U.S. Equity Investor Letter for the First Half of 2023
By Haruki Toyama, Head of Large and Mid Cap Equity, Portfolio Manager
We are halfway through 2023, and so far the U.S. stock market has continued its rebound from last year’s decline. The S&P 500 Index remains about 5% below its peak reached in January of 2022, so it hasn’t quite yet recovered all of its lost value, but as usual, the market averages don’t always tell the whole story. As well-chronicled in the financial press, the market breadth during this advance has been narrow. The top ten companies in the index by market cap accounted for over three-quarters of the index gain in the first half of this year. The phrase “narrow” implies a bit of tsk tsk, as if we should be admonishing the market that it’s not behaving well. In reality, and for some years now, we believe that the top-heavy nature of the general market averages has simply reflected the sea change that has been going on in the global economy for the past two decades, which is that technological and communication advances have allowed for unprecedented scale, global reach, and growth rates. Thus, when we hear observers comment that every one of the stocks of these behemoths is overvalued on its face because “no company of that size has ever grown at such high rates,” we wince a bit.
Such doomsayers have been peddling the same incredulity for many years. When Apple stock first approached $1 trillion in market capitalization five years ago, the financial press was replete with commentary saying no company should be worth that much, as if there was some ironclad law of physics that dictated such a constraint. Today, Apple is hovering around a $3 trillion equity market value, and that’s after the company has spent approximately half a trillion dollars on dividends and stock buybacks in the intervening period. It reported $100 billion in after-tax profits last year, double what it reported five years ago. We don’t own the stock and profess no strong view on whether it’s currently undervalued or overvalued, but Apple’s corporate performance has been nothing but stellar, and is a prima facie refutation that giant corporations can’t grow at robust rates, despite prior evidence to the contrary. Historical precedent is a good rough guide to future results, but by no means an infallible one. As Warren Buffett likes to say, if past history was all there was to the game, the richest people in the world would be librarians.
We don’t want to sound like apologists for unconstrained enthusiasm or expensive valuations. We hope we are neither. What we’re trying to say is that whether a stock is overvalued or not is independent from notions such as “market breadth” or historical precedent or any other secondary signal. Sometimes it’s difficult to detach oneself from these anchors, since human psychology is such that we tend to operate with heuristics, and not first principles. Our goal is to find the rare investments that are exceptions to the rule; by definition, we can’t do that if we don’t look beyond the heuristic.
In our portfolios, we own two of these behemoths, Alphabet and Amazon. In both cases, we believe that their target markets are so large that despite their current revenue run rates of $300 billion and $550 billion, respectively, they have as much growth ahead of them as many of our other portfolio companies with significantly less revenue. Thus, no special size discount need be applied; we value them as we would any other company.
Alphabet, in particular, is an interesting case study; it’s the holding company for Google and other technology businesses, and is currently the largest holding in our Large Cap strategy after multiple additions to our stake over the past year and strong price appreciation in the first half of 2023. We first invested in it in 2007, just a few years after its IPO. At the time, it was an unusual investment for a “value” investor like us; many of our value investor brethren shunned technology stocks, especially high-profile ones like Alphabet. (1) However, we believed we understood the power of Google’s search capabilities and the moat conferred by its continuous improvement model, and therefore, we were more than comfortable making what was then a fairly contrarian investment for the kind of investor that had scrupulously sidestepped the late 1990s tech bubble and flourished in those post-bubble years.
We then sold our stake several years later when we thought the valuation had gotten too extreme. But regretting our error, in 2014, we invested once again and have held it since. In the 2014 purchase, the prevailing sentiment was not dissimilar to the Apple example noted above: it was too big and mature a company to grow at anything but a plodding rate and thus deserved a low multiple. At the time, Alphabet reported $66 billion in revenue and $14 billion in net income. That net income figure was greater than the revenue of over a third of our portfolio companies we then owned. In the decade since that purchase, its sales and profits have increased almost fivefold.
Google’s moat in search is wider than it was when we first invested. It’s a business that benefits from scale: more customer searches equal more data, more data equals better results, and better results translate to more customer searches. Accordingly, the company’s lead in search volume has only increased during our holding period, and as expected, the superiority of its search results compared to its competition has increased as well. However, due to the advent of artificial intelligence (“AI”), particularly generative AI, as exemplified by ChatGPT, it’s possible that the moat is a little narrower than it was a year ago.
There are probably a certain minority of searches conducted on Google that make more sense as a generative AI query, but because of the lack of alternatives, are done as a search. As generative AI tools become widely available, some of those searches may end up moving over to generative AI providers, either resulting in a loss of usage for Google, or, if remaining within Google (through its AI tool, Bard), a reduction in monetization. In either case, the impact is likely to be relatively small, as the kind of searches that make logical sense to be generative AI queries are not the kind that is being monetized well anyway. Thus, in one of our quarterly letters in April, we expressed our skepticism that Google would lose any meaningful market share to Microsoft’s ChatGPT-infused Bing. So far, that skepticism has been well-placed, as recent data shows that after a small initial spike in usage for Bing, its volumes have fallen back to pre-ChatGPT levels.
PRINCIPLE VERSUS RULES
“Strength without flexibility is rigidity. Flexibility without strength is instability.”
– A wise yoga teacher
Among many things we like to do here at Madison, we try to challenge conventional wisdom. We’re very judicious in how we do it, as the goal is not to be knee-jerk contrarian but to be stalwartly independent. A simplistic contrarian attitude would have entailed purchasing airline stocks every year for the past century and would have been wrong the overwhelming majority of the time. Our goal is to be independent and right. By definition, a good investment must have contrarian aspects, since the market will reflect general opinion. Charlie Munger once likened the stock market to parimutuel betting, where stock prices are simply reflections of the odds as set by market participants. In a system like that, with millions of participants highly motivated by powerful economic incentives, the wisdom of crowds is wisdom best paid attention to. The art, though, is learning when to ignore it.
We use the word “art” because that’s what it is, no matter how much people may claim to be able to distill stock-picking down to scientific precision. We have yet to find a formula that consists of a comprehensive list of items that can be quantified or categorized and through some magic calculation, spit out market-beating returns without undue risk. Frankly, we hope we never do, because if superior results were achievable that way, our edge would inevitably disappear as other market participants figure it out and adjust the odds appropriately. We try to analyze the intangible and unquantifiable factors without getting caught up in the precision. I guess this is a way of saying that we operate under principles, not formulaic rules. The words principles and rules are often used interchangeably, and the line can be blurry. But the differences are real and profound. Perhaps by giving a few examples, we can convey some understanding of what we mean.
An investment firm may instill a rule that they will only buy stocks that trade at low multiples on current earnings. We think that is a terrible way to think about valuation. It’s possible that such a rule may help reduce the odds that one is overpaying for a company, but any such benefit would be minor compared to the opportunities missed because a company may have temporarily low earnings, or conversely, the risk that the current earnings are temporarily inflated. We prefer to think much more holistically about valuation, and have an overarching principle: we only buy stocks when they trade significantly below their intrinsic value, as calculated by the present value of the stream of future cash flows that can be generated. Adhering to this principle allows us to incorporate much more than one specific year’s earnings, while giving us a clear but non-negotiable guide on what we will buy and what we won’t.
Avoiding companies that have over a 3x leverage ratio (i.e. debt over three times their GAAP operating income) would be a rule. Avoiding companies with financial positions that will put them in jeopardy in tough times is a principle.
Buying companies that have grown revenues at a 15% average rate over the previous three years would be a rule. Buying companies with the ability to grow at above-average levels for many years is a principle.
Investing in companies where the CEO owns 5% of the company would be a rule. Investing in companies where executives and board members think like owners is a principle.
We can go on, but hopefully this communicates the point. Rules can become obsolete over time. Principles won’t. Rules, by necessity, can’t capture complexity. Principles can. Rules create an environment that stifles deep thought and analysis. Principles promote insight.
Managing a research team around principles, not rules, is easy to set up but hard to maintain. It requires constant vigilance around the principles. At times, this may mean an overbearing focus on what may seem irrelevant or minor. UCLA basketball coaching legend John Wooden began the first practice of every new season by teaching his players how to tie their shoes. Warren Buffett refused to take a $1 bet with his friends on the golf course, a bet on which he felt he had no advantage, saying that if he gave in on the little things, soon he would give in on the more important ones.
As an example, recently, we nixed the use of the word “unowned” that an analyst used in explaining our portfolio performance, referring to a stock in the index that had performed well and which we didn’t own. This may seem like a little thing. But the message in the directive is important – we should worry about what we own and not about what we don’t own. And we should have complete and unfettered freedom to be wrong on the stocks we don’t own. We believe the best way to beat the general market indices is by not thinking about the general market indices! Our maniacal focus is to look for investment ideas where we think we can have a strong view backed up by compelling insight and comprehensive information. If we attempt to have a view on a stock where we don’t have both, we’re not only wasting our time; we’re developing a bad habit.
ODE TO FRICTION
In the first half of this year, three large banks failed. Each was so large, in fact, that adjusted for inflation, the banks’ combined assets of $532 billion were greater than the $526 billion of assets held by the 25 banks that failed in 2008 during the heart of the Great Financial Crisis. This compares to a typical year in which few (or zero) banks fail, with perhaps some hundreds of millions in assets in total. So, what is going on?
In each of the failures this year, the culprit was liquidity. Depositors were withdrawing money so quickly that the outflow was jeopardizing the stability of the bank in question. By definition, banks borrow short and lend long, since deposits are technically a form of short-term borrowing. In practical terms, however, banks borrow long, because under the vast majority of conditions, deposits are sticky. Even if there is a lot of money flowing in and out on a daily basis, depositors’ money is fungible to the bank; thus, there is always an underlying amount of deposits available to remain lent out. The key phrase here is “under the vast majority of conditions.” Because there is no true barrier to a depositor wishing to withdraw money, panic and bank runs happen. Deposits are sticky until they aren’t.
This systemic flaw is what federal deposit insurance is meant to address, and for the most part, it’s done so beautifully for 90 years. We think the system probably needs some small tweaks, with one obvious modification being to raise the $250,000 limit for FDIC coverage. We have confidence that the FDIC and the federal government will do what it takes to maintain the soundness of the banking system. We’ve increased our investments in banks this year, as we think undue concerns about calamity risk have driven some share prices to cheap levels. Many are selling for mid-single digit price to earnings (P/E) ratios; at those prices, they more than account for the risk that unfavorable macroeconomic conditions will suppress profits for a number of years. However, given the overall tail risks involved (you lose the same amount of money if a stock goes to zero regardless of whether you paid a 5x P/E or a 15x P/E), we have always capped our exposure to banks no matter how attractively priced they look, and that won’t change. Over the past several months, we’ve essentially gone from very small exposure to banks to moderately small exposure, with positions in US Bancorp, Glacier Bancorp, and Charles Schwab Corporation, a brokerage firm with banking elements. Schwab is a new investment, but we owned the other two banks before the recent crisis, and their stocks remain underwater for the year.
Much has been written about how a defining feature of this year’s bank runs was how easy it is to withdraw money today than it was two decades ago, and how easy it is for fellow depositors to talk to each other and thus spur mass action. One news article noted how, amid the rumors regarding one suspect bank, a group of successful entrepreneurs were on a bus together at a mountainous retreat, heads down, busily tapping away on their mobile phones to withdraw money from the bank. In summary, these bank runs reflected something new – perhaps only in degree, but enough that it portends something qualitatively different: the ease with which we can move money around, and the speed at which information can spread.
Interestingly, these two things can be seen as the reflection of the same thing: our society’s inordinate drive to eliminate friction. Everyone wants to eliminate friction in their lives – who doesn’t want easier access, faster execution, and more choices? A remarkable number of technology start-ups have been founded on this premise, and many have become hugely successful. But there are side effects, and the societal implications are complicated. In some cases, there is likely a point at which the incremental detriments outweigh the incremental benefits. Some outcomes are better when we’re forced to use our slow twitch muscles instead of our fast twitch muscles.
Take stock trading, for example. It was probably a good thing when individuals were freed up to invest in stocks on their own, without having to pay high, regulated commissions and spend time on the phone with registered stockbrokers. But it probably isn’t a good thing that stock trading is now so easy that one or two clicks of the button on a mobile phone could allow one to trade for “free.”
From the standpoint of our job as investors, we think about the topic of friction a lot, from many angles.
We love to invest in companies whose products and services reduce friction for customers. Visa was founded about five decades ago to make paying for goods and services easier. We love its moat, derived from its network effect and scale, but what enabled it to build that scale and moat was the customer proposition of less friction in making payments. Visa has made it so easy to pay for goods that consumers have felt virtually no need to switch to, or even look for, alternative forms of payment. That’s why we’ve maintained our confidence and investment in Visa, even with the deluge of Silicon Valley start-ups and capital that have poured in over the past two decades to try to dislodge the card networks.
Amazon takes this idea a step further. Its management team is fanatical about reducing friction for its customers. They want to make goods as easy to buy as possible (search functionality, fewest clicks, faster page-loading, widest assortment, etc.) and deliver them as fast as possible. The company probably won’t be satisfied until it gets to the science fiction-ish scenario where a consumer will think of an item they want, and it will appear in his or her living room seconds later. (2)
We also love to invest in companies whose value proposition exists because of some inherent friction in the industry. For example, we’re invested in Ferguson, the leading distributor of plumbing and other building products in the U.S. While Ferguson obviously works to produce the best service to its customers, its value proposition exists because contractors need to find the exact product they need from among the thousands of products and manufacturers available in the market and want them all delivered at the right time. Looking for the right product can be a time-consuming process for contractors, and they gladly consult with Ferguson to figure it out. Having the right products delivered on time, in the right order, is a logistical issue that has physical constraints and can’t be “solved” away. The key here is that we believe some of the friction is a structural condition that we don’t think will disappear anytime soon. Thus, we see Ferguson’s value proposition as durable and shielded from obsolescence.
We love companies that obsess about reducing friction in their interface with customers. Our long-time portfolio company Copart went from physical auctions of its vehicles to entirely online auctions many years ago, well ahead of other industry participants. The move allowed it to reach global customers in a way it never could before. Its competitors have since established online auction capabilities but remain far behind Copart in developing the breadth and depth of their customer base and ease of access, and we don’t anticipate they will do so anytime soon. As long as that is the case, we believe Copart’s larger pool of buyers will give it a competitive advantage. Another portfolio company, Nike, has steadily been reducing its reliance on third-party retail chains to sell its wares and increasing its direct relationship with consumers. This sort of reduction in friction boosts the customer experience and also provides direct benefits to Nike, such as lower costs and better data on its customers.
Another kind of friction we think about with our portfolio companies is the switching costs that their customers may have to move to a competitive product. We love industries and products where the switching costs are high. We recently invested in Waters Corporation, which makes liquid chromatography and mass spectrometry instruments. These products are used to measure the chemical and molecular composition of pharmaceuticals and materials, and because companies need consistency in the way they measure such highly complex materials, they are reluctant to switch instruments. Often, they are prohibited from doing so by regulators that have approved a drug based on the measurements taken by a certain vendor’s products. The critical point to us is that the high cost to switch is not because the vendor is putting up walls or making it difficult to leave. The cost to switch is something fundamental to the product and service itself.
CONCLUSION
Speaking of switching costs, we understand that you have many options for your or your clients’ hard-earned savings. We, therefore, endeavor to do everything we can to deliver market-beating performance without taking on undue risks, and provide clear and honest communications about how we are doing so. We hope these semi-annual letters at least speak to the latter point.
Thank you for your confidence in us. We look forward to updating you at the end of the year.
Respectfully,
Haruki Toyama