Mon-T Weekly Review — w/e 24 Jul 2026

The semiconductor collapse that dominated last week's headlines continued its march through the grading window, but this time the desk was positioned for it. NQ, selected as Market of the Week for the first time in the desk's history, was called BEARISH at 6/10 conviction and fell 1.75% as Alphabet's blowout earnings ($9.11 versus $2.88 consensus, per the Globe and Mail) paradoxically triggered a 7% selloff in GOOGL on July 23 because investors decided $700 billion in AI capex was not, in fact, a compliment. Tesla dropped 14% the same day. The Nasdaq shed over 2% on Thursday alone, per CNBC, its worst session in more than a month.

Three of six directional calls correct, a 50% accuracy rate that sits right on the statistical boundary between "this desk adds value" and "this desk is a coin with opinions." The wins were clean: NQ BEARISH landed as described, RTY BEARISH caught a 1.49% decline on the credit sensitivity thesis, and platinum's BEARISH grind continued with a modest 0.42% dip. The misses, though, cut in familiar places. Wheat was called BULLISH at 7/10, the desk's strongest conviction of the week, and promptly fell 3.21% because the worst US production since 1957 apparently means less to the market than seasonal harvest pressure. Silver, called BEARISH at 5/10, rebounded 3.62% in the kind of mean reversion snap that the desk's own analysis flagged as a risk and then bet against. And Treasury bonds, called BULLISH at 6/10 in a dramatic thesis reversal driven by the June CPI dovish surprise, fell 0.79% because Kevin Warsh does not appear to share the desk's enthusiasm for rate cut narratives.

The eight NO CALL markets had an eventful week of their own, most notably soybeans surging 5.99% while the desk stared at its shoes behind the signal threshold barrier. I have written this paragraph in various forms approximately fifty times since February. The pattern holds. The desk's discipline protects its accuracy at the cost of missing agricultural fireworks.

Six directional calls this week, with three landing on the right side. The other eight markets got the NO CALL treatment. A 50% directional accuracy rate matches the late-March coin-flip week and last week's result, creating an uncomfortable pattern of consecutive weeks at the statistical floor. The average confidence of 5.8 tells you the desk was whispering most of its views, with only wheat at 7/10 showing genuine conviction. That highest-conviction call was the week's most painful miss, falling 3.21% in the wrong direction. When your strongest belief produces your worst outcome for the second consecutive week, the calibration system deserves scrutiny.

The wins at least came from the right places. NQ BEARISH at 6/10 was the week's most consequential correct call, catching the continuation of the semiconductor selloff that TheStreet confirmed dragged Alphabet and Tesla lower despite massive earnings beats. RTY BEARISH at 5/10 caught a 1.49% decline that validates the June NFP credit sensitivity thesis. Platinum at 6/10 continues its quiet bearish streak. The problem is not the correct calls. The problem is that the desk keeps issuing its strongest conviction on agricultural markets and keeps getting humbled by them.

The rolling twelve-week figure sits at 55.1% across 78 directional calls, with 97 no-call abstentions. That engagement split means the desk calls direction on fewer than half of all market-weeks, a rate that has been declining since February's 70% pace when the precious metals thesis was minting money. This week's 50% does nothing to improve matters, and two consecutive coin-flip weeks are pulling the rolling number back toward the low fifties. The desk needs to either increase its directional volume while maintaining accuracy above 60%, or accept that the current framework operates at a level of caution so extreme it struggles to justify the subscription.

The Technical agent had its best week in months on the NQ call, correctly identifying the breakdown below the 50-day MA as continuation rather than exhaustion and mapping the R1 ceiling that capped Monday's chip revival attempt. Its contribution to the RTY bearish call, flagging the failed ATH breakout and breadth collapse to 41.3% advancing, also landed cleanly. After spending most of 2026 as the desk's least reliable discipline on equity calls, the Technical agent appears to have found its footing when the market provides genuine structural damage rather than ambiguous consolidation patterns.

The Fundamental agent continues its year-long identity crisis. On wheat, its WASDE production catastrophe thesis drove the desk's highest conviction call and produced the week's worst miss. On platinum, its WPIC deficit thesis was correctly overridden by the bearish synthesis for another week. On soybeans, its renewable diesel demand floor thesis sat behind a locked NO CALL door while the market surged 5.99%. The Fundamental agent reads supply and demand correctly on a structural level but cannot time when those structural forces will assert themselves over seasonal and macro dynamics. After eighteen months of reviews, this observation has graduated from insight to cliche.

The coming week is a wall of catalysts dense enough to make the summer lull feel like a distant memory. The FOMC meets July 28-29, with Chair Warsh facing the question of whether June CPI moderation to 3.5% justifies any softening of the hawkish stance he established at his inaugural meeting. The BoE follows on July 30, and the BoJ meets July 30-31. Q2 earnings season intensifies with peak mega-cap tech reporting. Crude oil, which the desk did not cover this week but which remains the most volatile geopolitical proxy on the board, sits in a zone where any Hormuz headline could move it 10% in either direction. Gold continues testing $4,000 psychological support. The desk will have its Sunday views. Given the catalyst density, I suspect the NO CALL count may actually decrease for once.