Direct Derek - Public Market Edges

Form 4s Without a Press Release

2026-07-29T00:00:00+00:00

Three filings land in EDGAR the same morning. A director receives a scheduled restricted-stock grant. A CEO exercises vested options. A CFO uses personal cash to buy shares in the open market. A screen aggregates all three under one headline: insider buying.

The label is generous. A disciplined analyst wants to know what each person actually gave up.

Form 4s are useful because they are mandatory, arrive quickly, and may be the only fresh ownership event for months in a neglected company. A personal open-market purchase does not require a press release or create a Regulation FD obligation — that rule governs selective disclosure by the issuer to market professionals, and an insider's own purchase is simply a routine EDGAR filing. It appears in the system, mixed among grants, exercises, planned transactions, and administrative debris.

"Acquired" does not necessarily mean "bought." Compensation has a habit of arriving dressed as conviction.

Start With the Transaction</h2>
Section 16 covers directors, officers, and beneficial owners of more than 10% of a class. Form 3 establishes the initial position, generally within ten days of becoming an insider. Form 4 reports covered transactions within two business days of the trade date — a hard deadline with no grace period, and lateness is disclosed in the company's proxy. Form 5 handles certain transactions reported annually.
That machinery shows when ownership changed, and nothing at all about why.
The trade date and filing date also describe different moments. A Monday purchase reported Wednesday is Monday's decision, not a fresh opinion formed Wednesday morning. In a heavily followed stock, that distinction barely survives contact with the market. In a micro-cap with no analyst waiting for the filing, there may be no natural buyer assigned to notice.
So reconstruct the event. What security changed hands? Was cash paid? Was the transaction discretionary? Did total exposure rise? Three directors receiving identical grants on the same date usually reflects one compensation decision, not three independent conclusions about valuation.

The Sieve</h2>
Run every reported acquisition through five questions before it gets anywhere near an order.
First, what is the economic transaction?
Open-market purchases, equity grants, option exercises, compensation-driven acquisitions, 10b5-1 transactions, and administrative transfers can all appear on the same form. They do not carry the same information. What matters is the insider choosing to increase exposure, not merely receiving what the compensation committee already approved.
Second, how much personal capital is at risk?
Cash paid in an open-market purchase is the cleanest evidence of discretionary intent. A transaction under a 10b5-1 plan may still reflect a prior decision, but the relevant discretion was exercised when the plan was adopted, not necessarily on the reported trade date. A modest purchase made with personal funds can say more than a much larger grant.
Third, is the amount meaningful?
The dollar figure needs denominators: compensation, existing ownership, previous awards, and transaction history. There is no universal threshold. A small purchase may be a cheap gesture designed to be noticed. A meaningful one leaves the insider materially more exposed to being wrong.
Fourth, is this a transaction or a pattern?
Use Form 3 as the opening balance and turn subsequent Form 4s into a ledger: trade date, filing date, transaction type, cash committed, shares held afterward, and whether a plan was involved. Then strip away grants and mechanical exercises to see whether net exposure is actually rising.
An annual award followed by routine sales is compensation administration. Repeated purchases that increase an already meaningful stake deserve more work.
Fifth, who is buying?
A senior operating executive using personal cash generally deserves more attention than a director making a token purchase. Proximity to the business raises the signal's potential value, but it also raises the standard of proof. Incentives, compensation, liquidity needs, and prior behavior still have to be understood.
After that work, the filing becomes less exciting and more useful.

What the Historical Evidence Says</h2>
The evidence is asymmetric. Insider sales can reflect taxes, diversification, or personal spending. Purchases have fewer routine explanations.
Lakonishok and Lee, studying insider trades from 1975 through 1995, found that predictive power was concentrated in small, less-followed firms. Insider purchases in small-cap companies were associated with something on the order of 7% abnormal returns over the following twelve months. Jeng, Metrick, and Zeckhauser later found that insider-purchase portfolios earned abnormal returns exceeding roughly six percentage points annually, while insider sales showed no significant abnormal return.
Those are historical sample results, largely from before 2000, specific to their samples and windows. They are not expected returns for the next Form 4 that appears in EDGAR, and they do not rescue a weak business or an expensive stock.
What they establish is narrower: insider purchases have historically carried more information than sales, with the effect strongest where outside attention is scarce. That indicates where ledger work may be worth the hour.

A Useful Signal With Very Little Capacity</h2>
The setup worth wanting is deliberately inconvenient: a current reporting issuer, a senior insider spending personal cash on an open-market purchase, a material increase in an existing stake, little outside coverage, and enough liquidity for a small position but not enough for institutional size.
The capacity limit is part of the signal's survival.
A fund cannot extract much value from an opportunity that only supports a position too small to affect its portfolio. If it buys enough shares to matter, it can move the price and surrender the edge through spread and impact. Research on post-earnings-announcement drift shows the same liquidity gradient, and the magnitudes are worth reading as illustration rather than as a promised spread: in one study's extreme deciles, the strategy returned roughly 0.04% a month among the most liquid stocks against about 2.43% a month among the most illiquid — while transaction costs consumed somewhere between 70% and 100% of the paper profit.
The anomaly persists where harvesting it is expensive.
Before acting, check the spread, effective float, median and average daily volume, and the insider's purchase relative to normal trading flow. Estimate how many days entry might take, then perform the less cheerful calculation for the exit. If the stock trades OTC, check its SEC filing history and quotation status separately from the venue label. An issuer that stops making current public information available can lose eligibility for a published public quotation under Rule 15c2-11 and become materially harder to trade, which turns a slow exit into no exit without anyone announcing it.
A clean insider purchase cannot compensate for disappearing disclosure.

The Final Check</h2>
Reduce the filing to a short list:

Was it an open-market purchase made with personal funds?</li>
Was it discretionary rather than compensation or a previously arranged plan transaction?</li>
Is the size meaningful relative to compensation and existing ownership?</li>
Is total exposure rising across multiple filings?</li>
Is the issuer current in its reporting?</li>
Can a position be entered and exited without spending the expected edge on spread and impact?</li> </ul>
Then underwrite the company. The balance sheet, valuation, dilution risk, governance, and operating results still have to work. Form 4 activity merely indicates where informed capital may be moving before anyone bothers to narrate it.
The ledger has an obvious hole in it, and the discipline above does nothing to close it. It records what an insider did and never what they knew, and an officer who is confidently, expensively wrong about their own company files on exactly the same form as one who is right. The filter separates compensation from conviction. It has no opinion whatsoever about whether the conviction is any good.
Return to the three filings. The grant costs the director no purchase price. The option exercise may reflect compensation mechanics rather than a new view. The CFO's open-market purchase commits personal capital, but even that earns only a place in the ledger.
The useful question is whether ownership is building through deliberate purchases — or whether a routine allocation happened to arrive on a form carrying the word "insider."

The Float Is Smaller Than the Funds Trapped Inside It

2026-06-27T00:00:00+00:00

It's always comforting to open a Form 13F and find respectable institutional names sitting beside an obscure micro-cap. The ownership table reads like a letter of recommendation: a pension manager here, a boutique value fund there, names that have survived a cycle or two and have the scar tissue to prove it. Informed money, gathering quietly around an asset nobody else has noticed. It's reassuring.

Then you pull the tape. The stock trades a few thousand shares on a good day. The spread charges admission. Several elephants have wandered into the room, and the only exit is a service door built for one polite person at a time.

In a mega-cap, institutional ownership tends to arrive with liquidity. In a micro-cap, it arrives as an inventory of future forced selling. The filing tells you who is in the room. The tape tells you whether any of them can leave without redecorating.

Float math is subtraction</h2>
Shares outstanding is an accounting figure. Public float is the portion in public hands. Effective tradable supply is a further judgment about which of those shares would salute if you raised the flag.
Start with shares outstanding and subtract founder, insider, controlling, and strategic blocks. Forms 3, 4, and 5 help identify insider ownership and transactions. Schedules 13D and 13G generally disclose holders above 5% of a class registered under Section 12 — which means smaller OTC issuers may produce no such filings at all. Lockups, cross-holdings, and other restrictions remove more supply.
Then subtract the shares that are legally tradable but temperamentally absent. An index fund may hold until deletion and not a day sooner. A strategic investor may have no interest in selling at any price you would enjoy paying attention to. A founder may regard the quoted market as an administrative nuisance rather than a place one transacts.
Rule 144 can slow affiliate sales further. For affiliates relying on its safe harbor, sales during a three-month period are generally limited to the greatest of 1% of the shares outstanding or, for securities traded on a national exchange or an automated quotation system, the average weekly reported volume during the four calendar weeks preceding the notice. For a thinly traded security outside those systems, only the 1% test is available. It is not a prohibition on every conceivable transaction, but it makes the point: a block can be entirely real, worth a great deal on paper, and unavailable to the market for years. Wealth and access are not the same word.
Data vendors publish float estimates. Their classifications differ and their assumptions rarely introduce themselves. Treat the number as a starting point, not a verdict. Screens count shares. Markets clear only the shares whose owners are willing and able to part with them, which is a smaller and moodier population.

A hundred days of volume is not a hundred-day exit</h2>
The basic diagnostic needs three numbers:

Effective free float</li>
Each meaningful holder's position</li>
Average daily volume</li> </ol>
Divide a position by average daily volume and you get a clean, confident, misleading result. Consider a hypothetical fund holding two million shares in a company that trades twenty thousand shares per day. Its position equals one hundred days of reported volume. That's tidy.
It does not have a hundred-day exit. Selling every share that trades, every day, would make the fund the entire market — a distinction the fund would notice on the way down. At a 10% participation rate, the theoretical exit stretches to one thousand trading days, roughly four years, before you account for holidays, thinning volume, an unhelpful price, or other sellers discovering the same door at the same inconvenient moment.
The stress test:
`Exit days = Position size ÷ (Average daily volume × Participation rate)</code>`
`Run it at 10%, 20%, and 30%. Then refuse to believe any of it.`
`Average volume can be flattered by an index rebalance, a financing, a block trade, or a brief outbreak of speculation. Compare multiple windows, check median daily volume, and look at how much actually trades near the bid rather than somewhere inside a comforting historical average. Reported volume is throughput. It is not a standing invitation, and it will not wait for you.`
`The position receives a closing mark every afternoon, punctual and serene. Its realizable exit price, in size, may be a work of fiction.`
How institutions become trapped</h2> Nobody has to make a visible mistake to get stuck. An active manager can understand the business perfectly and still own more shares than the market can absorb. A specialist can know every footnote by heart and remain unable to trim without knocking over their own price on the way. An index vehicle can be cheerfully price-insensitive on the way in and legally obligated to sell on deletion — rules, not opinions. The trap becomes visible when a calendar appears: redemptions, index reconstitution, fund closure, manager replacement, mandate change, or ordinary portfolio housekeeping. Under redemption pressure, a manager usually sells what sells easily first — the reasonable instinct that quietly makes things worse. The liquid names leave, and the illiquid ones inherit a larger share of a shrinking fund. The positions that were merely inconvenient become urgent. The business hasn't changed. The shareholder's afternoon has. This is the inversion worth remembering. A position immaterial to a fund can dominate the market for the stock underneath it. The fund thinks in basis points of exposure. The order book experiences the same position as weather. Ownership data is a delayed map</h2> Form 13F applies to institutional investment managers exercising investment discretion over at least $100 million in Section 13(f) securities. It is filed quarterly and is generally due within 45 days after quarter-end. Useful, and stale by construction. The holdings reflect the quarter-end snapshot, not the portfolio on the day the filing surfaces. Form 13F also doesn't capture every institution, every security, or every economic exposure. It is a map of disclosed long positions within one reporting regime, not a live feed from the dealing desk, however much anyone would like it to be. Ownership databases add their own mischief by stitching together filings made under different rules. A quarter-end 13F, a fresher insider filing, and a 13D or 13G amendment can all describe different dates while sitting together in one clean, trustworthy-looking table. The table is clean. The dates are not aligned. Filings reconstruct structure — who owns the shares, how concentrated the positions are, what constraints may govern them. That structure is not today's order flow. Compare the filings with recent volume, insider transactions, index schedules, and any unexplained change in turnover. By the time the map reaches you, someone in it may already be halfway to the door. The small allocator's edge</h2> The opportunity appears when a seller's timetable matters more than the price. A small buyer does not have to swallow an institutional block. Small size permits limit orders, selective participation, and the patience to sit through a long liquidation without flinching. A position too small to matter to a fund can matter to a personal account. That is not only an edge in analysis. It is an edge in capacity — the rarer of the two. In a deep market, competing capital removes temporary pressure almost before you notice it. In a thin market, the very institutions capable of supplying that capital can't step in without becoming the market themselves, which they are paid specifically not to do. Capital is scarce, ownership information is imperfect, and the inconvenience keeps the opportunity safe from money that requires scale to bother getting out of bed. The fence works both ways</h2> None of this makes illiquid stocks cheap. Illiquidity is a priced risk, not a coupon for mispricing. Wide spreads and price impact are real costs, charged on the way in and again on the way out. A forced seller may hand you a discount, and you inherit, in the same motion, the exact market structure that trapped them. Congratulations. Position sizing is the whole discipline. Never buy a trapped fund's position in a size that quietly makes the buyer its heir. Market capitalization describes the building. Effective float describes the room. Daily volume describes the door. The shareholder worth watching is not the one eager to come in. It is the one who eventually must leave, through an opening everyone else has mistaken for a market.