Osh Trader

Weekly COT & positioning reads across 26 commodities. Macro narratives, vol surface, pre-earnings. One-man research desk.

[Osh Trader @Osh_Trader]

Why the Compass and Osh FX Matrix run on different inputs:

https://t.co/G5c0H8SyOy

[Osh Trader @Osh_Trader]

@JKempEnergy 'Cautiously' is the story. Post-2020 shale defends cash flow over growth — spikes no longer trip the old drilling reflex, and rigs lead output by ~9 months. The supply response lands after a war premium can fade: they're drilling a spike they expect won't last.

[Osh Trader @Osh_Trader]

@elerianm The leg that's rising is the lagging one. Eurozone services inflation tracks negotiated wages — backward-looking, set by last year's settlements. A services-led uptick is the weakest case against a cut: the ECB eases on momentum, not last year's wage round.

[Osh Trader @Osh_Trader]

@KobeissiLetter A regional barometer up +13.5 isn't 'recovery accelerating.' Even the national ISM at 54 sits on input prices near 4-yr highs and a 32nd straight month of manufacturing job losses. Price-led expansion is exactly what keeps the Fed boxed in.

[Osh Trader @Osh_Trader]

@mikemcglone11 Stretched managed-money longs are potential energy, not a trigger. In grains the unwind needs a catalyst — weather, USDA, demand — not just a crowded book. Positioning marks the fuel; fundamentals light it.

[Osh Trader @Osh_Trader]

@Geiger_Capital ISM is a diffusion index — it counts how many firms grow, not how fast. 54 = better breadth, not surging output. And manufacturing is ~10% of GDP; the cycle lives in services and hard data. Real print, narrow signal.


[Osh Trader @Osh_Trader]

@Macrobysunil DXY isn't "the dollar" — it's ~58% EUR, often just the euro's mirror. And its correlation with VIX flips with the shock type: USD as haven one regime, as the risk asset the next. One cross-asset sign isn't "the truth."

[Osh Trader @Osh_Trader]

@robin_j_brooks Intervention buys time, not a floor — that's the rate gap plus structural outflow. Any "new playbook" collides with debt-service math. The binding constraint is fiscal, not FX.

[Osh Trader @Osh_Trader]

@rektfencer Valuation percentile is a 10-year return prior, not a timing signal. The same reading in 1997 ran three more years higher. And a rich multiple can de-rate through the denominator — earnings catching up — not only price down. "Most overvalued ever" tells you little 12 months ahead

[Osh Trader @Osh_Trader]

@CryptoTice_ Action count is breadth, not stance. 31 cuts vs 12 hikes is still net easing — "smallest gap since 2023" is a second-derivative artifact of normalization, not a tightening pivot. A few EM outliers flip a 3M rolling tally. The regime is set by the marginal large CB, not headcount.

[Osh Trader @Osh_Trader]

@BullTheoryio $377B, one session, one "AI hardware" story — but that's two businesses. GPU compute is secular capex. Memory & storage is a commodity cycle: DRAM/NAND spot, supply discipline, inventory. Price the cycle like a secular trend and the cycle eventually reminds you what it is.

[Osh Trader @Osh_Trader]

Why the Compass and Osh FX Matrix run on different inputs:

https://t.co/G5c0H8SyOy

[Osh Trader @Osh_Trader]

A reader asked why I don't run one scoring engine across every market.

Because positioning in commodities has a long half-life. In FX it can die in an afternoon.

Same input, different shelf life.

[Osh Trader @Osh_Trader]

@WarrenPies Smart reframe — but the regression assumes wealth is wealth. In 2026, ~70% of HH net worth is equity + housing: unrealized, concentrated, mark-to-market. The debasement effect holds until the asset corrects. That red dot can slide left fast — and suddenly 2.6% is a problem again.

[Osh Trader @Osh_Trader]

@KobeissiLetter Same chart shape, opposite mechanism. 2008: cap-weight outperformed because large-caps fell less in a crash. Now: it outperforms because 10 names earn ~35% of index profits. Defensive concentration vs. offensive. The ratio looks identical — the signal isn't.

[Osh Trader @Osh_Trader]

@charliebilello 70 months isn't a drawdown — it's a regime change. The 40-year bond bull that made 60/40 work is structurally over. With term premium repricing and fiscal dominance baked in, the real question isn't when bonds recover — it's whether the old playbook ever comes back.

[Osh Trader @Osh_Trader]

@GlobalMktObserv Foreigners dumping Korea' is mostly foreigners rotating chip exposure — a sector trade wearing a country label, given two semis dominate the index. Outflows aren't a price verdict if domestic buyers and Value-up flows absorb them. The tell is who owns it, not where it's headed.

[Osh Trader @Osh_Trader]

@Ken_LoveTW Nikkei strength in yen terms and a weak yen are often the same trade told twice. So the flow thesis turns on what the headline hides: are foreigners buying hedged or unhedged? Unhedged, you're long Japanese equity and short yen in one position — rally and FX don't separate.

[Osh Trader @Osh_Trader]

@MikeZaccardi A contrarian sell signal isn't a timing tool — it's a froth gauge. >8 means positioning is crowded and the cushion's thin, not that the top is dated. These prints can stay pinned in the red for months while the tape melts up. It raises the cost of a shock; it doesn't schedule one

[Osh Trader @Osh_Trader]

@SoberLook Record-low implied correlation isn't 'all clear' — it's the market pricing out a systemic shock, a position not a fact. Dispersion looks free because everyone's short the same correlation. The tell isn't the low print; it's how fast it snaps toward 1 when a macro catalyst lands.

[Osh Trader @Osh_Trader]

@BullTheoryio Six charts, one trade. INR vs Dirham -10.3% ≈ INR vs USD -10.6% — Dirham's dollar-pegged, so half this list is the dollar in different hats. The -16% vs ruble and -3.8% vs yuan are those currencies' own stories. This is mostly USD strength read through the rupee.