way to play the coming rebound in travel stocks.  It has the strongest business model of all the OTAs.  Main question is when to buy.

Business and Strategy

This is a well covered stock, other people can explain it better than me.

glowing presentation from Sean Stannard-Stockton (youtube link) (starts at 5:30). Key points below:

  • 89% of revenue is from 87% of revenue from hotels and alternative accommodation. 20% revenue comes from alternative accommodation (but more corporate managed, not individual homes like airbnb).  
  • Most revenue is from outside the US.
  • Booking's business is "selling travellers to hotels".
  • Global hotel chains (eg: IHG, Marriott) already have a direct relationship with customers (eg rewards points).  Boutique hotels don't have this, so they pay a higher proportion of revenue to OTAs.
  • Europe/Asia have a lower percentage of hotel chains than the US:

  • Even though branded hotels are growing faster, it will be a long time before they significantly rebalance the market, especially in Europe.  Most of's business is in Europe.
  • Google recently added google hotel search.  This greatly impacts OTAs that reply on organic (ie: free) search results.  Estimate that less of Booking's earnings come from organic search:
  • Over 50% of the's transactions come from people who went directly to the site.  The company have been saying this for 18 months.
  • Booking is about 1 year behind airbnb in the alternative accommodation segment.

Balance Sheet

As of Dec 2019:

  • 6.3bn cash
  • 5.4bn short term liabilities.  This includes ~1bn of debt.
  • 7.6bn LT debt, plus 0.5bn operating lease liabilities.  1bn of the LT debt is due in Sept 2021.  After that, March 2022.  These are convertible notes, assume no one will convert.
  • A long term US tax liability of 1bn which we assume they don't have to pay now.
  • Interest expenses were only 266m.
  • Also have a 2bn revolving credit facility, only 5m used
How long can they last without revenue?  Last year, largest expenses were advertising 5bn, personnel 2.2bn, sales 1bn, G&A and IT 1bn.  Plus interest 266m.  So they should easily get to 3bn cash operating costs.  Thats 9 months until their credit runs out.  I am 80% sure they can outlast the crisis.  Need to see their 1Q20 results to see how fast they can reduce costs.

This guy thinks it can last a year.

Competitors and Industry Dynamics

Direct competitors are CTrip and Expedia.  A good comparison of Expedia and Booking is here.  Booking leads in Europe.

Hotel chains (eg: Hyatt, IHG, Marriott) are both customers and potential competitors.

The Google Monster is a supplier and the most threatening potential competitor.  Ben Thomson covered it well, he believes that google operates as an aggregator (advertising and recommendation platform), providing a frictionless service with zero marginal costs.  They would not want to do an OTA's messy job in the real world, such as signing up new hotels, bookings and cancellations.  This guy makes the same point looking at Google's and Booking's staff numbers.  I agree, though Google may want a larger share of the pie.

Airbnb is the largest threat, moving into boutique hotels and B&Bs.  Airbnb has listed 24000 hotels (as of Mar 2018 - less than 1% of their total listings) - they said they will only handle boutique hotels, not chains.  Booking and Airbnb have continually argued over who has the larger alternative accommodation listing, though by the numbers they seem equal.

Interesting March 2018 podcast on competition in the industry between hotels, OTAs and Airbnb:

  • (at 5:45) Airbnb so far does not compete with hotels.  Their original plan was hotels would come under pressure on high occupancy nights (eg: convention in town).  But this did not seem to happen for corporate travellers.  A hotel's convenience and consistency do not compete with the unique experience of an airbnb.  Airbnb plus is trying to move into this space.
  • (at 10:00) Airbnb's soft branding is the offer of a distribution system and loyalty program to boutique hotels, but the properties don't have to change their unique identities or rebrand.  Smaller independent hotels may prefer it.  Airbnb charges 3-5%, OTAs can charge 15%.  This is a threat to Booking, Expedia, and possibly hotel chains.
  • (at 13:08) Airbnb vs OTAs: Airbnb wants to be a 'superbrand' offering everything for holidays.  Expedia already does this, Booking is transitioning to one.  All of Booking's accommodation is instantly bookable, and it does not charge guests a fee.  Airbnb has better user interface and brand.
  • (at 17:15) Booking has the best business model and the most to spend on search engine and advertising, compared to Expedia and Airbnb.  Possible acquisitions by Amazon or Google may change the landscape.

Other articles:

Its a fluid industry with murky boundaries.  Customers are competitors.  We do not know who the survivors will be.  Based on sales, profitability and market share, Booking is the winner so far.

Misc is well followed, every hedge fund manager and his dog owns it.

Interesting article on Booking's past growth in Europe.

Wednesday, April 1, 2020

Other stocks I've looked at

I looked at these stocks but rejected them.  Keep this list, in case the market drops for another few months.

  • Malaysia Airport Holdings Berhad: They are changing their regulatory model - right now - due end March.  Previously they were only responsible for operational capex, now they will be responsible for all capex.  A massive capex increase is expected, from 200-300m per year, to 2-3bn (p28).  Since the business model is changing, the we can't look at past financials.  At least need to see the new regulations first, but it would be difficult for me to judge.
  • Sydney Airport: Simple business, far too leveraged and expensive.   At AUD 5.30, its at 30X earnings, even excluding D&A, its at 20X.  Debt is around 10 or 11bn, against 1.1bn CFO.  Forget it.
  • Chr. Hansen: Fantastic business, they make bacteria used for yogurt/cheese, with a 70/50% global market share.  Still too expensive.  Growth has been slow in the past few years, would need to drop more than 50% to get a price of DKK 170 to be at 15X earnings.
  • InterContinental Exchange (ICE): Stock/commodity exchanges, with 50% operating margin.  Still too expensive, wait till it hits $36 to $60 (trailing PE of 12 to 20).
  • Segro REIT: Fast growing industrial REIT, mostly UK, some Europe.  Still expensive on an EPS basis, would need it to drop to at least 50% to GBX 370 at 15X earnings before looking.

Maybe worth buying now:

  • Berkshire Hathaway.  Good description of BH's business model by ValueInvestAsia.

Theres still a lot of expensive stocks out there. Markets are not cheap yet.

Monday, March 30, 2020


Safran is a great business whose share price has been cut in half.

Its "Fly-by-the-hour" and maintenance revenues make it partially a proxy for air travel.


Safran is in the Aerospace industry, their 2019 segments are:

  • Propulsion.  Building (OE), maintenance (MRO) and spare parts for jet engines.  Almost half of company revenue.  56% of their propulsion revenue is from services (maintenance and parts), the rest is from OE.
  • Aircraft equipment, defence and aerosystems: 37% of company revenue.  1/3rd of this segment's revenue is from maintenance.
  • Aircraft interiors: 9% of company revenue.  One quarter of this segment's revenue is from maintenance.  From the 2018 Zodiac acquisition.

The crown jewel is jet engines.  In the narrow body market, their CFM-56 engine (co-owned 50/50 with GE) is the world's best selling aircraft engine, on the back of secular LCC growth.  They look to continue this dominance as they transition the old CFM engines to the new LEAP, which will work on the Airbus, Boeing and Comac planes.

Their only competitors in narrow body are P&W, and possibly a Chinese supplier for Comac.

2/3rds of 2018's Propulsion revenue was from CFM. So around 1/3rd of company revenue.  Could not find a profit breakdown.

In the wide body market, Safran has been involved the GE90 through its 23.7% stake.  This revenue will fall, as it is replaced by the GE9X, in which they only have an 11% stake.

In the mid range market, they have an 8% stake in the GENX engine, used by the B787 Dreamliner.  This revenue will probably increase in future, as B787 sales continue.

They target different margins for the operating segments:

Contract Liabilities

This is the by far the company's largest liability.  What are they?

Prior to 2018, "Fly-by-the-hour" revenue was recorded when paid, and lumpy costs (eg: 5 year overhaul) were artificially smoothed out:

Source: Safran IFRS 15 workshop (p6) 

For 2018 onwards, with the adoption of IFRS 15, revenue is adjusted to follow the real costs:

Payments in advance (in excess of revenue) are recorded as contract liabilities, against the new cash (asset).  When the lumpy costs occur, they are charged, and the corresponding contract liabilities  are reduced (along with the cash used to pay for them).

Financials and Liquidity

How long can they last in the current crisis?  Without external funding.

Their balance sheet is passable.  Against an adjusted (p15) PBT of 3.7bn, net debt is around 4.6bn.  The main liability is 10.4bn of contract liabilities, we don't know how many years later they are due.

For the income statement:
  • 25bn revenue, 21b costs, 4bn profit.  A 14% operating margin.
  • Of the 21bn: 14bn of the costs are variable (6bn raw materials, 5bn subcontracting, and 2bn external services).  1.6bn is D&A - ignore it.   6bn are personnel costs, take them as fixed.  Interest cost is neglible.
Last week they announced:
  • Net debt is has now decreased to 3.2bn
  • Dividend is cancelled
  • Cash is 3.1bn, of which 2.8bn is accessible within 90 days
  • They have an undrawn 2.5bn revolving credit facility till end 2022.
  • 2.8bn of debt is due this year, they are arranging a new 3bn credit line up to 2 years.  My guess is that this can be done, unless banks run out of money (everyone takes loans at once).  This is not a credit crisis like 2008.
In the worst case, I think this can last them at least 5 months.  Its a blind guess, no one knows how much things can choke up, from the airline customers to Boeing and GE, all the way down the supply chain to the company who makes the rivets.

Their finances are not as conservative as I'd like, a company with so many accrued contract liabilities should have more cash.  Still, leverage is low, and they are not as indebted as the typical US company.  A bit hard for me to make a decision, for three reasons:
  • There's a possibility the covid crisis lasts more than 5 month.  Will be a while before people book holidays again.
  • Many airlines will go bankrupt.  Demand for planes drops off a cliff.
  • There may be supply issues along a complex supply chain.  If any one part fails (one company under lockdown), the whole aircraft is unfinished.  Its not like a Big Mac, where you can still make one without pickles or sesame seeds.
I am undecided.  They can probably make it through, but not sure.  I'm sure that if they ran into problems the French government (or even private investors) would finance them, but I would want to buy shares after it happens.


Longer term:


Doing it based on their peak (pre virus) 2018/2019 earnings for cashflows.

  • On earnings, a price of EUR 90 is 22X 2018 earnings, and 15X 2019 earnings.
  • On company calculated free cashflows (p24), a price of EUR 90 is 22X 2018 FCF and 20X 2019 FCF.
I would probably buy somewhere between 12X and 15X peak FCF.  Around EUR 49 to 68.


Great company.  Not sure if I would buy at current valuations.  They can probably survive a covid slowdown/lockdown without external funding, but not 100% sure.

I think the bear market is not over yet - one month is too short.  Later on, the stock price may be lower, and we may have a better idea about how well they can survive.


When reading company materials, they use the word "recurring" for any operating income.  This includes OE, it does not just mean services.

They don't seem to release detailed quarterly results, which will make it harder to track their financial position in a few months.  Track their press releases instead.

2018 had substantial contract liabilities added (p87), which will make the CFO seem better than it really is.  Not sure about 2019, as the detailed cashflow statement is not release yet.

Sunday, March 29, 2020

Grupo Aeroportuario Centro Norte

Its worthwhile to look for travel related companies with high leverage hammered by the coronavirus. Ones that can survive.

This idea came from Lyn Alden's free newsletter.

Grupo Aeroportuario Centro Norte (OMA) operates 13 international airports in Mexico's central and northern regions till Nov 2048 (28 years from now).  1/3rd of their revenue comes from one airport in the wealthy industrial city of Monterrey.

65% of their revenue is from aeronautical services (passenger and aircraft charges), 20% from non-aeronautical (eg: leasing, hotel, industrial park, baggage services, F&B, and industrial park), 15% from construction (due to capex).  The first 2 are dependent on passenger traffic.


This is a regulated business.  Every 5 years, the airport operator submits a plan (MDP), which projects the traffic loads.  The maximum allowed rates charges for aeronautical services are determined by this.  If the rates are too high for one year (ie: because the operator underestimated traffic), then the government can 'claw back' the excess profits in the next year by lowering rates.  So upside is limited.  Conversely, the company may request adjustment on maximum rates due to natural disaster or recession (5% GDP decrease in 12 month period).

For non-aeronautical revenue, a substantial amount is royalty based (with a minimum and maximum bound).

The next MDP is to be submitted in 2021.


Balance Sheet looks very clean:
  • Total (not net) debt of 4.5m pesos is less that last year's operating income (before interest/tax).  
  • Apart from that, they have cash of about 3.4bn pesos.
  • No operating liabilities on the balance sheet (IFRS 16 adopted in 2019).
  • Only 2% of their debt is USD denominated, the rest are in pesos (as is their revenue).
Income statement looks straightforward:

Construction costs are the costs of construction (building improvements on the concessions' lands); these are added in revenues then deducted in operating costs (netted out). 

How long can they survive with zero revenue?

  • The first expense line "cost of services", seems fixed (page F-69), so keep it.
  • Remove concession taxes, technical assistance fee, and D&A.  
  • Keep construction costs (part of 5 year plan), and interest income/costs
We get around 2.5bn pesos.  With 3.4bn pesos cash, they should be able to last at least a year with no revenue!

Capital Expenditure

At the end of 2018, they projected 2.1bn pesos capex in 2019, and a smaller 0.5bn in 2020  (p122):

The 2bn is actually to make up for shortfalls in the 2016-2018 period.

However, at the end of 2019, they only invested 1.3bn (p1).  Giving a shortfall of 0.8bn in 2019, and expected 0.6bn in 2020; total 1.3bn for next year if they stick to the plan.  If they do stick to it, then they can only last 10 months without any revenue.  Still not too bad.

Capex for the next five year plan is expected to be 14-15bn pesos.


"In the future, we may face competition from Aeropuerto del Norte, an airport near Monterrey operated by a third party pursuant to a concession. Historically, Aeropuerto del Norte has been used solely for general aviation operations. The state of Nuevo León has requested in the past that the Ministry of Communications and Transportation amend Aeropuerto del Norte’s concession to allow it to serve commercial aviation operations. To date, the Ministry of Communications and Transportation has not amended Aeropuerto del Norte’s concession"

A positive risk is that the concessions may be renewed for another 50 years, though under new conditions.


At 80 MXN per share on the BMV exchange, thats a PE of just under 10 (pre virus).


Mexico has a 10% withholding tax on dividends.  Not sure if this is waived for Singapore residents (Article 10).

Friday, March 27, 2020

Sold Boustead. Waiting and looking for things to buy

Sold Boustead a few days ago at 59c.  Its a low beta stock, so convert it into cash which can buy something that goes up more.

The only stocks I have now are my dividend portfolio, and IAG which has fallen too much to sell (I would consider buying it now if I didn't own it).

I have SGD 250K invested, and 500K in cash.

I think the bear market is not over yet.  I am following Hedgeye for market timing.  They called market crash before it happened.

For the dividend portfolio, I will continue to slowly buy dividend stocks if they drop further. Targets are FCT and MCT.  My trades are recorded in InvestingNote.

For the rest, I am building up a list of stocks to buy.  I'll post them here on this blog, for the liquid ones.  I will buy when Hedgeye calls a turn.  This may be one or 2 months after the bottom - following this strategy means I may end up buying after some big up days.  Can't expect to get the bottom.

Tuesday, March 24, 2020

Interactive Brokers


IBKR is a fast growing international stockbroker.  They make most of their money from interest (2/3rds of revenue in 2019), and the rest from executing trades (1/3rd).  Trading commissions were only 7% of PBT in 2019 - that was an exceptionally bad year for trading - the average from 2015 to 2018 was 23%.  They also do some market making, which is negligible.

Institutional customers make up half their accounts and 65% of their total customer equity.   Over 60% of their accounts are form outside the Americas.

Competitive Position

I don't know who the players are for institutional investors, but assume its competitive.  For US retail investors, a market share estimate does not even include IBKR as its such a small player.  So numbers wise, they have a small share in a highly competitive market.

Their main product, IBKR pro, is for sophisticated traders/investors.  The user interface is complex and hard to use.  But they are one of the few retail brokers that allows you to trade a wide range of markets, currencies and instruments around the world.  IMO, they are not really a direct competitor to the other US brokers, but more occupy a high performance niche.

Despite being low cost, operating margins are amazing:
With these margins, they must have some competitive advantage.

Balance Sheet

Looks good.  Debt and operating lease liabilities almost zero.  2.8bn of cash acts as a buffer around 27bn of cash and securities in client accounts (Financial instruments are not included in cash, as they are mostly options - p86).


Like a bank or finance company, lending money is inherently a risky business.  One where the risks often don't appear on the balance sheet or cashflow statements.

Stockbroking risk is shorter term, as you mark-to-market every day.  But you have to deal with wild price swings:
  • In 2015, when the Swiss unpegged their currency, IBKR lost 137m (down to 115m in 2019 after some debt collection).
  • In 2019, there is an unusual loss of 42m (possibly up to 50m) for margin to "a small number of brokerage" customers who positioned in a US stock that had a drastic change in price (TSLA, maybe?)
These losses look OK in light of their cash and profit numbers.  A normal part of doing business.

However, financing businesses are a black box.  In every crisis a few go bust, and its hard to tell from the financials which ones.  If I buy, size this position accordingly.


Main growth driver is new accounts, which has been in the teens (number of accounts), or twenties (customer equity):

Source: 2019 Goldman Sachs Conference

IBKR is also a platform used by other stockbrokers ("introducing brokers").  The new Chinese broker Tiger, which they own a stake in, performs its trades through IBKR.


Most profits are from interest:

Short term profitability is affected by:
  • Interest Rates.  Lower rates mean "lower net interest income from customer deposits due to interest rate spread compression".  Management stated that 2019 profits would be 22m lower with a 0.25% decrease in rates.  I guess that with rates at zero, interest will be 130m to 200m lower.
  • Market volatility.  "Historically, our profits have been principally a function of transaction volume and price volatility... rather than the direction of price movement."  More volatility means more trades.  2019 was a bad year due to low volatility.  Now the VIX is over 60....yeah, they probably didn't want that much volatility...but lets assume trading revenue goes up to 2018 levels, which is around 200m higher.


This may be a Buffet buy-and-hold company.  No debt, minimum capital required, may have a sustainable competitive advantage.  What price would I pay for it?

  • For a company growing 15-20%, I'd catch a falling knife at 12X earnings. Thats $25.
  • In a more normal market, maybe 12-15X earnings would be fair.  Thats up to $31.

Alternatively, since the majority of earnings are from interest, wait till we see signs of interest rates rising.  Right now, we probably get a recession, then one or two quarters of growth (off a low base)... then we start to expect interest rates to rise.


Glowing report on IBKR from SumZero.

IBKR used to have a larger market making operation, but have reduced it due to HFT.