The Next Berkshire Hathaway Will Be Built on Technology

Technology is eating your portfolio

What is the best way to invest in a time of great technological disruption? Everywhere we look, it seems like new, emergent technologies are disrupting old ways of doing business. Ride sharing, self-driving cars, hotels and hosted stays, stock picking and trading, retail, computing, machine learning, telecommunications, blockchain and Bitcoin — make no mistake, these new technologies are tearing through traditional industries, doing the same old jobs faster, for less money, and with greater security.

Technology is no longer a vertical, technology is every vertical.

Because value investors seek predictable but mispriced businesses, their universe of investable assets is shrinking. We learned from the late 90’s tech bubble that while the prices of technology assets may temporarily rise to irrational prices, eventually, the promise of technology will come to fruition and disrupt old ways of doing business. Witness’s disruption of retail — and computing, and now possibly a slew of other industries with Alexa.

While value investors can hide for a time in increasingly obscure and esoteric businesses (industrial sprayers, chemicals, automotive seat making), eventually, these businesses will most likely be disrupted. Rather than retreat from this reality, capital allocators must seek to understand the forces at work in technological disruption and create a way to allocate rationally and responsibly, in a manner consistent with the best value investing principles of Graham, Buffett, and other luminaries like Joel Greenblatt.

This point was underscored by Buffett and Munger at the 2016 Berkshire Hathaway Annual Meeting. They expressed their respect and admiration for Jeff Bezos, and shared how they could have understood Google and probably should have invested in it, but did not because it involved too much new technology. While it is of course perfectly fine for the best investors in the world to miss a few big opportunities, if you are looking at investing for the next few decades, you need to understand how technology is impacting the world and business, and develop a strategy for allocating within it, to protect your wealth from disruption, and take advantage of new technology-enabled opportunities — all while remaining true to timeless value investing principles.

A unified theory of capital allocation

In this article, I’d like to suggest a framework for investors to think about allocation to technology-enabled companies, at various stages in their growth. My objective here is to present “a unified theory of capital allocation,” bringing together the best of value investing and venture capital/angel investing into a single coherent framework that an intellectually flexible investor can follow. At its heart is the core of all true value investing — share ownership represents a stake in a real business with underlying cash flow; those cash flows can be analyzed from a fundamental, long-term basis; rational capital allocation for the enterprising investor involves buying businesses at a discount to their long-term intrinsic value, defined as the present value of all future cash flows produced by the business.

First, it is important to note that there are precedents in the value investing canon for investing in assets with asymmetric payoffs, in which the downside is X, but the upside is a multiple of X — say 10x. Joel Greenblatt popularized the idea of LEAPS — long dated call options and warrants on undervalued stocks. If the full value of the underlying stock is realized, these securities will pay off a multiple of their cost. If the underlying stock does not realize its intrinsic value, or some other event like a strike price, these securities are worthless. Greenblatt:

“See, there is an added benefit to owning the calls [LEAPS] — it’s the benefit of not losing any more money after the stock falls below the strike (or exercise) price…. The bottom line is that buying calls is like borrowing money to buy stock, but with [downside] protection. The price of the call includes your borrowing costs and the cost of your “protection”, so you’re not getting anything for free, but you are leveraging your bet on the future performance of a particular stock. You are also limiting the amount you can lose on the bet to the price of the call.”[i]

The upside is highly asymmetric to the downside. The key is in analyzing the underlying stock to understand the likelihood of full value being reflected in price. While far from easy, it is possible.

The other precedent for investing in these types of securities/assets is Nassim Nicholas Taleb. In his book Antifragile, he discusses the concept of being “barbelled”. A barbelled portfolio is one in which 90% of the assets are deployed in very safe assets, such as cash, secure bonds or the equity of high quality companies acquired at reasonable prices. The remaining 10% is in a diversified pool of assets that may go to zero, but also have a reasonable probability of returning a multiple of the capital invested. Across that diversified pool of assets (within the 10% portion), meaningful asymmetry is achieved — “heads I win, tails I don’t lose that much.” Taleb:

“I initially used the image of the barbell to describe a dual attitude of playing it safe in some areas (robust to negative Black Swans) and taking a lot of small risks in others (open to positive Black Swans), hence achieving antifragility. That is extreme risk aversion on one side and extreme risk loving on the other, rather than just the “medium” or the beastly “moderate” risk attitude that in fact is a sucker game (because medium risks can be subjected to huge measurement errors). But the barbell also results, because of its construction, in the reduction of downside risk — the elimination of the risk of ruin…. Let us use an example from vulgar finance, where it is easiest to explain, but misunderstood the most. If you put 90 percent of your funds in boring cash (assuming you are protected from inflation) or something called a “numeraire repository of value,” and 10 percent in very risky, maximally risky, securities, you cannot possibly lose more than 10 percent, while you are exposed to massive upside. Someone with 100 percent in so-called “medium” risk securities has a risk of total ruin from the miscomputation of risks. This barbell technique remedies the problem that risks of rare events are incomputable and fragile to estimation error; here the financial barbell has a maximum known loss.”[ii]

The quality of the assets in the 10% pool with asymmetric payoffs depends on the quality of the fundamental analysis done before investing in them.

Technology startups conform to this asymmetric payoff structure. They may go to zero, or they may turn out to produce meaningful future cash flows and return a multiple of capital invested, even after dilution. Not all startups are speculative, though some are. The best startups share a handful of qualities which, taken together, constitute a contrarian, variant view on how a given job should be done or industry should be structured. It is possible to develop a coherent investment thesis for many startups, and to do the fundamental analysis needed to understand both the opportunity the startup seeks to capture, as well as the industry and what most players fail to understand about it.

How to Research Startup Investments

While there are many frameworks for analyzing startups out there, I’d like to advance one that is more focused on the company itself rather than the founder or team. If you have the chance to invest in Elon Musk’s next venture, you should probably take it, given his remarkable track record of wealth creation and success. Most of us won’t have that opportunity, however, and as such we need a framework to analyze startup businesses and their potential payoffs.

My objective in this section is not to say that other startup investment frameworks are wrong, but rather to compliment the good work already done with a means of evaluating nearly any investment, including startups, at least from a qualitative perspective. The other key to investment success — valuation — will also be discussed.

Startup investors with a value orientation (angel, VC) should look at four things, at a minimum, to assess the potential of a startup to create wealth. I don’t mean to say that these are the only 4 things — rather, these should constitute an initial screen based on reasonably observable traits and publicly available information (such as that pertaining to the industry size or total addressable market).

1. The job to be done. This idea comes from Clayton Christiansen in his book The Innovator’s Dilemma. What is the startup basically trying to do? What problem is it solving? Next, look at how many people have this problem to try to get a sense of the size of the opportunity or market. Another way to think of this is as a proxy for the startup’s industry. Are they trying to provide people with a better way to get educated? Are they improving how manufacturers procure raw materials? Did they make a better artificial heart valve? Answering this question is often relatively straightforward, and can be done by looking at both industry revenue figures (top down) and analyzing how many people or companies have a given problem and what not solving it costs them (bottom up). Sometimes solving for this can be challenging — for example, what problem does Facebook solve? Human interaction, communication? It may not always be easy.

2. The technology the startup deploys to accomplish the job to be done. I would argue that every business relies on some form of technological base, even if it is low tech or analog. Take agriculture for instance — there was a time when humans relied on hoes and gardening implements to cultivate the land. Then we domesticated animals, resulting in an order-of-magnitude improvement in productivity and scalability. A horse drawn plow, for example, enabled humans to cultivate far more land than before, resulting in all kinds of surpluses and knock-on benefits for civilizations involved. The same is true for the invention of the tractor. Now we can produce huge amounts of food for de minimus labor. A good startup does something similar — it identifies an important job to be done, one that is shared by many people, and couples it with a technology that involves a radically discontinuous improvement in cost, time to be done, ease of use, security, or any other relevant metric. It collapses the investment required to do the job, and captures a portion of that value. Examples include improvements in telecommunications, computing power, transportation, and now potentially financial services with blockchain and cryptocurrencies. There are many more — the key question to answer here is: how is the technology leveraged by the startup radically better than existing solutions?

3. The moat. How will the company eventually, preferably sooner than later, create a defensible business? Technological improvements, if they can be copied, do not constitute a moat. A moat is required to capture the surplus value created by the startup. A moat is defined as the ability of the startup to earn returns on capital above its cost of capital over a prolonged period of time — ideally until the next technological innovation comes along for that particular job to be done/industry (think decades). According to the great stock research firm Morningstar, moats come in five forms, which conform to the acronym LINES: Low cost provider; Intangible assets like patents and regulatory licensing; Network effects; Efficient Scale, in which case a market is fully supplied with a handful of producers; and Switching costs. It is worth asking if and how the startup can achieve a moat, with sufficient capital and time, in its market to help ensure defensibility and protect against commodification by competitors. Moats often take huge amounts of time and money to build — rarely are they endemic to the nature of the product or company. Even so, identifying a moat strategy and pathway to building one ex ante is desirable in a startup investment.

4. The idea maze. Successful entrepreneurs have navigated the idea maze for their startup. The idea maze is a concept that comes from VC firm Andreeson Horowitz (in fact, investors in any company must go through a similar process, regardless of the investment size or stage). It denotes a process whereby a founder, through research, critical thinking, incessant questioning by themselves and their friends, family, colleagues, potential investors, competitors, and so on, has come to understand their product, their market, go-to-market strategy, and the full range of scenarios that could impact the company, as well as any key risks that could cause their thesis to shift, and how. In my own startup journey, this process involved asking everyone I could for feedback on the idea, and then noticing their responses. Over time, these responses begin to cluster around a handful of key questions, risks, and objections. If the idea is exposed to enough people, the founder will see the common issues people see in it, and she can develop strategies to address those. She’ll see where people misunderstand the market or idea. From time to time, she’ll encounter a person with exceptional knowledge or a unique insight, which often throws her for a loop and in turn must be incorporated into her overall thesis. After a sufficient amount of time, she knows how to answer nearly every objection and mitigate just about every risk. Investment-ready entrepreneurs have already been through the idea maze. When you talk to them, asking them questions about the idea, you get the feeling that you’re learning from an expert that has a depth of knowledge far beyond your own, rather than talking to another generalist or lay person with only basic knowledge of the subject. If you ask the founders questions, and they are surprised or don’t know how to answer them, they aren’t ready for outside money.

Other metrics, such as unit economics, founder pedigree, tenacity/fanaticism, and background (have they been coding since birth?), deep technology assessments, product functionality and design, and so on all matter. They’ve been written about elsewhere, and rather than reinvent the wheel, I want to focus here on how value investors can wrap their heads around startup investing, should they decide technology investments have a place in their portfolio. Indeed, my hope is that not only startups, but also mature companies, can be evaluated by looking at the job, the technology, and the moat possessed if any. This framework becomes not only a means of looking at startup investments, but a way to evaluate the potential disruption to any mature businesses that one may own. If you are a major investor in financial services companies, for example, understanding the activity in the blockchain startup investment space is both prudent and in your best interest. The same applies for other industries.

Now the other key piece — how to value startups?

Value investors know that successful investment involves a process of buying assets at a discount to their intrinsic value. This means that valuations play a huge rule — understanding how much cash can be taken out of a company over its lifetime (or at least during the holding period), as well as the price at which those future cash flows are for sale in the market. The objective (simple, but not easy) is to buy cash flows for less than they are actually worth — buying the proverbial dollar for 50 cents. Or, if a company is worth $250 per share, buying it for $150 or less per share, to use one example.

In a startup context, valuation is a huge challenge. Financing instruments, such as convertible notes, emerged to essentially bypass the valuation question. While it is possible to get pricing of startup shares wildly wrong, there is a high degree of variability in what a rational pricing figure might be. The potential accuracy of the valuation exercise is to a large extent a function of the stage of the company. Below is a rough map of the funding rounds of startups, the key milestone they need to achieve at each stage to proceed to the next.

Seed — Product/Market Fit. The company has made something people want, understands what their core product value is, and is able to deliver it using at least one channel.

Series A — Build a functional small company. Hire key staff to begin to achieve early scaling milestones. Begin to understand, invest in, and execute upon moat building. Fill out any holes in the product and gain clarity on the product roadmap. Investments result in high returns in the form of revenue and market share gains.

Series B — Build out the executive team and key employees to scale. Here the company knows clearly what the moat source will be and they’re in a sprint to invest to achieve scale as fast as possible. Product here should be stable, complete and competitive in the market. Market share grows very quickly.

Series C — Hire middle management and invest as much as possible in the moat. This can be through debt or equity capital, or both. Dollars invested here turn into tens or hundreds of dollars later. The company is gobbling up market share. The product is stable and robust.

IPO or liquidity — The company is a relatively stable, if fast growing business, with a clear product, strategy, and moat source. The company is (re)investing external funds and the substantial internally-generated funds in anything that can help them win market share and widen the moat.

As the company progresses through these fundraising rounds, valuation will become easier and easier, although there will still be meaningful uncertainty. Because the company is so high growth, traditional valuation metrics probably won’t come into play here. Remember the key here — until they’re public or late-stage private ventures, startups have a dual payoff structure of zero on the downside and a huge multiple on the upside. Therefore revenue, revenue growth rate, and total addressable market will play a bigger impact than metrics like free cash flow, EBITDA and profitability. Less tangible features like the product, the technology base, and the moat should be strongly considered here.

At the very early stages, unless the valuation is clearly wrong (EG a $50 million valuation for a company with no users and no revenue), valuation is very hard to do well. Simply put, if the startup works out, you will have bought at a huge, material discount. If it does not work, you will have overpaid at nearly any price. Because the payoffs tend to very binary, especially for early stage startups, escaping this duality on valuation will be hard. Protection will come less from a traditional margin of safety, in the value investing sense, and more from the depth of your research, your conviction on the technology, product, or team, and the diversification you will get by investing in not one or two, but ten, twenty, or thirty startups — probably over a period of several years. Fat pitches just don’t come around very often, and there are no strikes called in startup investing, either.

Constructing a portfolio is thus an exercise of coupling the investment in mature, stable, cash flow producing businesses bought at reasonable prices, with a tiny percentage of your portfolio in a handful of well-researched startups with the potential to remake entire industries. If you’re able to access the deal flow, investing in later-stage startups may also take a place in your portfolio. If you’re not able to access that kind of proprietary and exclusive deal flow, most likely the best way to gain access to it, other than building out the networks to the startup and professional VC community, is to invest early at the seed stage, and then follow on with more funds as the startup grows. Thus, startup investing becomes a process of putting a tiny amount money with a handful of promising companies, and then following on with more substantial funds with the companies in your portfolio that do well. In short, selectively plant your seeds then double down on the winners.

The Next Berkshire

The next Berkshire Hathaway will incorporate an understanding of how technology companies are built, and the role technology plays in business, and will play in the future. It will combine a handful of large, profitable, mature businesses with a portfolio of tiny startups — experiments in doing important jobs in society on a discontinuously-better technology base — with growing businesses that are nurtured within the company with the reinvested cash flow of the larger business lines. In fact this company already exists — it is

The next generation of investors would do well to learn from and replicate as much as possible what Bezos has done. He couples mature businesses with countless startup experiments, most of which will fail, some of which will become medium-sized businesses, and a tiny fraction of which may become another billion dollar business within Nothing fundamentally disallows investors from doing this — all the pieces are out there in the public domain — but so few actually do it.

The next Berkshire will inevitably involve technology and will, like Buffett and Munger, be focused on a relentless search for mispriced moats, built on technology that is durable for the foreseeable future, with talented and moral managers. For the next several decades, investors should focus on finding mispriced or cheap moats on the most efficient, cheap, effective, secure or reliable technology, protected by substantial moats, wherever they find them. They might find this in a startup, a growth company, or in a mature business. While the ease of analysis, valuation, and position size all shift based on the size and scale of the business and the probability of loss, the core discipline of understanding the company (job to be done, technology, moat, and management competence) is the same regardless of the stage of the company. It is my hope that some of the ideas here can help investors find mispricings wherever they may occur, and provide the reader with a unified theory of investing across the full lifecycle of businesses.

[i] You Can Be a Stock Market Genius, by Joel Greenblatt. Page 217.

[ii] Antifragile, by Nassim Nicholas Taleb. As quoted from Accessed 7/29/2017.