Hand-Crafted Artisanal Liquidity Provision
Before Goldman Sachs was a famous financial colossus, it consisted of Marcus Goldman, whose job involved a lot of brisk walking: he'd visit small businesses that had accounts receivable, buy them at a discount, then trot over to local banks and sell those accounts receivable at a slightly smaller discount. Evocatively, the tradition was for bankers to stuff the loans they're purchased into their top hats, so you could measure the general financial situation by looking at how high their hats were on their heads.
This is a nice story because it illustrates two parts of the financial system:
- The passive accumulators and buyers of financial claims: small businesses ended up being owed money in the course of their business, and could wait for Goldman to show up and convert those claims into cash. Meanwhile, bankers with large balance sheets and hefty cash reserves could sit behind the counter and wait for someone more entrepreneurial to show up.
- Intermediating between these two fairly passive financial actors was Mr. Goldman, whose job turned over his capital in proportion to how quickly he wore out his shoes, and could scale profits purely through harder work.
Mr. Goldman's firm has come a long way since then, though senior bankers there will certainly accumulate their share of airline miles. The firm's origins reveal something important about the financial system: that there are numerous strategies that technically earn some return on capital, but are mostly defined by the input of effort required, rather than the input of money. You can compare them to other strategies, but if you just judge them based on return on capital, you're missing something critical.
One classic example of this is "odd-lot arbitrage": sometimes, companies will run a tender offer for their stock, offering to buy shares from a large set of shareholders all at once. Often, the structure of the deal will be that shareholders can agree to sell into the tender, and if there are more shares tendered than the company wants to buy, they get bought pro-rata. So if there's a tender for a million shares, and owners of four million participate, each shareholder who tendered gets a quarter of their shares bought. But there's often an exception to this: someone who tenders fewer than 100 shares, or in some cases fewer than 1,000, will get all of them bought.
Since companies typically tender at a premium to their current valuation, and since the price after they announce the tender reflects some bet on how many shares the average holder will successfully tender and where the price will be post-tender, odd-lot trades usually work. A company trading at $20 might tender at $24 and see its price rise to $22 (implying that holders think they'll successfully tender half of their stock, after which the price will drop back down), while the odd-lot holder gets to buy at $22 and almost certainly sell at $24 in a few weeks.1
A look at this useful site shows why this free-money strategy is not an especially popular one: in the last year, investors could have pocketed all of $3,277.89 from these trades. On the other hand, the largest position they'd have to take would be $8,686.26 worth of Trinet, so an odd-lot strategy would earn unlevered returns of 38% annualized with very little risk.
(The natural question arises: why do companies bother to give away smallish amounts of money to tiny shareholders? One reason is that they're trying to reduce their shareholder count, in part because shareholder communications is an expensive monopoly operated by Broadridge Financial Solutions, which charges up to $0.25 per email for shareholder communications ($, FT).)
This is not the only non-scalable strategy out there, though it's because of how straightforward it is and just how little capital can be deployed in it. Others include market-making in obscure markets, where the annual profits probably aren't large enough for a proprietary trading firm to devote a seat to it but where the spreads are wide and there's at least some demand; small-cap special situations generally; and shorting microcap scams. These strategies might be characterized as, not just active management, but hyperactive management—data- and effort-intensive, with a fairly short-term payoff, and an input measured more in time than money.
It sometimes seems to be the case that when you peel back the layers and look at the history of trading companies, they often start with exactly this kind of situation—some small-scale trade that isn't worth a big firm's time, but that can pay the bills at a smaller one. (A similar case is Michael Milken's career; when he started trading junk bonds, the market was tiny, illiquid, and low-status; he was stuck with a bad job at a smallish firm. But in that case, the market wasn't structurally small; it was limited in part because nobody had been willing to make a market. Liquidity begot liquidity, and Milken's trading profits were high enough that he was sufficiently well-capitalized to keep up with the market as it exploded in size.)
A great term I’ve seen on fintwit a few times is “Return on brain damage”: there are some cases where, yes, you have positive expected value in financial terms, but not only will you do a lot of work to get there, but you’ll spend a lot of your non-work time distracted and angry along the way. Two standard instances of this:
Shorting overvalued concept stocks with high borrow rates and promotional management: If you think a stock is going to zero in a year, it can be a perfectly good deal to pay 30% to borrow the shares.2 Since shorts are volatile and have unbounded losses, the position has to be small. And that’s especially true for investors who are running a sizable portfolio, since many of these stocks are small- or micro-cap. What ends up happening in these trades is that 1) there’s a material cost to holding the position every single day, 2) promotional management sometimes finds ways to talk up the stock, and 3) when this happens, dumber investors make money, and there is a correlation between poor lifetime investment returns and propensity to gloat. Size the position at 1%, pay a 30% borrow, see the stock decline 50% in a year, and you have a position that has contributed a net ~31 basis points to your annual return and probably shortened your career and/or life expectancy by a few months due to added annoyance and stress. And that assumes a steady pace of decline; there’s probably a brief period in there where the stock doubles and the borrow triples, and even if you know you’re right, you may not be able to hold on. Sure, you could diversify, but that also increases the flow of annoying company announcements to keep track of.
The other instance is cheap-for-a-reason stocks that are simply too cheap. The reason is usually either a structural problem with the business or a controlling shareholder who is indifferent to the concerns of other owners. Sometimes you get both! There is a microcap stock out there that recently traded at a mid-single digit P/E and below net cash, because management had consistently been so awful to shareholders and had failed to sell the company when good opportunities presented themselves. On one of their last earnings calls before they stopped talking to investors entirely, one of the analyst questions was "has the board considered just resigning and letting someone else come in who knows what they're doing?" Is the prospect of eventually seeing the stock double and trade at 5x earnings plus cash on hand really worth being mad enough to dial in to a conference call and yell at management?
Both of these situations probably have positive financial returns and negative overall returns after accounting for the shear annoyance entailed. Of course, tolerance for that kind of annoyance varies from person to person, and some people absolutely thrive on it; in early interviews, Carl Icahn sounds absolutely thrilled that some board of directors has the temerity not to do what he says. Unfortunately, 1) most of us don’t have this trait, 2) most of the time, sufficient success at this kind of work leads to higher assets under management, meaning that the smallest and thus gnarliest of these opportunities no longer make even financial sense, and 3) getting habituated to brawling with management makes it harder to pivot into gentler strategies with higher capital capacity.
So why do these situations matter at all? They matter because they're such a good training ground. If I were talking to a high school student who had decided to become an active investor (something I'm familiar with!), I'd point them in the direction of odd-lot tenders: the capital requirements are low, but to do the strategy effectively you have to trawl through lots of boring-by-design SEC filings, which is excellent practice. Non-scalable strategies one tier up from this can pay the bills early on while investors design more scalable ones. And there's a fairly smooth continuum from extremely inefficient and illiquid markets in microcap stocks and some foreign countries to more liquid but more efficient markets. The math is usually easier, and the qualitative concerns are a lot more salient—"bad management" at a Fortune 500 company means overpaying for acquisitions, while "bad management" at a tiny company in a lightly-regulated jurisdiction can mean outright embezzlement and fraud. Artisanal liquidity provision basically ages investors in dog years, which means they offer a lot of painful but informative experience.
A Word From Our Sponsors
Tegus is the first port of call for M&A professionals and institutional investors ramping up on an industry or company.
Get access to a database of 35,000+ expert call transcripts, spanning 5+ years, or schedule expert calls through the platform for a fraction of the usual cost.
When thousands of research analysts are pooling their expert calls into an on-demand database, using Tegus is table stakes. It's the leading platform for due diligence and primary research.
See the power of a Tegus subscription, and get up to data parity with your competitors, with a two week free trial through the Diff.
Legendary media investor John Malone has dropped one of his occasional interviews, where he notes that the streaming business is not nearly consolidated enough. (Malone was a player in the Warner Brothers/Discovery deal, which created a large and challenged company whose bets on legacy content distribution are partly offset by an increasingly important streaming bundle.) Per Malone: "The smaller guys can’t get to scale... They’ll inevitably have to combine in order to try and become profitable." That consolidation can take the form of horizontal mergers, but it can also mean what Paramount is doing: offering its product as part of a bundle with Walmart's membership program, since even a sub-scale streaming business can be a nice complement to something else.
The Diff has been writing for a while now about how shareholder activism in Japan might finally collapse the gap between Japanese valuations and the rest of the rich world's ($), but that's not the only approach: Warren Buffett has taken vocally non-activist positions in Japanese shares, arguing that we shouldn't assume Japanese companies are mismanaged just because they finance themselves differently from US companies.
The simplest way to describe the difference is that it's not uncommon to find Japanese companies that have half of their market cap in cash, or more. This is an extreme level of financial conservatism. The tricky thing about this kind of situation is that it's very sensitive to catalysts: if a cash-heavy company decides to return that cash to shareholders, whether it's through dividends, buybacks, or selling out, the return is realized immediately. But if it takes a long time, unlucky investors are stuck with a business that doesn't grow much because there isn't any reinvestment, plus a cash position that doesn't yield much, and that has, in the last year, lost 20% of its value against the dollar.
And on the topic of corporate cash, US non-financial and non-utility investment-grade companies have reported a 22% year-over-year decline in cash on hand ($, FT), a record going back to at least 1998. Some of this is because companies stockpiled so much cash during the pandemic. On the other hand, bank loans are growing massively year-over-year, too, so it's not as if corporate America is getting collectively more conservative. These data points would be very strange-looking in a recession, at least in nominal terms.
(If you want to describe a recession in real terms, I'd start with the decrease in quality for service-related jobs—from the perspective of customers, but also from that of employees—since March 2020. In a very significant sense, real output was hit hard by the pandemic and direct policy responses, and the fiscal and monetary response prevented a financial crisis but didn't mitigate the wealth destruction. But that kind of recession is endlessly debatable.)
PE and Software Compete to Eat the World
An early Diff* post argued that SMB-focused software companies and PE rollups were both executing the same model: applying big company efficiencies to fragmented industries and capturing a large share of the profits. The WSJ has a good piece on how this is playing out, with a focus on the car wash industry ($). That industry has historically had lumpy profits and few economies of scale, but a shift to membership models makes the business more predictable—and creates a high-margin recurring revenue stream that can be borrowed against—turning it from a naturally fragmented sector to one that can support large-scale operations. Figuring this out creates a nice long-term growth path, since the company can find tricks to increase performance at individual locations while continuing to acquire new ones at prices that are set by the business's historical economics.
And see also this Diff writeup of PE-backed Mister Car Wash ($).
Reinsurer of Last Resort
There will always be debates about the proper role of government, but from a financial perspective one thing governments do is offer the "reinsurance of last resort," by dealing with problems that are catastrophic, expensive, and hard to coordinate through market mechanisms. Such as, to take the easy example, military conflict.3 The insurance industry can cover some risks, but not those, and Lloyd's has announced plans to exclude nation-state hacking from cyber insurance coverage ($, WSJ). Part of the point of this is to be explicit about what is and isn't covered; insurance is all about predicting uncertain events, and one element of doing it well is reducing the odds of litigation over what specific form of uncertainty manifested itself. But it's also evidence that state-sponsored hacking will be a growing threat for some time.
Diff Jobs is our service matching readers to job opportunities in the Diff network. We work with a range of companies, mostly venture-funded, with an emphasis on software and fintech but with breadth beyond that.
If you're interested in pursuing a role, please reach out—if there's a potential match, we start with an introductory call to see if we have a good fit, and then a more in-depth discussion of what you've worked on. (Depending on the role, this can focus on work or side projects.) Diff Jobs is free for job applicants. Some of our current open roles:
- A company offering working capital to small businesses is looking for an operations and strategy lead. (SF)
- A startup that helps companies' financial systems seamlessly communicate is looking for a mid/senior-level engineer, C#/.NET experience ideal. (Austin)
- A company that is building web3 tools for brands to better attract and retain customers is looking for engineers. Web3 experience is not necessary, but always a plus. (US, remote)
- A company in the alternative data space is looking for an equity research specialist. (US, remote)
- A company in the education space is looking for senior product people to manage new strategic initiatives. (US, remote)
Technically minded or contrarian readers will note that this is not strictly true, for a few reasons. First: if the company really was worth $20, the tender reduces its value a bit because some shares have been repurchased at a premium to its real value. If the company had 50m shares outstanding, and bought them back at $24, then after the tender the fair price would be slightly lower, at $19.92. But tender offers also have signaling value! Sometimes they indicate that management is convinced that the stock is cheap, and worried that it will rebound fast. Investors who sold Teledyne into Henry Singleton's various offers made good money, but the ones who held made more.
And, quite relevant to the matter at hand, Warren Buffett has a story about how in the 1950s, a chocolate company swapped cocoa beans for shares, presenting an easy arbitrage—but that the better deal was to buy the stock and hold it, since the tender continuously reduced its price-to-beans ratio despite occurring at a premium.
On the other other hand, the trouble with any kind of signaling behavior is that other people understand signaling, too, and sometimes companies will tender specifically to indicate that they're more confident in their prospects than is really warranted. ↩
If you ignore borrow costs, short interest is a great predictor of excess returns. Which means that part of being an effective short seller is structuring the trade so you minimize those borrow costs. There are all sorts of tricks here: for smaller investors, using options instead of shorting directly is basically outsourcing borrowing to the options market-maker, who probably gets a better deal than you do. Some funds lock up more shortable shares than they really need, and temporarily offset them with long swaps, because at the time when they’d want to scale up the position, it may be harder to borrow. And on the other side of the equation, being long shares of a company that’s in terminal decline, but that’s good at staving off the inevitable, can be a winning proposition: if the shares lose, say, 30% of their value every year, but you’re collecting a 100%+ annual borrow, you still win.
These concerns apply less to purely valuation-driven shorts and more to fads and frauds. But anecdotally very, very few people have assembled enviable track records by focusing their shorting on catalyst-free bets on valuation. ↩
Though this can end up semi-privatized in cases where there's a weak government in a country with abundant natural resources. Squint carefully and coups are an unusually literal interpretation of the term “hostile takeover.” And there are earlier versions; this Caesar biography describes the Roman approach to war in a roughly similar way, with politicians going into debt to get the votes they needed to raise armies, then paying off their debts with the proceeds of invasions. ↩