The
Options
Layer

Part 1 of this investigation documented that SPY and QQQ traded at more than three and a half times their normal dollar volume on April 7, 2025 — two days before the 90-day tariff pause that ended Liberation Day. Part 2 established that no comparable reading exists anywhere in seven years of presidential trade policy events, including Liberation Day itself.

Those two pieces answer the first question: was the equity volume anomalous? The answer is yes, unambiguously, by any reasonable measure.

They do not answer the second question: was it directional? A large volume spike on a volatile day could reflect panic selling, institutional rebalancing, or two-sided chaos. It does not by itself tell you whether whoever was moving that capital expected the market to go up or down.

Options flow answers the second question. Not conclusively — nothing short of subpoena power answers it conclusively. But directionally, and with enough specificity to narrow the field of explanations considerably.

What Options Flow Measures

An option is a contract that gives its buyer the right to purchase (a call) or sell (a put) an underlying asset at a set price before a set date. Call buyers profit when the underlying rises. Put buyers profit when it falls.

The call/put ratio — total call premium divided by total put premium in a given window — is a directional indicator. When more money goes into calls than puts, the ratio is above 1.0. When more money goes into puts, it is below 1.0. In a market that has been falling for five consecutive days with no public indication of reversal, the rational position is puts. The ratio should be well below 1.0.

Market Sentinel measured call and put premium for near-ATM contracts — strikes within ±3% of the closing spot price, expiring within seven days — for both SPY and QQQ across the Liberation Day sequence. The closing price is used as the ATM anchor because it is the most publicly verifiable reference for each date; the ±3% band is wide enough that normal intraday price movement does not materially change which strikes are included. Each bar represents five minutes of trading. Premium is calculated as VWAP multiplied by volume multiplied by 100 (standard contract size).

All data is from Polygon.io's options API. All computations are reproducible from the published event list and parameters at policytorque.com/market_sentinel_article.

One instrument in the broader dataset does not appear in this options analysis: TQQQ, the 3x leveraged QQQ ETF. TQQQ has a thin options market — thin enough that its C/P ratio is not a reliable signal. But TQQQ itself functions as what might be called options-lite exposure: it gives a buyer three times QQQ's daily move, long or short, without requiring an options account, the capital requirements, or the complexity of strike selection and expiration management. On April 7, 2025, TQQQ equity dollar volume was 180% above its 5-day baseline. That is a meaningful signal in its own right — directional positioning through the leveraged instrument rather than through the options market, producing an equity anomaly rather than an options anomaly. Both are consistent with the same thesis.

Two Windows, One Day

Before presenting the sequence data, the methodology requires one clarification that is central to understanding what April 7 shows.

Market Sentinel anchors each measurement to its event. The April 7 event — the Truth Social post — is anchored at 09:30 ET with a window of 60 minutes before and 120 minutes after, capturing approximately 08:30 to 11:30 ET. This is the morning window. The April 9 event — the tariff pause announcement at 13:18 ET — uses a window of 120 minutes before and 240 minutes after, capturing approximately 11:18 to 17:18 ET. Because April 7 falls within the five prior trading days used as baseline for April 9, it appears in both datasets — but at a different time of day.

These are two non-overlapping windows on the same date. They measure different market sessions and produce different readings. Both are reported here.

In the morning, around the Truth Social post, QQQ was exactly flat at 1.00 and SPY was put-heavy at 0.53. The market processed the post and did not respond directionally. Options traders who saw "GREAT TIME TO BUY" on their screens did not, in aggregate, buy calls.

By afternoon — same day, no additional public news — QQQ had moved to 1.29 call-heavy. SPY was still put-heavy at 0.78 but had pulled significantly away from its morning reading. There was no announcement, no scheduled event, and no change in the public information available to the market between the morning and the afternoon. The post was old news. The market had already shown it did not believe it.

The Liberation Day Sequence

The afternoon window — anchored to the April 9 announcement time — allows comparison of April 7 against the full Liberation Day sequence for both instruments.

The sequence tells a consistent story except for one entry. Liberation Day itself (April 2) was call-heavy at 1.19 for QQQ and nearly flat at 0.94 for SPY — the tariff announcement came after close, and options markets had not yet fully priced in the implications. April 3 and April 4 show the market processing the reality of the trade war: put premium built steadily as the freefall deepened. Both instruments tracked the same direction. April 8 hit the lowest readings in the sequence — 0.43 and 0.36 — consistent with a market that had been falling for five straight days and saw no public path to recovery. April 9 spiked to 2.11 and 2.15 the moment the announcement hit.

April 7 afternoon does not fit. In an unbroken sequence of put-heavy readings on April 3, April 4, and April 8, the afternoon of April 7 produced call-heavy flow in QQQ and a significant pull toward neutral in SPY. No public event occurred between the April 7 morning session — when the market had already shrugged at the Truth Social post — and the April 7 afternoon session to justify the shift.

Eliminating the Alternatives

Three alternative explanations for the April 7 afternoon options signal are worth examining directly.

The Truth Social post drove retail call buying. The morning window rules this out. The post was published in the morning. The morning C/P for QQQ was 1.00 — exactly flat — and for SPY was 0.53, put-heavy. If retail investors had responded to the post by buying calls, the signal would appear in the morning window. It does not. The afternoon call buying came after the market had already demonstrated it was not moving on the post.

Short sellers covered positions, creating upside pressure. With SPY down over 20% from its Liberation Day high, short sellers covering would appear as equity buying and could drive some call buying as a hedge. But short covering would appear across all instruments, including DJT, the president's own company. DJT was below baseline on April 7 — not elevated. Short covering also concentrates at market open, not in the afternoon session. And short covering of this magnitude across multiple instruments simultaneously, with a call/put ratio shift specific to the afternoon, is not a pattern short covering produces.

A contrarian investor believed the selloff was overdone. A rational contrarian with no privileged information would have had the same thesis on April 3, April 4, and April 8 as on April 7. Market was oversold on all four days. The macro thesis for buying the dip was identical. The C/P readings on those days were 0.91, 0.57, and 0.43. The contrarian thesis does not explain why April 7 afternoon specifically produced call-heavy flow when none of the other freefall sessions did.

The Combined Signal

Parts 1, 2, and 3 of this investigation measure the same event through three separate lenses: equity volume magnitude, comparison against a seven-year control group, and directional options positioning across two independent time windows. Each lens produces a different measurement. All of them point the same direction.

Three and a half times normal equity volume. A morning options market that processed the president's post and went flat. An afternoon options market that turned call-heavy with no new public information. The president's own company below baseline while everything else surged. The highest single-instrument residual in seven years of data. Forty-eight hours before one of the largest single-day market rallies in years. The next day, with identical market conditions, the options market went back to deeply put-heavy. Then the announcement itself — on a personal platform, mid-session, the only positive trade reversal in seven years that gave the market less than three hours to react.

Each of these findings has an innocent explanation in isolation. Together, accounting for the two-window options structure, across two instruments, against three alternative explanations that have been tested and eliminated, with a fourth signal in the announcement timing itself, they are very hard to explain.

A Fourth Signal: The Announcement Window

The three prior sections measure market behavior. This one measures something different: the structure of the announcement itself.

The April 9 tariff pause was not announced at a press conference, through the USTR, or via a White House statement. It was posted on Truth Social — the President's personal platform — at 13:18 ET on a Tuesday. That detail is worth examining against the full record of Trump trade reversals.

The events database covers every major Trump tariff reversal, pause, and positive trade announcement across both terms. Setting aside the G20 Buenos Aires truce, which was announced at approximately 20:00 ET because it concluded at a dinner in Argentina and the time was not a choice anyone in Washington made, every deliberate Washington-based positive reversal in the dataset came before the market open or at the open:

Four out of four prior deliberate announcements gave the market a full session or more to react. April 9 gave the market 2 hours and 42 minutes. It is also the only positive trade reversal in the dataset announced on Truth Social rather than through an official government channel, a press conference, or a platform with broader institutional reach.

The structure of that announcement window has a specific consequence for market participants. Institutional investors — pension funds, asset managers, the firms that move capital at the scale required to influence the residuals measured in Parts 1 and 2 — operate on overnight cycles. Risk management systems price positions at close. Rebalancing decisions are made between sessions. A pre-market announcement gives every participant, informed and uninformed alike, the same overnight window to read, analyze, and position accordingly. That is how markets are supposed to work.

A 13:18 announcement eliminates that window. An institutional investor who was correctly positioned for the prevailing public information — a five-day freefall with no public indication of reversal, a C/P ratio of 0.43 on April 8, active White House denials of any pause as recently as that morning — had 2 hours and 42 minutes to unwind before the close. The market does not pause for position adjustments. At 13:18 the clock was already running, and anyone on the wrong side of the trade had no overnight to digest the news and no way to avoid the loss that followed.

The only participant structure that benefits specifically from a mid-session announcement over a pre-market announcement is one holding a short-dated long position — calls expiring same day or within days — that was acquired before the announcement and closed into the rally before bell. A pre-market announcement produces the same directional gain but eliminates the time pressure that distinguishes a pre-positioned holder from everyone else. A mid-session announcement preserves that distinction for as long as the session runs.

This is not an accusation. The President can post on his own platform whenever he chooses. There is no law specifying that market-moving policy announcements must come pre-market. What the record shows is that every prior deliberate positive reversal did come pre-market or at the open, and this one did not. The departure from pattern is documented. The question of whether it was consequential is one that trading records — available only through subpoena — would answer.

Market Sentinel measures anomalies. It does not establish intent, identify actors, or prove a crime. A call/put ratio of 1.29 on April 7 afternoon means someone bought calls. It does not mean they had advance knowledge of a presidential announcement. A +262% equity volume residual means an unusual amount of money moved. It does not tell you who moved it.

Establishing who traded on April 7, what information they had, and whether that information was obtained illegally requires access to trading records, account holders, and communication logs. That access requires legal authority — a subpoena, a formal investigation, or a congressional inquiry.

What Market Sentinel can do — and has done across three parts of this investigation — is demonstrate that the public data is consistent with advance knowledge, inconsistent with every available innocent explanation, and unreachable by any further analysis that does not have subpoena power. That is a lower bar than proof of wrongdoing. It is the bar required to justify asking the question with legal authority.

Federal securities law — Section 10(b) of the Securities Exchange Act and SEC Rule 10b-5 — prohibits trading on material non-public information. A pending policy reversal that produced a 9.5% single-day market rally is material by any reasonable standard. It was non-public until 13:18 ET on April 9.

Neither the SEC nor the DOJ has opened a public inquiry into the April 7 trading pattern. The Senate Banking Committee has jurisdiction. The House Financial Services Committee has jurisdiction. The tools to answer the question this investigation raises exist. They are not being used.

Michael Russo is the founder of PolicyTorque. Market Sentinel is a live data pipeline measuring anomalous equity volume and options flow around presidential policy announcements. All findings are drawn from publicly available market data via Polygon.io. Full methodology and event database are published at policytorque.com/market_sentinel_article. This is Part 3 of a 5-part investigation.