I’ve started an experiment where I’m keeping a public portfolio of large cap stocks via the Motley Fool CAPS system- you can view it here. I would like to see if I can generate value on the long side… something that I find much, much more difficult than shorting.
My criteria are:
- Pick only the quality businesses in a particular sector. For example, Dollarama is the clear leader among discounters.
- Not overpriced. For that reason, Amazon and Netflix don’t make the cut (just look at what happened to Amazon during the Dot-Com Bubble).
- I avoid industries where there are no clear industry superstars. Oil and mining stocks simply don’t make the cut as none of them are quality businesses.
- Lastly, I avoid dying or shrinking industries. Profits ultimately don’t grow in dying industries and therefore those stocks almost never do well.
Empiricism in investing
The problem in investing is that it’s incredibly difficult to tell whether or not your process is correct. Having a large enough sample size is one problem, so that’s why I’m trying to keep at least 30 stocks in the model portfolio. I am sticking to large cap stocks so that the results are more reproducible (no illiquidity issues) and so that it’s easier to track- the CAPS website will have more accurate performance data for large cap stocks (as data errors exist in financial data). Nonetheless, there will still be several problems.
Timeframe: It may take several years for the model portfolio to have reasonably reliable results. Events such as stock market bubbles can cause investing strategies to temporarily outperform or underperform for years.
Reflexivity: The stock market changes as market participants learn what works, leading to some strategies no longer working. The classic deep-value strategy of buying low P/E stocks and low P/B stocks hasn’t worked in the past decade.
Unintentional macro investing / sector speculation: I avoid industries where I can’t see any standout stocks, so the model portfolio has zero utilities and commodity stocks. The portfolio is also overweight healthcare insurers and technology stocks.
Predicting dying industries can be subjective: The portfolio doesn’t have TJX or ROST because I don’t understand whether that part of retail will eventually die off like department stores.
Valuation can be subjective: The interpretation of the accounting affects valuation, as accounting shenanigans and distortions created by GAAP accounting factor into the correct accounting. Also, the distinction between overvalued and reasonably valued is highly subjective. (Perhaps I should have specified some type of cutoff such as PEG, but that would take detailed valuation work.)
Results so far
The fastest way to estimate the portfolio‘s result is to take the CAPS Score and divide by 35, which results in 2.6% out-performance relative to the S&P 500. The first trade was on October 5 of last year so the performance of 2.6% has been for a timeframe of slightly over a year. However, it’s too soon to tell if there is anything here.
My attempt at backtesting suggests that my general strategy would have outperformed the S&P 500 by roughly four percent in the past. So, the portfolio is currently lagging expectations.
The backtesting portfolio was based on picking stocks that I would have thought were quality stocks in 2006, e.g. MSFT, INTC, AAPL, ETFC, MO, LUV, WFC, FITB, XOM, EBAY, XLNX, and others. There is hindsight bias so don’t take it too seriously. However, I did try to pick financials as the financial sector did poorly in the 2008/2009 crash. Unfortunately, picking the quality investment banks meant that the portfolio did not have Bear Stearns, Lehman, Citigroup, etc. The backtest portfolio included Etrade, which performed quite badly due to its investments in subprime mortgages. However, its inclusion has little effect on the portfolio as bankrupt stocks actually don’t have much effect on the overall portfolio’s performance. Portfolio returns are dominated by a handful of stocks that skyrocket, e.g. AAPL.
Comments on why particular stocks did and didn’t make the portfolio
Notable tech sector omissions- INTC, MSFT, TSM, NVDA
(This section is mainly an update on my post about what happens when technology is good enough.)
Intel and Microsoft are clearly the #1 player in their respective industries. However, I would argue that their industries are shrinking. Microsoft’s key cash cows are Windows and Office. Soon after those segments saw shrinking revenues, Microsoft re-organized their divisions and no longer provides much data on how these products are doing. Given that the personal computer market has been slowly shrinking since 2011, I doubt that Microsoft will see growth going forward. The story with Intel is similar, as sales of desktops and laptops have gone down. While the sale of smartphones and tablets have gone up, Intel has very little market share in that market due to terrible management. If Intel had followed the contract fab model like TSMC and its ilk, it would have likely dominated the smartphone industry. Unfortunately, Intel management was obsessed with manufacturing exclusively its own designs and didn’t want to license its technology to other players. It bought Infineon’s wireless business in 2011, which ultimately turned out to be a disaster as that division couldn’t produce a working wireless design in a timely fashion. So Intel’s smartphone SoC designs (which were relying on a functional wireless component from Infineon) weren’t very good. As well, Intel didn’t want to manufacture Nvidia’s Tegra chips for tablets so Intel’s excess capacity sat idle. Embarrassed, Intel gave up on the mobile market entirely. They still don’t want to manufacture chips for Nvidia. (I am aware that both Intel and Microsoft stock have risen dramatically in the past few years, but I think that the market has been wrong.)
NVidia is a somewhat different story. It has maintained impressive earnings growth, although I could never figure out how much of it was from the cryptocurrency bubble driving up the prices of GPUs. Its core business, video graphics cards, may slowly die over time as consumers lose some of their appetite for the latest and greatest in computer graphics. While video games will someday catch up to 2009 Hollywood movie Avatar, faster graphics chips haven’t been translating as effectively into stunning graphics that wows audiences into buying the latest and greatest graphics card. I don’t know if there core business would still be growing without the crypto bubble.
TSMC (NYSE:TSM) is hard to figure out. It has thrived because of the rise of the smartphone market and Intel’s incompetence. (Intel, which was already an ARM customer, could have manufactured the ARM-based cellphone chips that TSCMC currently makes.) However, TSMC will likely face the same headwinds as the personal computing market. What happened in personal computer will likely happen to smartphones- consumers will be satisfied with the speed/performance of previous generations’ technology and will instead start buying based on price, causing average selling prices to go down as the cost of production comes down. TSMC’s growth has been buoyed by crypto (see pg 199 of Bitmain’s financials for a breakdown of TSMC’s largest crypto customer) so it isn’t growing as fast as it appears.
One of the stock market’s hottest growth stocks is trading at a much lower multiple than shrinking companies like Microsoft- FB has a P/E of 22 versus Microsoft’s 49. Facebook has grown revenues at 42.7% over the past 5 years while operating income has grown 80.9%. I don’t get it.
Facebook has no serious competition on the horizon as they own Instagram (the hot new social media platform that continues to gain popularity) while competitors like Google+ and Vine are shutting down.
Stock exchanges and commodity exchanges- CME
While they used to have great moats, the stock exchange business has slowly been getting more and more competitive. Whereas stock trading was previously dominated by NYSE, AMEX, and Nasdaq, then number of competitors has increased dramatically. Interactive Brokers gives its customers the choice of 16 different venues such as BATS, the DirectEdge exchanges (bought out by BATS), and IEX. The list of competitors gets even bigger when you include all ECNs, dark pools (most major investment banks have their own dark pool), and off-exchange venues where most retail order flow is handed to market makers. I don’t like the increasing competition as competition is not good for profits. The commodity exchanges on the other hand don’t have this problem. For commodities with a physical settlement option, competitors cannot create an identical competing product.
Verisign’s monopoly on .com domains is a wonderful business- as long as ICANN continues to give them that monopoly. However, I couldn’t figure out the surrounding politics in regards to ICANN and the US government. For that reason, I could not figure out how to value this stock. Its future cash flows depend heavily on the terms at which their .com monopoly is renewed. In an ideal world, ICANN would put the contract out to bid and the price charged would drop. One article by The Register suggests that the price could have dropped to $2 rather than the $6 per domain that Verisign was charging at the time. Realistically, the cynic in me thinks that Verisign successfully influences ICANN to once again provide it with a lucrative monopoly.
Healthcare and pharma – AMGN, GILD, ANTM, CNC, ESRX, HUM, UNH
Among pharma stocks, which is not a sector that I understand well, Amgen and Gilead seem to have a better track record than their peers.
Miscellaneous technology companies – IPGP, CDW, BKNG, GOOGL
IPGP is a laser manufacturer that spends about 43.5% of its operating expenses on R&D. I don’t know if it’s the correct analogy, but I would compare IPGP to Intel. They are in a commodity industry where scale is a huge advantage. IPG photonic’s R&D spending clearly benefits from scale.
CDW is a computer hardware/software reseller and not really a technology company. They are a high-touch reseller that will provide IT advice to its customers so that they only buy IT systems that work. If there are incompatibility issues, the customer can return incompatible software/hardware. Customers pay a premium for CDW so that they don’t have the pain of buying IT systems that don’t work (see Reddit for some discussions on CDW). They can be thought of as a sales organization and an IT services company.
Priceline / Booking Holdings and Google / Alphabet are well-known companies so I will not cover them here. For more information on Google, see my old post about online advertising.
Semiconductors – TXN, SWKS
Whereas Intel makes digital semiconductors, Texas Instruments (TXN) makes analog semiconductors and is the largest company in its field. Unlike Intel, TXN’s has a broad product portfolio as its products do many different things. Moore’s Law doesn’t apply to analog semiconductors so product lines tend to have very long lives rather than just a few years. Something about Texas Instrument’s size has allowed it to retain its leading market share and maintain high profitability for a very long period of time, although I must admit that I don’t understand the industry very well.
Skyworks Solutions (SWKS) is also technically in the semiconductors industry, but they don’t compete with Texas Instruments. They design various technology products and sell them. They do not have an obvious moat but have a pretty good financial track record.
Insurance – PGR, BRK.B
Progressive is one of the leading auto insurers, alongside Berkshire Hathaway’s GEICO. Berkshire Hathaway, as you probably know, is a collection of fairly high quality businesses.
Brands – CLX, MNST, KHC, HRL, MO, TSN
Established brands tend to retain their market leading positions while generating higher profits than competitors.
Altria (MO) is particularly interesting. Let’s go on a digression for a second and talk about the Nifty Fifty, a basket of 50 quality companies similar to what my model portfolio is trying to do. The Nifty Fifty traded at very high valuations in the 70s, with an average P/E ratio of 41.9 versus the S&P 500’s 18.9. Not surprisingly, this basket of overvalued stocks underperformed the S&P 500 over the next decade. Nonetheless, compounding trumped overvaluation if you were willing to wait over three decades. Out of the Nifty Fifty, Altria performed the best (its predecessor split into Altria and Philip Morris). It turns out that a tobacco stock would go on to become one of the greatest growth stocks of all time. That being said, please understand that I have no idea what I’m doing when it comes to these types of stocks. Altria continues to grow its earnings despite selling fewer cigarettes (due to less smoking in the US). That trend does not seem sustainable.
Retail – HD, COST (with omissions of TJX and ROST)
Home Depot and Costco are retailers that are fairly well-managed, grow their earnings with high returns on capital, and are unlikely to see much disruption from e-commerce. I could not figure out if TJX and Rost Stores will someday see an inflection point from Internet retailing hurting their business. Most of the items being sold at those particular off-price discounters (e.g. apparel) are items that people also buy online.
Moats – SIRI, V, MA, MCO, SPGI, LBRDA/CHTR
Sirius XM has a monopoly in satellite radio, although their product does compete with free terrestrial radio and streaming audio. Its monopoly business continues to grow. LSXMA is another way of owning Sirius XM.
Visa and Mastercard have very strong positions in the payments industry as their network effects have been very difficult to overcome. Their peers such as American Express and Discover (DFS) haven’t shared the same financial success.
Moody’s and S&P Global are two of the big three ratings agencies. The oligopoly of the three agencies has been difficult to break, even by scrappy upstarts like KBRA. Institutional Investor has a great article by Julie Segal on the difficulties faced by newcomers like KBRA.
Charter Cable (which you can own via LBRDA or GLIBA) typically competes in duopoly markets, where customers can purchase services (phone, video, Internet) from a cable company or a telephone company. A bet on Charter is a bet on Thomas Rutledge’s ability to execute well and squeeze more profitability out of each coaxial cable heading into a consumer’s home. Otherwise the stock should be considered overvalued if you don’t think that its operations become more profitable.
Auto parts – LKQ
LKQ distributes auto parts to businesses, although their junk car lots do sell salvage parts to consumers. Their industry has its own quirks and needs. A commercial car repair shop prefers to have parts arrive quickly so that their workers do not have to wait around for parts to arrive so that they can complete a repair job. They also like to check inventory and place orders online as it is faster for their workers than using the phone.
LKQ focuses heavily on “alternative parts”, which refer to parts other than official (original equipment or OE/OEM) parts that can be sourced from a car’s manufacturer. Alternative parts come from aftermarket parts (unofficial replacement parts), used parts that have been remanufactured or refurbished, or working parts pulled from wrecked cars.
As far as competitive advantages go, management tries to maneuver the company into a position where their scale can be an advantage. Scale can benefit the customer as having a broad inventory means that more customer parts requests are fulfilled and in a timelier fashion. There are also some minor economies of scale that would improve profitability. Historically, LKQ has grown revenues and operating income at around 15-20% per year.
Homebuilders – NVR
NVR has consistently posted the highest margins while taking less risk on land price fluctuations than its peer. It is the best managed homebuilding stock (see this old post on shorting homebuilders as the post goes through homebuilding accounting issues).
Thor Industries makes recreational vehicles (RVs) and has a long track record of compounding. If case you don’t know what RVs are (they are called motorhomes in Europe), they are mobile homes that can be driven around. Many people use them for vacations as they travel across the country.
Thor has rolled up the industry to become the largest company in its industry, followed by Forest River (acquired by Berkshire Hathaway in 2005). Another publicly-traded company, Winnebago Industries (WGO) is a distant fourth-place competitor. Thor’s Q3 2018 presentation shows the market share of all industry players:
Here are the respective share price performances of the publicly-traded companies:
Please note that the chart above excludes dividends, so THO’s performance has been quite understated. Also note that the Forest River acquisition was in 2005 and likely had very little effect on Berkshire Hathaway’s performance. Regardless, it is obvious that Thor has outperformed its only pure-play and publicly-traded competitor Winnebago.
Currently, THO trades at a P/E ratio of around 8.3. This is only partially explained by the industry’s cyclicality:
Thor managed to eke out small profits at the market bottom in 2008 and 2009. Nonetheless, the industry suffered several lean years and has since rebounded dramatically. Overall, Thor seems undervalued given its low valuation and history of compounding capital for shareholders.
We’ll see how this portfolio does. Significant out performance will likely come down to whether or not the portfolio contains superstar stocks like the Intels, Microsofts, and Wal-Marts of the world. My hopes are on Facebook, Google, Charter / LBRDA, Sirius XM / LSXMA, NVR, the ratings agencies, and the health insurers. Some of the stocks in the portfolio seem far less likely to turn into superstars (due to relatively much lower historical growth) and that may be a potential mistake on my part.
*Disclosure: Of the stocks mentioned in this post, I own THO, CDW, CNC, FB, GOOG/GOOGL, IPGP, LBRDA, LSXMA, SPGI, UNH, NVR, BKNG.