In December 2017, I co-authored an article in Forbes about The "Free Lunch" Portfolio, a strategy to invest in 15 stocks across Uber Cannibals, Shamelessly Cloned Ideas and Spin-Offs. The portfolio rebalances its Cannibals in April and Cloned Ideas and Spin-Offs in December. I have been publishing updates to the portfolio here on Chai With Pabrai.
The Free Lunch portfolio was up just 3% in 2020, vs. 16% for the S&P 500. Although the portfolio fell 37% in the Covid-driven sell-off in Q1 2020, it has since recovered meaningfully. Since April 2020, the Free Lunch is up 64% vs. 45% for the S&P 500. Since inception on January 1, 2018, on an annualized basis, the Free Lunch portfolio is up a very modest 1% while the S&P is up 13%. The S&P 500 has trounced the Free Lunch portfolio so far. But, three years is also a very short duration. With bubblesque additions like Tesla and overheated valuations, I think the S&P 500 may have a few challenging years ahead.
Holding on to Compounders and Spawners
The Free Lunch’s underlying strategy is a sound one. It biases the portfolio towards businesses that generate (and return) a lot of cash and have wide moats. Over the years, the algorithm has picked up stakes in exceptional businesses, including NVR, Alibaba, Alphabet (Google), Micron Technology, Chipotle and Berkshire Hathaway. The algorithm is a great stock picker. However, it tends to sell businesses that are Compounders, Uber Cannibals, or Spawners way too early. This leaves money on the table and stunts the power of long-term compounding.
Take NVR as an example. NVR is an asset-light U.S. homebuilder that builds pre-sold homes. It has a low-risk business model that generates a ton of cash. At NVR, buying back shares is religion. For 25 years, management has used every penny of excess cash to buy back stock, and they have have bought back a mind-blowing 75%+ of their shares outstanding! Mr. Market has rewarded this Uber Cannibal handsomely as a result. An investment of $1,000 in NVR in 1995 would be worth $413,500 today. In January 2018, the Free Lunch picked NVR, which it promptly (and incorrectly) sold in the April 2018 rebalance.
The Free Lunch strategy is also very good at selecting Spawners, which are businesses that persistently add and incubate related and unrelated businesses. These have the potential to be massive growth engines. The Free Lunch has invested in exceptional Spawners over the years, including Alibaba, Alphabet and Berkshire Hathaway. But it has also sold too early. In January 2018, the Free Lunch picked Alibaba, which it (incorrectly) sold 12 months later.
New 2021 Free Lunch Portfolio
As a result of this introspection, we have made three key changes to the Free Lunch approach. First, although we will continue to leverage the algorithm as a stock picking tool in December and April of every year, we will not automatically follow its buy and sell decisions. We will only make changes if doing so improves the quality of the portfolio. Second, we have replaced the Spin-Offs bucket with Spawners, because they are superior businesses to buy and hold. Third, we will not be required to select 5 businesses from each of the three buckets. The number from each bucket will depend on the businesses that the algorithm offers. We want to invest in compounders, not buckets.
We will continue to evaluate the Spawners and Shamelessly Cloned Ideas in December of every year, and the Uber Cannibals in April.
Here are the constituents for the upcoming year:
If you are a new investor to the Free Lunch Portfolio, you can just equal weight these 15 stocks (i.e., invest the same amount of money in each of these 15) between now and early January 2021.
If you are already invested in the Free Lunch Portfolio, and you rebalanced the Uber Cannibals in April 2020 with the New Uber Cannibals, then sell all of the 2020 Spinoffs and Shamelessly Cloned businesses except Alphabet, Berkshire Hathaway, and Chipotle, and invest the proceeds equally among the 6 new kids. You can do this in early January 2021.
If you’re investing in a taxable account, you may try to sell the losers (Citigroup, Hilton Grand Vacations, Athene Holdings, RMR Group and Vectrus) in December 2020 to capture short-term losses and sell the winners (Fiat Chrysler and FirstService) in early January 2021 to capture long-term gains.
These are the stocks to sell, along with their full-year 2020 returns:
As a reminder, the new Uber Cannibals get published every April, while the new Spawners and Shameless Cloning businesses are released in December. When we publish the new Uber Cannibals in April 2021, sell the Ubers that are no longer on the new list and invest the proceeds equally across the new Uber Cannibal picks. Then in January 2022, you’ll rebalance the Spawners and Shameless Cloning ideas.
Note, anyone who invests in any strategy needs to do their own research/due diligence and are themselves fully responsible for the outcome.