In Praise of the Crowded Register

A founder I worked with once held roughly seventy percent of his company after a listing that floated just enough to satisfy the exchange's free-float rule and not one share more. He was proud of it. The register was a short list of names he trusted, a couple of strategic holders he had known for years, and a thin sleeve of public float that traded by appointment. By the concentration-and-duration test, it was a beautiful register. It was also, on the days the stock moved, almost untradeable. A single mid-sized institution deciding to exit could not find the other side of its trade for a week, and the price gapped down on volume that would not have registered as a rounding error in a liquid name. The holders he had were exactly the holders he wanted. They were also, collectively, a liquidity problem he could not communicate his way out of.

That register is the live counter-example to an essay I published. The essay said you choose your investor base. This founder had chosen his, precisely, and the choosing had a cost he had not priced.

So let me put my own essay on the stand.

1The position we took

Essay 08 of the Bedrock Papers, The Investor Base You Have Is the One You Chose, opened with a confession and built to a verdict. The confession: "we do not really know who we are targeting." The verdict: the practitioner tradition's instinct to chase more, meaning more institutional ownership and more analyst coverage and more breadth, is optimising for the wrong variable.

The essay leaned on Cunningham's clientele effect: "a company's policies, communication style, capital allocation decisions, and disclosure practices organically attract shareholders whose preferences match the company's actual approach and repel those who don't. The shareholder base is not exogenously determined by markets. It is endogenously chosen by management's communication and operating decisions, whether management realises this or not."

And it was pointed about breadth. Growing the institutional base broadly, the essay said, "is not a neutral activity. It is a positive choice to dilute the shareholder properties that the literature on long-run governance, narrative coherence, and intrinsic-value pricing suggests are correlated with corporate success."

I believe most of that. But the verdict on this essay is "hold," not "vindicate," and a hold has to survive its best objection out loud. Here is the objection.

2The counter-case, steelmanned

The first thing the practitioner literature does is treat the crowded register not as a failure of discipline but as the goal. This is not a fringe position the Bedrock essay can wave away as the consensus it had already discredited. It is the operational recommendation of the people who actually run IR departments, made for reasons that are mechanical rather than aspirational.

The investor-base-diversification material, from Bragg's operational guide to running an IR department, states the ideal directly: shift "from small number of owners with massive holdings to larger shareholder base with proportionally fewer shares each." And it names the specific harm that concentration causes, the one my founder discovered the hard way. "Excessive institutional concentration is problematic—large-block sales drive down share price." There is the block-sell risk, stated as a first-order concern. And then the liquidity point, which the Bedrock essay never engaged: institutions "create illiquidity by accumulating so many shares that few remain for trading."

That sentence is the whole rebuttal in miniature. The very property the Bedrock essay prized, large positions held a long time, is the property that drains the float. Concentration and duration, celebrated as quality, are mechanically the same thing as illiquidity, deplored as a defect.

You cannot have a register of patient holders sitting on big blocks and a deep, tradeable market in the same shares.

The cost of losing that pull is not soft. The capital-liquidity-premium material, from Bragg's public-company manual, is explicit that liquidity is the core economic reason to be listed at all. A public market eliminates the illiquidity discount that suppresses private-company valuations, adding "a premium to what the shares would have been worth if the company had remained privately held." The kicker: this premium "reduces cost of capital by several percent." Several percent. That is not a rounding error in a discount-rate model; it is the difference between a project clearing its hurdle rate and not. A company that engineers a concentrated, low-float register to cultivate quality is, in the literature's own terms, partially forfeiting the liquidity premium it listed to capture. It is paying a higher cost of capital for the privilege of a prettier register.

And the diversification case is not only about liquidity and cost of capital. The global-investor-base-diversification material, from Yücel's playbook, adds the volatility argument and frames it geographically. Geographic diversity in the investor base provides "a buffer against the volatility of local markets" and stabilises stock performance; broader distribution "reduces concentration risk." This matters enormously for an Asian issuer. A register dominated by domestic holders moves with domestic sentiment, domestic liquidity cycles, and domestic policy shocks. Spreading the base across geographies is not breadth for breadth's sake. It is risk management: the kind that keeps the stock from gapping when local money all heads for the exit at once.

Then comes the part of the steelman that genuinely bounds my thesis, and I want to give it full weight because it is the strongest blow. The Bedrock essay's title asserts a choice: the base you have is the one you chose. The literature says you cannot fully choose it. The investor-revealed-preference material, from Stout, makes the point through an empirical observation that cuts at the clientele effect's premise. When companies go public, "many adopt dual-class equity structures that concentrate voting power and control in insiders' hands. Google, LinkedIn, and Zynga are prominent recent examples." Investors "eagerly buy shares in companies that strip them of power." Meanwhile, charter provisions that would give shareholders more leverage over directors "are so rare as to be almost nonexistent."

Read against the Bedrock thesis, that does real damage. The clientele effect says the base is endogenous to management's communication and governance choices. But Stout's evidence shows the base is also constrained by structural facts the communication cannot touch. If you carry a control block, your free float is what it is, and the holders who buy the float buy it knowing they cannot vote it to any effect. You did not communicate your way to that register. The capital structure delivered it. The investors who arrive are revealing a preference for the structure, not responding to the equity story. The base, in other words, is not fully a thing you choose. It is partly a thing your float and your control block hand you.

So the steelman, at full strength: the crowded register is the recommended goal, not a mistake. Breadth raises liquidity, dampens volatility, buffers against the block-sell that craters a thinly held stock, and recovers a cost-of-capital premium worth several percent. And the founding claim, that you choose your base, is overstated, because float and control-block realities bound the choice in ways no communication programme can override. The Bedrock essay sold cultivation as the answer. The practitioner literature says diversification is the answer, and the governance evidence says you do not have as much choice in the matter as the essay implied.

3The verdict: hold

This is the strongest counter-case in the set I have argued against, and it changes the essay in two real ways before it leaves the rest standing.

The first concession is liquidity, and it is not a small one. The Bedrock essay treated the crowded register as a defect to be corrected, and it simply did not price the cost of correcting it. The literature is right: a register of concentrated, long-duration holders is mechanically an illiquid register, and illiquidity is not a cosmetic problem. It widens spreads, deters the very long-duration institutions that have minimum-liquidity rules, and forfeits a cost-of-capital premium the company listed to capture. The spread cost is not trivial: an Amex study found over-the-counter stocks carry relative spreads nearly triple those of exchange-listed names. My founder's beautiful register was a genuine liability on the days it mattered. The essay should have said that a quality base and a tradeable float are in tension, and that the tension has to be managed, not wished away.

The second concession is the choice point. Stout's dual-class evidence lands. You do not fully choose your base. Where a control block sits on the register, which in Asia is most of the time, the float is a residual, and the holders of the float are taking the structure as given. The Bedrock title's confidence ("the one you chose") needs the qualifier the essay omitted: you choose the character of the holders you can attract into the float, not the size of the float itself.

But notice what neither concession touches, because this is where the hold is earned. The counter-case proves that diversification delivers liquidity and lower cost of capital. It does not prove that diversification and quality are opposites you must choose between. That framing is a false binary, and the literature, read carefully, does not actually assert it. Bragg's diversification ideal is about not letting any single block grow so large it dominates and dries up the market. Cunningham's quality ideal is about who, within the register, anchors it through cycles. These are answers to different questions. One is about the shape of the float: how many hands hold it, how deep the market is. The other is about the character of the holders: whether they price on the business or on the quarter. You can have a deep, liquid float and still have it anchored by a curated set of concentrated, long-duration holders. Indeed the Bedrock essay's own recommendation said exactly this: add "a small, deliberately curated set of additional concentrated long-duration holders" without diluting the existing quality. That is not anti-diversification. It is diversification with a thesis about who you want at the top of the register.

So the synthesis is both/and, not either/or. Choose the register's character through the clientele effect: your communication, disclosure, and capital-allocation choices genuinely do attract some holders and repel others, and that part of the Bedrock thesis survives the cross-examination intact. Then size the float for liquidity, deliberately, so the cost-of-capital premium and the volatility buffer the practitioner literature documents are not sacrificed on the altar of a tidy register. The endogeneity claim holds. The breadth-is-the-goal claim holds too, for the part of the problem it actually answers. They are not competing; they are operating on different variables.

Choose the character. Size the float.

The Asian angle is where this stops being theoretical and becomes the central design problem. The Bedrock essay noted, correctly, that most Asian registers are already concentrated by default: family stakes, state holdings, founder positions. That is exactly why the "you can't fully choose" point is most live here, and why Stout's evidence reads almost as a description of the regional market. An IDX, Bursa, or PSE issuer with a sixty- or seventy-percent control block does not get to choose its base in the Bedrock essay's strong sense. The block is the base. The choice that remains, and it is a real one, is over the residual float: who you cultivate into it, and how wide you make it.

That reframes the regional IR job more usefully than either essay alone. It is not "dilute the concentrated base toward an international-looking register," which the Bedrock essay rightly rejected. Nor is it "cultivate quality and ignore liquidity," which the crowded-register literature rightly warns against. It is: take the control block as given, then engineer the float to do two jobs at once. Deep enough to earn the liquidity premium and buffer local volatility, and anchored by a curated set of long-duration holders who hold through the bad quarters. For a controlled Asian company, free float is not a number to minimise to the exchange's floor. It is the one lever of register design management actually controls, and it has to be set with both liquidity and quality in view.

So the register can be crowded. It should be, in the part of it that trades. The Bedrock essay was right that the character of your holders is endogenous to how you communicate, and it was incomplete because it forgot that the float beneath those holders has to be deep enough to trade. Choose the character. Size the float. The crowded register and the quality register were never the same shareholders, and a company that understands that gets to have both.