Coffee Can Portfolio – 5 Years On

Inspired by Chris Mayer's influential book, “100 Baggers”, the Coffee Can portfolio advocates for the selection of high quality companies, which are then entrusted to the passage of time, shielded from constant tinkering and reactionary market sentiments.

I was first introduced to the concept of the coffee can portfolio while reading the book “100 Baggers” by Chris Mayer. Out of all the investing books I’ve read, this one has had the most significant impact on me (stay tuned for a future blog post about it). Although there have been other more influential voices in my investing journey, the likes of Matt Joass and Andrew Page come to mind, this book sits at the top of my reading list.

The coffee can portfolio strategy is refreshingly simple: buy and hold. Resist the urge to tinker. Avoid selling after a quick 20% profit that makes you feel like a champion, and don’t be swayed by fear-inducing headlines that lead you to sell at a 10% loss. Heck don’t even sell when a company is down 70% and it seems unlikely to recover. As one of the first investment strategies I tested (and continue to test), it’s probably responsible for my habit of being extremely slow to sell, even for the positions outside of my dedicated Coffee Can portfolio.

Let’s dive back into the strategy itself. The book recounts the inspiration for a journal article from investment advisor Robert Kirby from 1984. Kirby describes how a couple invested over the same time frame and purchased all the same stocks yet achieved markedly different results. The wife followed all the buy and sell recommendations of the investment advisor, while the husband did his own thing. After the husband passed away, it was discovered that ‘his own thing’ was to buy all the recommended stocks but ignore the sell recommendations. Surprisingly, this unconventional approach outperformed the wife’s portfolio. By making the ‘smart’ decision to sell a stock based on fundamentals, valuation, fear of fraud, worsening macro conditions, or other reasons, it ended up costing the wife a substantial amount of money. The husband held onto some companies that eventually went bankrupt, but he also had several that doubled, and one company (Xerox) that became a 100-baggers. Kirby and, consequently, the wife missed out on the massive gains because they sold too early, missing out on potential returns of over 1000%.

This story inspired the concept of a coffee can portfolio. It suggests putting your stock certificates into a metaphorical coffee can and stashing it away for at least ten years. Take it out after a decade or more and see the results.

I read “100 Baggers” around 2018 and decided to create my own coffee can portfolio. I considered companies with the potential for long-term, high returns and committed not to touch them for at least ten years. The companies I chose were:

  • Amazon
  • Apple
  • Atlassian
  • Carvana
  • CrowdStrike
  • MongoDB
  • Square (now Block)
  • Twilio

There were other companies I would have loved to buy, but at the time, I was also renovating a house, so cash was limited.

generated by pstoedit version:3.44 from NVBadge_2D.eps

The pure coffee can strategy suggests equal weighting your starting positions, but I deviated from that. I heavily overweighted Amazon (which has impacted my returns). I didn’t purchase all these companies simultaneously either. For instance, Crowdstrike didn’t go public until June 2019, while most other companies were bought in late 2018. There was also a minor market correction in December 2018, providing an opportunity to average down on some positions. Nevertheless, my commitment at the time of each purchase remained the same: buy the stocks and put them in the “coffee can” under the mattress for at least ten years.

I was able to find some notes that I recorded ~2018. Here are some shorthand notes from that time:

Amazon (AMZN)

“Eating the world. AWS growth out-pacing other revenue line items and at higher margins. We seem to continually underestimate how much data and storage the world needs. Dad hates having to maintain the services at his school (solving a problem). AWS to be a key cog in a worldwide oligopoly.”

  • Purchased a bunch of time for an average split adjusted $88.182 ($0.712 AUD/USD).

Apple (AAPL)

Love their products and they have built an ecosystem which makes it hard to break away from them (switching costs – moat). They actually have an inferior product, yet the cult-like followers don’t even care (might actually be a long-term risk). Another stream of income – services – is growing like crazy and at a higher margin (kinda razor and blade style here).

  • Purchased twice (average down) for an average buy price of a split adjusted $48.188 ($0.7214 AUD/USD) and received 19 dividends.

Atlassian (TEAM)

“Hometown bias. Cuttla Aussie blokes taking it up to the big dawgs overseas. I’ll back ‘‘em in. Plus, they are gushing free cash flow.”

  • Buy price $72.50 ($0.7082 AUD/USD).

Carvana (CVNA)

[Paraphrasing a collection of notes] “I had recently purchased a second-hand car and the process sucked. I walked out of a dealership after four hours feeling grimy and agitated. Thought Carvana could be the answer and carve out a niche for themselves.” [They did… and lost money on every car they sold. You may call it idiocy, but I am sticking to the process. They might be a walking zombie or perhaps they surprise to the upside and the (many) shorts get squeezed thanks to the angry mob behind WSB? I’ll take a win any way I can get it with this company].

  • Average buy price $66.50 ($77.65 AUD/USD)

CrowdStrike (CRWD)

[A later entrant into the coffee can] “As more of the world moves online privacy and security will become increasingly important. Back the dominant player. [paraphrasing] I had a mate that worked for the company and having just read Jim Collins’ book “Good to Great” I asked him “what company or product do you talk about inside Crowdstrike as being the market leader / best product / fastest innovator / cultural target?” His response, “absolutely no one, they talk about us”. This friend was otherwise a very humble guy, so his response really stood out to me.”

  • Purchased three times. As low as $74 and as high as $220 for an average buy price of $104.26 ($0.6854 AUD/USD).

MongoDB (MDB)

“Seemingly novel open-source approach to data management. Not necessarily within my circle of competence but if the good Dr likes them [Dr Anirban Mahanti – Motley Fool analyst at the time, now with 7Investing] it is worth a closer look. The higher margin cloud offering, Atlas, is starting to make up a higher percentage of overall revenue. SaaS style revenue stream is hiding progress.”

  • Purchased twice (average down) for an average buy price of $151.28 ($0.6890 AUD/USD).

Nvidia (NVDA)

“Best in market and in ten years there will be more everyday devices with chips in them (leader in important sector with growing TAM). Can see a world where competitive computer gaming will take off in a huge way.” [It somewhat has but not to the extent I envisaged].

  • Purchased for a split adjusted $40.095 ($0.732 AUD/USD).

Square (SQ)

[paraphrasing a series of notes] “I wanted to buy Visa but thought it was too old-fashioned. [I’d like that decision again]. I was astonished to discover the limited penetration of digital finance across the US. At the time, I hadn’t used cash for five years in Australia, and if someone asked me to write them a check, I would have had to google how to do it. Yet, the US population still relied on cash transactions. I believed that Square, riding on the proliferation of iPhones, would digitize finances for the masses.”

  • Purchased twice (average down) for an average buy price of a split adjusted $88.46 ($0.7214 AUD/USD)

Twilio (TWLO)

I am almost prepared to put this in the same ‘permanent capital loss’ bucket that Carvana sits in. [to paraphrase] “I thought all interactions and communications would happen exclusively online. Therefore, any company or content creator that needed to communicate with their audience would require Twilio’s technology stack on their websites/apps. I believed Twilio would dominate this space given their head start.” [At least I was making money from currency exchange].

  • I averaged up on Twilio, resulting in an eye-watering average buy price of $320.46 ($0.723 AUD/USD). Bummer.

Portfolio Results

After five years, the returns are already resembling the story from Chris Mayer’s book. Carvana appears to be heading towards zero (from its previous lofty highs of $30b MC), while Nvidia is up ~900% since my initial purchase. The gains from Nvidia have more than compensated for the losses or sideways movement of other stocks Carvana and Twilio.

One of the obvious advantages of the long-term, coffee can, approach is that you only have to be generally right. It is much easier to develop a thesis and be generally right rather than specifically right, including right around the timing of your expected catalyst. Across most of my assumptions, I was in the right postcode and that has been enough to generate solid returns thus far.

Company Name

Average Buy Price

Current Price (10 Jun 23)

Capital Gain

Currency Gain

Dividend

Total Return (excl FX)

Total Return (pa)

Amazon

$88.182

$123.43

9.07% pa

1.17% pa

0%

39.97%

10.23% pa

Apple

$48.188

$180.96

64.65% pa

1.54% pa

1.84% pa

275.5%

68.03% pa

Atlassian

$72.50

$171.99

31.26% pa

1.09% pa

0%

137.2%

32.35% pa

Carvana

$66.50

$19.07

-20.68% pa

3.82% pa

0%

-71.3%

-16.86% pa

CrowdStrike

$104.26

$149.53

21.7%

5.92% pa

0%

43.4%

27.63% pa

MongoDB

$151.28

$374.51

39.7% pa

0.75% pa

0%

147.6%

40.45% pa

Nvidia

$40.095

$387.70

207.71$ pa

1.89% pa

0.42% pa

867%

210.02% pa

Square (Block)

$88.46

$64.94

-6.65% pa

6.37% pa

0%

-26.6%

-0.28% pa

Twilio

$320.46

$63.89

-26.68% pa

6.17% pa

0%

-80%

-20.51% pa

 

Chart showing Coffee Can Portfolio with initial weightings
Chart showing Coffee Can Portfolio with current weightings

Closing Thoughts

Whilst I deviated slightly from the purist coffee can approach, I have committed to letting time do its thing and am relying on the high-quality names in the portfolio to do the heavy lifting. The size of the Nvidia slice of pie in the above graph would make a lot of investors nervous. But, isn’t that the whole point of the coffee can approach? Let your winners keep winning. It is mid-2023 and Nvidia’s stock prices likely well over its skis but the (simple) original thesis still holds and so I am confident to let it play out for at least another five years.

There is significant volatility amongst some of the names. Apple has had two 30% corrections within the last twelve months and the round trip Nvidia has done over the 18-months to July 2023 is nauseating. Yet it hasn’t impacted me emotionally at all. Why? Because I have committed to holding for at least ten years. Given the volatility, I can see a need to time the exit slightly if/when I start to unwind the portfolio. Even the best passive portfolios have to be actively managed at some point.

Overall, I am huge fan of the coffee can portfolio as it leverages the importance of time in the compound interest equation and attempts to counter the constant Chicken-Little narratives. Whilst the half-way performance is excellent on this portfolio if I was to repeat the process I would be far more deliberate with my company selection. Insider Ownership, high ROIC and consistent revenue growth matched with margin expansion are the earlier indicators I anticipate focusing on. But more on that later…

Leave a Reply

Your email address will not be published. Required fields are marked *