Email: matthewthagerty@gmail.com

ANALYSIS

When Theatrical Signals Convert to Streaming (And When They Don’t) - Evidence from Netflix original films and synthetic box office controls (June 2026)

Traditional studio film finance operates in a spot market relying heavily on pre-release signals like franchise history, star power, and massive budgets to convince viewers to pay a sunk cost at the box office. Streaming video platforms, however, largely run on subscriptions where different economic rules apply: revenue is pooled across a bundle, and algorithmic distribution routes viewers to new titles at near-zero marginal cost, allowing platforms to cross-subsidize a massive portfolio of smaller, exploratory bets.

This empirical analysis asks whether the expensive signals that price theatrical risk actually translate to streaming success. By matching over 400 Netflix original films to synthetic box-office comps, I find that theatrical pedigree converts to streaming only at roughly one-fifth strength (a slope of 0.204). Because a 10-point gain in expected theatrical rank buys only ~2 points of streaming engagement rank, the priciest theatrical bets—like sequels and massive franchises—drive disproportionately high engagement, but less high than their theatrical pedigree would predict. Others genres, on the other hand, including documentaries and romantic comedies, seem to slightly outperform on Netflix. The data suggests that content underwriting under subscription logic looks largely like venture capital: a high volume of lean, original bets, with few or low-grossing theatrical comparisons, while the priciest signals, while driving disproportionate engagement, do so less on the Netflix subscription platform than their box office comps would suggest they would theatrically. Slide deck available here.

Data provided by The Numbers, via OpusData.

Caracas Stock Exchange (BVC) - Institutional Capacity Assessment (March 2026)

When is a market actually tradable? Measuring price impact and exit horizons in Venezuelan equities after the January 2026 transition, this deck finds a $5M position would take over a year to exit and carries ~2,000× the price impact per dollar of comparable Colombian stocks. This is a structural ceiling to capital market growth in Venezuela that political change alone cannot lift. Slide deck available here.

DOCTORAL RESEARCH

Dissertation: Floats, Frictions and High Frequency Fundamentals: Essays on Modern Valuation

Advisors: Samuel Hartzmark, Vincent Bogousslavsky, David Solomon

See my dissertation defense slides here.

The Subscription Float: Recurring Revenue as Corporate Financing (2026)

Subscription revenue models have come to dominate the technology, media and entertainment industries—among others—with the share of all public firms citing them rising from ~20% to ~50% over the last two decades. I find that management most frequently cites “revenue predictability” and “customer stickiness” in regulatory filings as reasons for subscription adoption.

I also document a shift in corporate finance, where subscriptions act as interest-free loans from customer to firm, paid upfront and decoupled from usage. This has significant capital structure implications: firms that adopt subscriptions hold less cash and grow faster relative to their pre-adoption baselines. The result is a valuation sensitivity to regulatory shifts: a study of abnormal stock market reactions of subscription-exposed firms around FTC regulatory events suggests that markets may price the slice of this float exposed to regulation at roughly 2% of firm value. Available on SSRN.

The Value of Velocity and the Velocity of Value (2023)

In an amusement park, a longer queue usually signals a better ride; all else equal, however, an empty queue is an ideal experience. An empty nightclub, by contrast, is always a failure. So it goes with financial assets: for some, a trade merely signals a value that already exists, and a lower price for the same asset is strictly better; for others, like Bitcoin, trading both signals value and creates it, because the network is, in large part, the value.

Using high-frequency data across major exchanges, this paper shows these competing valuation mechanisms leave a fingerprint even at the millisecond scale: the compensation for trading against better-informed players vanishes within seconds, yet price impact stays outsized in a way comparable equities never show. Available on SSRN.

The Pedagogical Price: On the Three Layers of Learning in Empirical Asset Pricing (2022)

An essay arguing that price is not a number but a process — that everything markets "know" lives in three inseparable layers of learning: Semantic, Syntax, and Structural. Surveys the contemporary academic literature on how market participants learn and how information is impounded in asset prices. Ranges from Bayesian updating to the Count of Monte Cristo to Gödel, Escher, Bach, to the strategic behavior of orcas. Available on SSRN.