Back to index

Four-Round TA Synthesis: Risk-On Reclaim, But Buy The Retest

This memo synthesizes the completed four-round recursive TA run over [[BTC]], [[SPY]], [[QQQ]], [[GLD]], and [[tlt]]. It is a 10-bar tape memo, not a macro or earnings memo. The latest factor snapshot available to this workflow is dated 2026-04-25 and is mostly unavailable across rates, credit, dollar, commodity, volatility, breadth, and liquidity inputs, so the memo cannot honestly claim fresh macro confirmation.

Part 1: Howard Marks Voice

The important change in the four-round loop is not that the final ranking became QQQ > SPY > BTC > GLD > TLT. The important change is how it got there. In round 003, the book had become tactically short [[SPY]] and [[QQQ]] into rejected resistance, long [[GLD]] from an accepted support band, wait-only in [[BTC]], and wait-only in [[tlt]]. By round 004, after the market opened, the old equity-index resistance fades had been invalidated: QQQ reclaimed 668.90, SPY reclaimed 719.79, BTC accepted the prior 77493.67 trigger, GLD accepted 425.46 but stayed balanced, and TLT remained boxed.

The chart layer therefore supports a risk-on reclaim, but only conditionally. QQQ is the best expression because the post-open thrust cleared 668.90, the read names a round-number air pocket toward 680.00, and the trade has a defined reload at 668.90 with invalidation below 664.51. That is not the same as saying "buy QQQ here." The trader pass explicitly waits for a 668.90 retest hold because current price had already moved to 674.97, shrinking the chase reward relative to the reload level.

SPY confirms the same direction, but with less asymmetry. It reclaimed 719.79, and its old base at 702.28 sits well below the tape, but its mapped upside was already near the post-open high around 724.50. The final expression is therefore smaller and conditional: wait for 719.79 to hold on retest, size only 0.5 ATR, and invalidate below 715.63. QQQ is the cleaner risk-on reload; SPY is confirmation and secondary expression.

BTC is the psychologically tempting trade because it finally did what the first three rounds demanded: it accepted 77493.67. But its next obstacle is also the cleanest contrary evidence in the whole memo. The reader says BTC is still below the prior daily high at 79523.00, and the adversary's rival argues that a 1.0 ATR trigger-retest long is vulnerable to range-top distribution because the target remains rejected while the stop is close. That is not a reason to ignore BTC; it is a reason to demote it below QQQ and SPY and require the retest to prove itself.

GLD and TLT are what make the conclusion more precise. GLD resolved the prior sizing objection by accepting 425.46, but the final trader correctly refuses to turn that into a current long because upside to 428.22-428.52 is smaller than the support-band risk to 422.76. TLT is even less expressive: support at 85.55 is accepted, resistance at 85.90 is unresolved, and 86.37 is rejected. There is no structural downside target below 85.55, so a short would be anticipation rather than expression.

The synthesis is therefore conservative but not neutral: the tape says risk-on won the open, defensive expressions failed to become better trades, and the right implementation is to buy reloads rather than chase extensions. The key tripwires are also mechanical. QQQ below 664.51 breaks the best risk-on expression. SPY below 715.63 makes the secondary confirmation suspect. BTC below 76852.56 turns trigger acceptance into a failed upper-range push. GLD below 422.76 breaks the support band. TLT needs either 85.90 acceptance or 85.55 failure before it deserves capital.

Part 2: DeepFuckingValue Voice

Everyone wants the heroic version: "risk-on is back, smash the button." The tape is better than that. It gives a better trade because it gives a price. QQQ at 674.97 chasing 680.00 is not the trade. QQQ at a 668.90 retest hold, with 664.51 as the wrong-price line and 680.00 as the first payout, is the trade. That is roughly 4.39 points of risk for 11.10 points of target from the reload, about 2.5:1, and it is the cleanest expression in the set.

SPY is the beta cousin. It works if the same tape works, but the payoff is mediocre: from a 719.79 reload, the target is only the post-open high near 724.50, with invalidation at 715.63. That is about 1.1:1. Fine as confirmation, not the lead vehicle. If I am taking the same risk-on bet twice, QQQ gets the bigger sizing and SPY gets the smaller one.

BTC is the one that can embarrass you. The chart finally accepted 77493.67, which matters, but the old high at 79523.00 is still rejected. The adversary is right to call out the sizing problem: if the next BTC entry is treated as a full-size clean reload while the market is actually distributing under the old high, the R/R math will look beautiful right until the tape rolls back through the shelf. The only honest BTC stance is "make the retest hold first."

GLD and TLT are not shorts by default. They are just not good longs here. GLD accepted the level people wanted it to accept, but the next shelf is too close and the downside band is too close. TLT is a tiny box between 85.55 and 85.90. The money is not made by pretending every chart has to be traded. The money is made by ranking the expressions and refusing the bad ones.

So the answer is simple: QQQ is the lead reload, SPY is smaller confirmation, BTC is conditional and adversary-flagged, GLD is wait-for-better-price, TLT is no-trade compression. This is not a bearish memo. It is a memo against paying bad prices for a bullish tape.

Frame To Expression Check

Thesis frame: post-open risk-on reclaim with defensive assets failing to offer superior immediate expression. The expression that matches is not "long the basket"; it is conditional reloads in QQQ first, SPY second, BTC only after retest proof, and flat GLD/TLT. This matches the recursive-TA decision chain because QQQ and SPY flipped from round-003 tactical shorts to round-004 retest-hold longs, while GLD flipped from full-size long risk to flat and TLT avoided an unsupported downside target.

Contradictions: BTC is the main one. The support for BTC is accepted trigger behavior; the contradiction is rejected overhead at 79523.00 and adversary range-top distribution risk. The memo keeps BTC active but demotes it below QQQ/SPY and requires a retest hold, which is the smallest expression that still respects both sides of the evidence.

Tripwires

Stance Ledger