Friday, July 16, 2021

Why I want to buy Gold

 Gold may have stopped falling:

In the coming years, I expect periods of high inflation like in the 40's or 70's.  Gold serves as a decent inflation hedge, even if its not a great one:

In the coming quarters, I expect a lower q-on-q growth rate.  Gold does well when growth slows.  One caveat is that even through the growth rate should be slower, GDP growth itself should still be high (mid-to-hight single digits).

In the next few months I think there's a good chance of a market correction.  When market exuberance hits slowing growth.  Think it'll be a small correction like Dec 2018, not a crash like 2020.  Gold hedged against past market corrections well.  I put the chances of a correction this quarter around 50:50.

Gold is the best bet for me now.  Copper may be a better inflation hedge long term, but does badly with slowing growth.  Gold miners will go up more than gold over several quarters, but won't hedge against a correction.  So just hold the shiny yellow stuff paper.

I bought a small 2% position nought during last month's shakeout.  Right now its hard to buy more, its always overbought.  Got a feeling it won't be oversold for a while.

Lots of reasons to buy it.  No reasons not to.  This is my only high conviction trading/investment idea now.

Thursday, June 17, 2021

Where are we in the Palm Oil price cycle?

There are a lot of listed CPO producers in Malaysia, Singapore and Indonesia.

The CPO price is really volatile:

The latest peak is probably caused covid-19 supply disruptions, and is probably temporary.

Where are we in the commodity price cycle?


I don't think we can predict demand in the next few years.  There's too many moving parts:
We are better off looking at supply.


Palm Oil can only be grown at plus or minus 10 degrees from the equator.  In 2018, Indonesia supplied 56%, by far the most.  Malaysia supplied 28% with the remaining 16% from other Asian, African or South American countries.

The long term factors affecting supply are the number of hectares planted and the age of the trees.  The trees start producing 30 months after being planted, and reach peak yield at 7-13 years.  Production slowly declines since then:

Source: Palm Oil World

This December 2020 CPOPC report explains supply well:

Other figures corroborate this: Malaysian Palm Oil production has been flat for the past 5 years, while Indonesia's has been rising:

Source: Indexmundi (MsiaIndon)

In summary:
  • Malaysian CPO production has been flat, while Indonesia's growth has slowed (though is still currently growing enough to meet demand).  
  • If both Malaysia and Indonesia catch-up in replanting new trees, it would temporarily remove around 2.6m tonnes (estimated 3.5% of world production).  I guess it would take 4-5 years for production to recover from this, after which it starts exceeding today's numbers.
  • El-Nina has unpredictable short term effects on prices.  It may increase or decrease palm oil production, and affect soybean production (substitute for palm oil) in Latin America.
I would say that 2019/20 may have been the bottom of the cycle, but its not clear.  We did not see the type of demand destruction typically associated with the bottom of commodity markets.  But we did see some curtailing of growth.

If we do get a CPO boom, expect supply to catch up with demand in 3-6 years time.

Tuesday, June 15, 2021

Bought Uranium Stocks

Been looking for a chance to buy Uranium producers.  For the inevitable uranium bull market.  Was waiting for a market correction, but this news ("Chinese nuclear plant 'performance issue' reported by its French joint operator") hammered Uranium stocks last night.

Decided to take the plunge and buy a 3% position.  For something this risky, its gotta be a small enough position so that it doesn't hurt too much if it halves.   Otherwise I can't stick with it.  This is a "buy and forget" trade.  I want to wake up a few years from now and find we're in a Uranium bubble.

Theres only 2 publicly listed Uranium companies currently producing:

  • Kazatomprom - the world's lowest cost producer in Kazakhstan (LSE:KAP).  Bought 364 GDRs @ USD 31 each. 
  • Cameco - a Canadian mid-cost producer (TSE:CCO).  Bought 570 shares @ 23.83 each

Its risky.  Short term, the knife can keep falling, even if the fears turns out to be false.  After rising so far, these stocks may need a few more more days to shake out weak hands.  Long term, they can go up 3 or 4X if a real Uranium bull market occurs, like in 2007.  They can also drop by 2/3rd if the worst turns out to be true and we get another Fukushima.

(Edit July 2021): Realised Cameco is a mistake, sold it off at a small loss.  Its a barely profitable company, not one that I want to hold when I may have to wait years for a recovery. Sold it off at a tiny loss.

Saturday, May 22, 2021

Does Paper Gold hedge against stock market corrections?

Gold is an unprintable zero-yield currency.  So it should do well when real interest rates fall, or when they are expected to fall.  Like in a stock market correction.  Lets check the behaviour of GLD in past corrections.

Since GLD's history only goes back to 2004, we start in 2008.  The chart below covers the the 17-month bear market from Oct 2008 to Mar 2009.  S&P 500 (the purple line) was down 55%, while GLD (the candlesticks) was up 40%.

It wasn't smooth. GLD had a vicious 7-month 1/3rd drawdown, followed by a swift 4-month 50% recovery, bringing it back to where it started the drawdown. 

Why did it fall?  Interest rates were falling throughout this period, but there was a pause from May to Sept 08, corresponding to GLD's fall:

Source: MacroTrends Federal Funds Rate - 62 year historical chart

Flat rates would have stopped gold rising, but why did it fall?  Probably due to the panic, with funds needing to liquidate anything they could.  No one knows, but the important thing is that it did.  How would you buy-and-hold or trade through this?

Next we have a series of smaller corrections.

  • A brief two-month 16% S&P 500 correction from May-June in 2010, shown by the dotted rectangle below.  GLD up ~3%.  Its too short a time for the chart to mean anything.  And this was within a wider multi-year gold bull market.
  • A more substantial 5-month 19% S&P 500 correction from May-Sept in 2011.  GLD was up 4%.  We can see GLD's big spike as the stock market corrects, but GLD's final red candle has it falling 9% in a week.  Excluding this fall, GLD was up 13%. 
  • Again, we can see how hard it is to trade or buy-and-hold.
  • After the above spike, gold entered a 7 year bear market.  We only get one stock market correction in this time, 3-month correction from Nov 2015 to Jan 2016.  GLD exploded when the market dived:
  • The 3 month 2018 bear market from Sept to Dec 2018.  GLD was up 3-4%, and a lot less volatile than the S&P 500 which was down 20%:

  • With hindsight, the above was the beginning of a new bull market in gold.  Which we may still be in now.
  • GLD fell with everything else in the 2020 crash.  It was down 7% vs the S&P's 33%.   With a maximum drawdown of 8%, it held up ok.


On average, GLD has risen during small corrections.  But not during large bear markets or liquidity crises.

I'm expecting a short correction in the second half of this year.  Like 2018 or 2015-16.  I'm thinking of going long gold.  The odds are on your side if you buy gold before a (brief) downturn.  But the charts above give some idea of the risks involved.  2015-2016 in particular shows how dangerous it can be to be early.

Friday, May 21, 2021

Selling My Trading Account

I'm now selling my trading account which made up around 15% of my assets:

I'm selling now because:

  • Hedgeye sees quad 4 next quarter.
  • Money inflows from the Treasury General Account will stop.
  • The market is too frothy.  Every man and his dog is in.  Shitcoins are flying as crypto will "change to world".  Headlines proclaim "Copper is the new oil"...where were all these guys in 2020?

Think I've got another month to sell.  Then I look to buy gold to benefit from falling growth.  Don't know how long that trade would last.

Longer term, I think the correction will be short, like Dec 2018.  I want to be in the market again.  Still in commodities for higher inflation.  I'm looking for new targets, plus sticking with the old ones:

  • Copper again, for electrification.
  • Aluminium, copper's only substitute.  It also increases energy efficiency (lightweight).
  • Oil.  Rising energy prices are the definition of inflation.  
  • Fertiliser, rising crop prices mean more planting for next year's harvest.
  • Maybe palm oil, it takes 4 years for newly planted trees to bear fruit.
  • Battery metals are tough, they are prone to substitution.
  • Met Coal.  Probably US based, to support infrastructure spending and re-indistrialization there.  Australian listed coal producers have too much thermal exposure.
  • Maybe evil thermal coal?  Would favour Indonesian producers, as they have a strong domestic market (unlike the US or Oz) and are best placed to export to Asia & India.
  • Uranium is hard to place bets on.  The small caps are losing money, and Cameco is too tame.
  • Crypto.  I like ETH, but they are losing market share from high fees and a clogged network.
Basically anything that can't be grown too fast or printed.

Wednesday, April 21, 2021

Sold ADP and AENA

 ADP has been showing signs of distribution in the past month.  Falling prices on higher volume:

It could be due to India's covid situation.  But when reading charts: "Ours is not to wonder why".  

Normally I could ignore little wobbles in a company's share price.  But I think the bull market is coming to an end.

Source: Hedgeye, Jonesy's twitter

For AENA, it was a close decision on whether to let this stock go or hold it long term for dividends.  Its got a decent yield, monopoly position, and could survive a Eurozone breakup.  But I have enough dividends to live off, and it doesn't have enough growth potential - with a 5% yield while paying out 100% of their 2019 earnings.  I need more dry powder for the next cycle.

Sold 300 shares of ADP @ 101.10.  Profit SGD 8.8K, or 22% over 6 months.  And 210 shares of AENA @ 134.25.  Profit ~ SGD 5.1K, or 12%.

As a "covid recovery" trade, they've been disappointing.

Friday, March 5, 2021

Fibra Macquarie

Mexico has good demographics and its low-cost manufacturing should benefit from the China trade war and NAFTA (h/t Peter Zeihan).  So I've been digging around for Mexican Industrial REITS.

Basics on Mexico

The only things I know about Mexico are from TV.

So let's starts at the beginning.

Inflation: Mexico has had bouts of high inflation.  

Although its now under control, the steady inflation rate of 3-5% is higher than developed countries.  Since everyone in Mexico has experienced inflation, most rental leases are either tied to the CPI or in USD.  So peso inflation is not directly an issue for investors.  But when you see a peso denominated rate-of-return (interest rate or cap rate), mentally deduct 3-5% from it to account for inflation.  Mexico's official interest rate is 4%.

Property rights: foreigners can own freehold property, except perhaps up to 50km from the coastline or 100km from the border.

Crime: the drug war started in the 90's and has worsened since the Cartels splintered.  The level of violence is like something out of a movie and is unimaginable by people living in Asia/Europe.  Sometimes powerful gangs are welcomed as law enforcement.  The government does not control some areas of the country.

Fibra Macquarie

Properties,  Tenants and Leases

Fibra Macquarie has around 230 industrial and retail properties throughout Mexico:

Mostly industrial.  In 2020, 82% of the revenue and 88% of Net Operating Income (NOI) was from Industrial.  From the above map, around 1/3 to 1/2 of the industrial properties (by GLA) are near the US border.  They don't have much in Mexico city. 

83% of their NOI is denominated in USD.  98% of industrial leases are triple net (p13).  Over 70% of their industrial tenants are light manufacturing (p18).  38% are from automotive (p10).  This is a play on NAFTA, not e-commerce.

Fibra Macquarie did not list their properties or tenants in their annual report.  Their website has a browse properties page, but only shows vacant ones.  I probably can't evaluate 200+ Mexican industrial properties anyway.  Theres some talk of their properties being older than other REITs - the average age of its industrial building is said to be 15 years (p70).

The scant information I have on their tenants is from their 2019 Annual Report: the largest industrial tenant accounted for 3.9% of industrial rent.  The top 10 largest tenants contributed 25.8% of industrial rent.  Most REITs list their top 10 tenants proudly, but its strange I couldn't find any here.  What are they hiding?

WALE is short at 3.5 years.  Management stated that for leases expiring next year, current rent and market rent is similar, and they are not expecting any change (40:10).

Management and Incentives

The REIT is managed by an external party (Macquarie México Real Estate Management), who owns 4.8% of the units.  They get paid (p273 in Adobe Acrobat):

  • 1% of the REIT's market cap, annually.  
  • Plus an additional bonus of 10% of the amount of "total return based on market cap" (over a 5% hurdle plus inflation rate), payable every two years.
The first is OK, but not great.  The fee is reasonably cheap.  I thought it may encourage reckless issuing of units to fund dilutive expansion.  But this has not happened - they have instead bought back units over the past 5 years.

The second is OK, since it based on market cap.  As I read it: excluding issuances and buybacks of units and dividends.  Its reasonably aligned with shareholders.

As a plus, they are not paid any extra for making new leases, construction, acquisitions or disposals.

Overall, their administration fees are reasonable, aligned with shareholders, and easy to understand.

Mexican Fibras generally have high management fees and complex/distorted incentives:

Cashflows, FFO and AFFO

FFO is usually lower than CFO:

To get AFFO, they subtract future (non-cash) costs  from FFO.  Around 14-20%.  Mostly for property maintenance or improvements, with a little for leasing commissions and platform costs. Straight line rent is negligible:

The terms are defined on slide 39 here:

Source: 4Q2020 Supplementary Information, slide 39

In summary, they are generating cash, and provisioning 14-20% of their FFO.

Pipeline and Growth

Their last big property purchase was announced in 2016.  Since then, growth has come from expansions and redevelopment (p27):

The 11.6% cap rate sounds really high.  Maybe we should mentally deduct 3-5% from it, I'm not sure.

Management stated that they expect to be able to add 1 to 1.5m GLA a year, funded via retained FFO, new loans and opportunistic asset sales (34:50).  Thats an increase in GLA of 2.5 to 4% a year.

Seems like they continue with their slow, low-risk growth strategy of "extend and redevelop".


Net LTV is 36.4%.  Their Regulatory LTV is slightly lower at 35%.4% (p35).  The maximum is 50%.

All their loans are fixed (p36).  Some use properties as collateral: 30% of their properties are encumbered.    97% of their loans are USD.  The interest rate is pretty high, averaging over 5%.

A lot of their loans expire in 2024 (p34), loan expiry is a bit too concentrated:

All their loans are interest only, but they have being slowly paying off debt over the past 5 years.  

Management anticipates debt may rise in future to fund the development pipeline.  Short term, LTV may rise into the low 40s. (34:00).


A big tail risk is a sudden USD depreciation.  71.8% of their leases are denominated in US Dollars.  If the USD was to drop 30-40%, lets say in 2-3 years, their revenue would follow.  Their costs will be the same, except for interest, whoch is the largest cost.  If the USD dropped by 40%, I model that their 2020 operating profit before tax would drop by 40% too.  The company would survive.  The short WALE works in their favour, in this case.

Other risks are:
  • I do not know the quality of the properties or tenants
  • I do not know the property cycles in their local markets.
  • It takes a year or two to build an industrial property, so there is no moat in this business.
  • Mexican political risk.  The drug war.
  • 3D printing disrupting manufacturing.  EVs disrupting automotive.


2020 AFFO was 2.59 pesos per certificate.  Projected 2021 AFFO is 2.27 to 2.32 pesos per certificate.  So at its a price of 26 pesos, its trading around 10 times earnings (AFFO), with a 7% yield (80% AFFO payout ratio).


  • Excellent financials, and cheap.
  • Biggest long term risk is Mexico, especially the drug war. Historically, holding Mexican stocks and pesos has been a losing game.  Maybe this will change.  Like China in 2001.
  • Biggest short term risk is the lease expiries, and I know nothing about their properties or local property cycles.
In the end I bought some, its 5% of my portfolio.


Mexico's Withholding Tax is 10%.   It should be zero for Singapore residents (p7), I'll need to check if Interactive Brokers applies this later.

Other links:

  • Detailed pre-covid 2019 report on Mexican Fibras by BTG Pactual.
  • Short 2014 write up in Value Investors Club (free registration required)