Net P&L
−$900
387 trades, 41 days
It's probably not your strategy. It's the five days a month you can't see, the news-day stop-outs, the revenge re-entries, the one symbol quietly eating the account. Evalytics reads your broker export and writes the entire debrief for you. You type nothing.
Silver is where your discipline goes to die. Trade your morning edge, cut the midnight revenge trades, and stay flat into the news, and this account flips green.
This isn't an anonymous demo, it's my own funded account. I dropped the export in, and Evalytics surfaced the patterns I couldn't see across hundreds of fills. Losing month and all, because that's exactly when a journal earns its keep.
Every journal on the market imports your trades and then asks you to do the hard part by hand. That blank box is where the habit dies.
✕ Imports the numbers, you write every note yourself
✕ A blank "what went wrong" box on each trade
✕ Feels like homework; you quit inside a week
✕ Dashboards full of charts you never open
✓ Writes the full review for you, in plain English
✓ Names your exact leaks, with the evidence attached
✓ Done in seconds, nothing to type, ever
✓ Ends with the three things to change next week
Unlike TradeZella, Tradervue or Edgewonk, Evalytics doesn't hand you a dashboard and a blank box to fill in, it writes the review itself.
Download your trade history from your broker, one click, in whatever format it gives you.
Evalytics reads every fill and finds the patterns hiding across hundreds of trades, session edges, symbol leaks, rule breaks, and the trades that put your eval at risk.
A written journal lands in seconds: what's working, what's quietly costing you, and the exact rules for next week.
Under the hood: a deterministic analysis layer computes your stats, flags behavioral patterns, revenge trading, averaging into losers, missing stops, and matches every trade to the economic calendar. Only then does the writing get generated, grounded in your real numbers. Not a chatbot guessing.
Same account as Exhibit A, mine. I uploaded my export and typed nothing. This is the raw output, unedited.
You're down $900 over 41 days and 387 trades, profit factor 0.95. That isn't bad luck, it's a math problem. Your win rate and average win/loss sit near break-even, so this account isn't dying to bad risk-reward. It's dying to volume and venue.
You took 359 of your 387 trades in silver, that one symbol lost you $1,360. Everything else you touched, oil and BTC, was profitable. And your Asian session is carrying the entire account at +$2,508, while New York and the late-night hours quietly hand it all back.
But the behavior is the real story: 39 revenge re-entries, jumping back into the same symbol within minutes of a loss. 173 trades with no stop-loss set. And on May 14 you stacked 13 losing silver longs into one bag-held position, closed together for −$2,291. That's the stuff a chart will never show you.
The news days confirm it: on scheduled high-impact days (FOMC, CPI, NFP) you're net −$1,496; every other day you're net +$596. You don't have a strategy problem, you have a discipline-on-news-days problem.
The same engine that writes your debrief powers our research desk, short, sourced stories on how markets actually move. Prop trading, prediction markets, macro mechanics. New episodes daily on Instagram, TikTok and YouTube, each with the short film and the full written research note below.
At Jackson Hole, new Fed Chair Kevin Warsh signaled he's serious about killing inflation, and overnight the odds of a September rate hike jumped from 35% to 57%. Stocks dipped, tech led the drop. Everyone panics about the Fed, but most people don't know why a rate hike actually hurts their stocks.
It's three things at once: higher rates make companies borrow at higher cost (profits shrink), make bonds pay more (money leaves stocks), and, the deep one, make future profits worth less today (discounting), so the same company is worth less on paper. Tech falls hardest because it's mostly future profit. The Fed doesn't sell one share, it changes one number, and the whole market reprices itself.
Companies run on debt, for expansion, operations, buybacks. When the Fed pushes rates up, the interest on all that borrowing rises, which directly eats into profit margins. A company that could fund growth cheaply at low rates suddenly faces a higher hurdle for every project. Less cheap capital means slower growth and thinner earnings, and earnings are what stocks are ultimately priced on.
Stocks always compete with the "risk-free" return, what you can earn on government bonds or savings with essentially no risk. When rates are near zero, stocks are the only game in town. But when a Treasury bond suddenly pays a healthy yield, the calculation changes: why hold volatile stocks when you can earn a solid, guaranteed return? Capital rotates out of equities and into bonds, and that selling pressure weighs on stock prices.
This is the part most people miss. A stock's value is the sum of all its future profits, translated into today's money, because a dollar earned years from now is worth less than a dollar today. That translation is called discounting, and the interest rate is the key input. When rates rise, future profits get discounted more heavily, so they're worth less in today's terms. The company hasn't changed, but the math that prices it just shrank its value. This is also why growth and tech stocks fall hardest: their value is concentrated in profits far in the future, which are the most sensitive to the discount rate.
The striking thing is the Fed doesn't buy or sell a single share. By changing one number, the expected path of interest rates, it simultaneously shrinks profits, pulls money toward bonds, and reprices every future dollar in the market. That's why "don't fight the Fed" is Wall Street's oldest rule, and why a single sentence from the Fed chair can move trillions in an afternoon.
Gold hit ~$4,700 (its best month since 1999) and Bitcoin passed $80,000, both ripping at the same time. Most people see two separate wins. It's one message. US government debt just hit an all-time high of $39.7 trillion, growing roughly $7 billion every day, and a surprise Treasury move pushed the dollar to a 3-month low.
Wall Street calls it the "debasement trade": when a government borrows endlessly and its currency slowly loses value, investors flee into assets it can't print. Gold adds ~2-3% supply a year; Bitcoin is capped at 21 million by code; dollars are unlimited. The flip: gold and bitcoin rising together isn't them winning, it's the same thing measured twice, the dollar losing trust. They're not the winners, they're the scoreboard.
The debasement trade is the move into assets no government issues, gold and bitcoin, when investors worry about dollar weakness, fiscal strain, and the erosion of purchasing power. It rests on two ideas. First, that a currency loses value when its issuer runs large deficits and finances them by expanding the monetary base, a process economists literally call debasement. Second, that assets whose supply isn't controlled by that same issuer, either because they exist as physical metal or because their issuance is fixed by code, should hold real value when the currency does not.
Two things lit the fuse. US federal debt hit an all-time high of $39.7 trillion, growing by roughly $7 billion a day, an unsustainable trajectory that keeps the debasement thesis alive. Then Treasury Secretary Scott Bessent made a surprise move to double bond buybacks, which pushed long-term yields down and sent the dollar index to a three-month low. The size of the purchases was trivial relative to the market, but the signaling effect (that policymakers will suppress yields and provide liquidity to fund the debt) was powerful, and both gold and bitcoin responded.
The whole trade hinges on supply that can't be inflated away. Gold adds only about 2-3% new supply per year through mining, a pace no central bank can accelerate. Bitcoin's supply is fixed by protocol at 21 million coins, with roughly 19.7 million already mined and issuance designed to keep halving. Dollars, by contrast, can be created without limit. When trust in the unlimited asset wavers, capital rotates toward the ones that are mathematically scarce, which is exactly what "hedging against debasement" means in practice.
Not all of the recent move is pure debasement. Part of last week's bitcoin surge was a short squeeze (forced buying), and on individual days gold and bitcoin have actually diverged rather than moving as one. The real test of the thesis is whether they keep rising together around upcoming fiscal catalysts in September, Treasury's first buyback operation and new tariff dates. If they move in unison on those days, it's the debasement trade. If only the asset with more shorts moves, it was something shallower. Big names (Ray Dalio, Deutsche Bank) are backing the structural view either way.
Gold surged past $4,400 an ounce, its highest since June, driven by a weak jobs report raising Fed rate-cut bets. It caps one of gold's biggest runs ever, up roughly 95% over the past year. To most people, it looks unstoppable.
But it just flashed a warning it hasn't shown in five months: gold closed "overbought" for the first time since March 10, its first such signal in over 100 trading days. RSI hit 77, a doji candle appeared at the top, and volume thinned on the highs. Bespoke's finding: historically, once gold hits this signal, forward returns have usually been negative.
Overbought is a momentum condition, not a verdict. Gold closed a full standard deviation above its 50-day moving average for the first time in 103 trading days, and its RSI (Relative Strength Index, a 0-100 momentum gauge) hit 77, where anything above 70 is considered stretched. It means the buying has been so fast and one-directional that price has run well ahead of its own trend, historically a point where the move tends to pause or reverse, because the pool of new buyers thins and early buyers start taking profit.
Per Bespoke Investment Group, this was one of the lengthier streaks on record without an overbought reading, and their key finding is the uncomfortable part: once gold notches this kind of overbought close, forward returns have generally been negative. Add the technical tells that clustered at the top, a doji candle (a classic indecision/reversal signal where open and close are nearly equal) on August 9, and thinning volume on the highs, and you have a textbook "the move is tiring" setup.
Overbought is a risk alert, not a sell signal, and this matters. Gold's run isn't only speculative: central banks, especially China, have been buying steadily as a hedge against the dollar and against currency debasement, and that structural demand can keep price elevated far longer than a momentum signal suggests. Markets can stay overbought for a while. The signal raises the odds of a pullback; it does not schedule one.
Gold hasn't had a 10%+ correction in over two years, itself unusual and arguably overdue. Trading roughly 2.5% above its 20-day average and well above its 200-day (~$4,133), it's stretched by most measures. None of that means sell, but it means the crowd is piling in at the highs precisely as the historical math counsels caution, which is exactly the kind of gap between what feels safe and what is safe that a disciplined process is built to catch.
Gold surged to about $4,411 an ounce, up roughly 95% over the past year. The strange part isn't the price, it's the context: gold is rallying at the same time stocks sit near record highs. Those normally move opposite, gold rises on fear, stocks rise on greed. Both at the highs at once is a contradiction.
The trigger: a brutal July jobs report. The US lost 23,000 jobs when economists expected to add 80,000. A weak economy raises the odds the Fed cuts, rate cuts erode the value of cash, so money runs to gold. But the deeper signal is the contradiction itself, investors riding the rally with one hand and buying insurance with the other.
Gold and equities usually trade as opposites because they express opposite emotions. Stocks price optimism, growth, and risk appetite; gold prices fear, hedging, and the desire for something that can't be printed or defaulted on. When both climb together, the market is holding two contradictory beliefs at once: that the AI-driven rally will keep delivering, and that something is fragile enough to need insurance against. That tension, not the headline price, is the real story. It says the confidence at the top is nervous confidence.
The July report was a genuine shock, a loss of 23,000 jobs against an expected gain near 80,000. Counterintuitively, weak labor data is often bullish for gold, because it raises the probability the Fed cuts interest rates to support the economy. Lower rates reduce the appeal of holding cash and bonds (which now yield less), and combined with heavy government spending that erodes currency value over time, they push investors toward hard assets. Gold pays no yield, so it thrives precisely when the yield on everything else is falling.
This isn't only a rate story. Gold is up roughly 95% on the year on the back of record inflows into gold ETFs, sustained central-bank buying (Poland and others adding to reserves), and a broad wave of investors treating gold as protection against currency debasement. Note the important nuance: at ~$4,400, gold is recovering strongly but is not at an all-time high, its record was near $5,600 in January before a sharp pullback. This is a powerful rebound within a violent, momentum-driven market, not a fresh record.
The next inflation prints and the Fed's September decision (do rate-cut odds firm up or fade), whether gold ETF flows keep accelerating, and whether the gold-and-stocks-together pattern persists, which would confirm the market is running on fear and momentum rather than settled fundamentals.
For the first time ever, the US is about to get a real crypto rulebook, the CLARITY Act, and the industry that ran from the SEC for a decade is the one pushing hardest to pass it. This week the Senate advanced it, with a key vote set for September 15.
The flip: for years, nobody knew who was even in charge. Was a token a stock (SEC) or a commodity like oil (CFTC)? That confusion is what got Coinbase and Ripple sued. Crypto realized the enemy was never the rules, it was not knowing the rules. The catch: prediction markets give it just a 22% chance of passing in 2026.
For years, US crypto sat awkwardly between two regulators with fundamentally different rulebooks. The SEC oversees securities (stocks, bonds) under strict disclosure and registration rules. The CFTC oversees commodities (oil, wheat, gold) under a lighter regime. A cryptocurrency could plausibly be either, and no one, not the companies, not the regulators, not the courts, could say definitively which. That ambiguity is what let the SEC sue Coinbase and Ripple, and what left every builder guessing whether they were breaking a law that might not even apply to them.
The counterintuitive shift is that the industry now wants rules, even strict ones, more than it wants the old ambiguity. A clear rulebook, however demanding, lets companies plan, raise capital, list assets, and operate without existential legal risk hanging over every decision. Uncertainty is more expensive than regulation, because you can build a business around known rules but not around a coin flip. That is why Coinbase's CEO called the bill "the one-yard line" and even Goldman Sachs' CEO backed it.
It draws the SEC/CFTC boundary for digital assets, gives stablecoin issuers like Circle a federal framework, lets exchanges like Coinbase list more tokens with legal certainty, and, for the first time, creates federal legal protection for self-custodied crypto. It also closes the "DINO loophole" (Decentralized In Name Only) that let platforms fake decentralization to dodge anti-money-laundering rules while operators still held control.
Passage is far from certain. Prediction markets price roughly a 22% chance for 2026, and the vote has already slipped multiple times. That is why crypto stocks keep rallying on anticipation, Coinbase and Circle jumped double digits earlier this summer, then fading as each deadline passes. The market is trading the hope of clarity, not clarity itself, and hope has a habit of getting repriced.
When a country "prints money," most people picture a printer running. There's no printer. The central bank simply types a number into a computer, money that didn't exist a second ago, then uses it to buy government bonds, pushing that new money out into the economy where it's real and spendable.
The catch is what didn't happen: the country didn't make more stuff. Same amount of goods, more money chasing it, so each unit is worth a little less. That's inflation, not just prices rising, but the value of money falling because supply grew faster than the economy. And who pays? Everyone already holding the currency, quietly diluted.
A central bank doesn't physically print the money it "creates" in this context. It credits accounts electronically, expanding the money supply with a keystroke. This is what quantitative easing actually is: the creation of new central-bank reserves out of nothing, used to purchase assets. The money is real the moment it's typed, because in a fiat system money is simply an entry on a ledger that everyone agrees to honor.
The new money doesn't go straight to citizens. The central bank buys government bonds (and sometimes other assets) from banks and financial institutions. Those sellers now hold cash instead of bonds, which pushes money through the financial system, lowers interest rates, and encourages lending. That's the transmission mechanism: newly created money enters at the top and flows outward as credit.
The value of money is a ratio: money supply versus the amount of goods and services in the economy. Create more money without creating more stuff, and the ratio shifts, more currency chasing the same output, so prices rise and each unit buys less. This is why "printing money" and inflation are linked, though the relationship has lags and depends heavily on whether the new money actually gets spent or just sits in the financial system.
Inflation is often called a hidden tax. It doesn't take money from your account, it reduces what your existing money can buy. Savers and holders of cash are diluted; borrowers and asset-holders often benefit, because debts shrink in real terms and hard assets reprice upward. That redistribution, from savers to debtors and asset owners, is the quiet consequence of money creation that rarely makes the headline.
On July 29 the Fed held rates steady for the 5th meeting running, and chairman Kevin Warsh insisted he's serious about inflation: "There is no soft inflation target. There's only a target, and it's 2%." The bond market called the bluff. The 30-year Treasury yield spiked past 5.2%, its highest since 2007, a 19-year high, and the Dow fell over 840 points.
Here's the tell almost nobody clocks: at the same time, the 2-year yield fell. Long rates up, short rates down. The Fed only sets short-term rates. The bond market sets the long ones, the mortgage ones, and this week it tightened policy itself, over the chairman's objection.
The Fed directly controls one number: the overnight rate banks charge each other, which anchors the short end of the curve. Everything longer, the 10-year and 30-year yields that price mortgages, corporate debt, and long-dated borrowing, is set by the bond market, by millions of investors deciding what return they demand to lend for that long. Usually the two move together. This week they violently diverged, and that divergence is the entire story.
When the 30-year jumps while the 2-year falls, it is the market saying two things at once: we don't expect the Fed to hike soon (short end down), but we demand far more compensation for inflation over the long run (long end up). It produced one of the sharpest yield-curve steepenings after a Fed meeting since the mid-1990s. In plain terms, investors decided that a chairman talking tough without acting wasn't credible, so they did the tightening for him, pushing long-term borrowing costs to a 19-year high on their own.
The 30-year yield is the reference rate for the cost of long-term money across the economy. When it rises, mortgages get more expensive, car loans get more expensive, and every long-dated borrower pays more, regardless of what the Fed's headline rate says. This week the market raised the cost of borrowing for every American, over the explicit objection of the person supposedly in charge of it.
Whether September's inflation data forces Warsh to actually act, whether long-term yields keep climbing or stabilize, and whether the bond market's warning shot turns into a sustained repricing. Analysts called it one warning, not a verdict, and September is the test.
Wall Street is racing to put everything on-chain. Tokenized real-world assets tripled to $33.5 billion in a year, with BlackRock, JPMorgan and Goldman all in, and the DTCC (which clears nearly every US stock trade and holds over $114 trillion in securities) piloting it with 50+ firms. It sounds like the future arriving.
Then someone checked whether any of it moves. 56% of tokenized assets over $100k had zero on-chain activity in a typical week, more than half the market sitting completely still. And 6 of the top 7 tokens to bet on the trend lost money, ranging from -44.7% to -98.8%. The building went up. Nobody moved in.
The headline is real: on-chain tokenized value tripled to roughly $33.5 billion, held across 167 platforms by nearly a million holders. But a joint report from rwa.xyz found that of 1,289 tokenized assets worth over $100,000, only 379 recorded any transfers in a typical week. The other 910, representing more than half the market's notional value (about $32.9 billion), sat completely still. Only around 10% of tokenized RWA value actually flows into DeFi. Tokenization has quietly won the issuance battle, minting a compliant token is now nearly a commodity, but usage (real distribution, redemption rails, secondary liquidity) remains mostly unsolved.
This isn't a crypto-native experiment. The DTCC, the plumbing behind nearly all US stock settlement and custodian of over $114 trillion in securities, is running a pilot with more than 50 firms including BlackRock, Goldman Sachs and JPMorgan, with a possible commercial launch by October 2026. BlackRock's tokenized Treasury fund BUIDL (~$2.5 billion) became tradeable on Uniswap in February. The infrastructure is being built by the most serious money on earth, which is exactly why the emptiness underneath it matters.
For anyone who tried to bet on the trend directly, the tokens were brutal: 6 of the top 7 RWA project tokens posted negative returns from January 2025 to March 2026, ranging from -44.7% to -98.8%. The market growing and the tokens that represent it are two different things, the same lesson that keeps recurring on this desk: being right about a trend and right about the instrument are separate bets.
Whether the DTCC pilot converts to a live October launch, whether tokenized Treasuries (the one genuinely active category) pull the rest of the market into real usage, and whether on-chain activity starts closing the gap with issuance or stays a ghost town with a skyline.
South Korea's Kospi fell 10.8% in a single day, its worst since the war began. Samsung dropped 13.4%, its worst day in almost 20 years. SK Hynix, the memory supplier inside Nvidia's chips, fell 14.7%. The index is now down over 30% from its record high in just 25 trading days.
The twist: this isn't AI failing. Samsung's quarterly profit just beat both Nvidia and Apple. It's crashing because the spending got too big to trust, AI investment hitting $870 billion this year, up 77% in twelve months, and investors panicking about whether it can ever pay off.
The clearest tell that this is a sentiment event, not a demand event: Samsung posted a quarterly profit that surpassed both Nvidia and Apple, and the stock fell anyway. When record results aren't enough, the problem was never the fundamentals, it was the price expectations were already set at. Analysts described it as the despair phase of a selloff, where investors rush for the exit because the tape says so, not because anything changed in the business.
JPMorgan flagged that AI-related spending is set to reach roughly $870 billion by year-end 2026, up 77% from a year earlier, with hyperscalers like Amazon, Meta, Microsoft and Alphabet accounting for about $750 billion of it. Combined with Nvidia's fresh round of AI deals worth over $750 billion, the sheer scale triggered a fear that the demand is artificially inflated, that so much money is being spent it can't possibly earn a return. The boom got so large it scared its own believers.
Samsung and SK Hynix together make up nearly half the Kospi, and both are among the world's largest suppliers of the high-bandwidth memory that AI servers depend on. That makes the Korean market a pure-play proxy for AI hardware sentiment, so when the mood turned, the whole index cracked with it. Japan's Kioxia fell 18.3% and Taiwan's MediaTek fell about 10% on the same fear.
Whether the megacap hyperscalers confirm or cut their spending forecasts in upcoming earnings, the direction of memory pricing, and whether this was a sentiment flush or the start of a genuine repricing of the entire AI-capex trade.
For two weeks, oil went straight up, more than 25% on the Iran war, Brent touching a two-month high near $102. Everyone braced for $100 oil and a fresh inflation shock. Then Friday it reversed: WTI down ~4% in a session, gold's safe-haven bid fading from ~$4,140 back toward ~$4,050.
The part nobody says out loud: the Strait of Hormuz never actually closed, the US military kept it open the whole time. Oil didn't spike on lost supply. It spiked on the fear of lost supply. And when the fear cracked, oil fell even as the war kept going.
The two-week rally traced to a single escalation: Iran-backed Houthis attacked two Saudi oil tankers in the Red Sea, widening the disruption beyond Hormuz. But actual barrels never stopped flowing, US Central Command kept the strait open through 13 consecutive nights of strikes. What moved was the risk premium: traders pricing the probability of a supply cut that hadn't happened. That's the tell that oil in a geopolitical spike is a fear gauge, not a supply gauge. The price measures how scared the market is, not how many barrels are missing.
The reversal came just as fast: reports of Chinese diplomatic movement eased the worst-case fears, and a 2.0 million barrel build in US crude inventories reminded everyone supply was still ample. WTI gave back half of a 6% single-day rally in one Friday session. Gold confirmed it, its safe-haven premium fading from ~$4,140 toward ~$4,050 as the panic cooled. Even after the drop, oil was still up ~10% on the week: enormous two-way volatility, which is exactly what a fear-driven market looks like.
Here's what makes it more than a story: even with oil falling, markets still price roughly an 80% chance of a Fed hike in September, with the FOMC meeting July 28-29. The oil scare already did its damage, it reset inflation expectations upward, and that doesn't un-reset just because the price came back down. The fear passed; its consequence didn't.
Hormuz transit volumes (actual barrels, not headlines), the July 28-29 FOMC decision, and whether oil's risk premium keeps draining or the next escalation re-arms it.
Q2 revenue came in at $16.13B, up 25.4% year over year, against a ~$14.4B estimate. Adjusted EPS landed at $0.42 versus roughly $0.19 expected, more than double. A clean beat on every line, in the middle of one of the more dramatic turnarounds in recent memory.
The stock is down roughly 28% on the month. Because markets don't price good news, they price news relative to what was already believed. After a +317% twelve-month run, a beat was the baseline, not the surprise.
A stock price isn't a scoreboard for how a company is doing, it's a bet on what everyone already believes. Intel entered this print up more than 300% over twelve months, and that run wasn't paying for the past. It was pre-paying for a future where the 18A node ships, the foundry business wins real external customers, and the restructuring works. By reporting day, "the comeback is working" was the assumption, not the news. The market pays for the delta between reality and expectation, not the level of reality, which is the same reason a struggling company can post a loss and rip 20% higher when the loss is smaller than the funeral everyone had planned.
Nothing in the quarter says the strategy is failing. The 18A node is in production, and Apple and Microsoft have joined as early design partners on the foundry side, the thing Intel needed most, a credible outside customer base for its manufacturing. The cost was brutal: roughly 15% of the workforce. The company is executing; the stock had simply already been paid for that execution in advance.
Being right about a company and being right about its stock are two different bets. You can nail the thesis, watch the fundamentals confirm it, and still lose money, because the price you paid already contained your thesis. Entry price isn't a detail; it is most of the trade. That gap between a correct call and a profitable one is exactly the kind of pattern a written debrief surfaces and a P&L chart hides.
Whether foundry converts design partners into volume commitments, the direction of gross margins, and whether July's drawdown was sector rotation or a repricing of the entire AI-capex trade.
A fifth of the world's oil passes through the Strait of Hormuz, and the strait is under attack. Brent crude is above $88, a one-month high. Oil is inflation in a barrel, so inflation won't fall, so the rate cut everyone waited for never comes. Cuts are now priced near zero, with roughly a 78% chance of a hike by September.
The old rule says war means the Fed cuts. That rule assumed conflict makes the economy weak. It doesn't hold when the war is the inflation, a supply shock, not a demand shock. Same word, opposite trade. FOMC decides July 29.
Oil isn't one line in the inflation basket, it's the input to freight, plastics, fertilizer, and the cost of moving anything anywhere. When a supply shock lifts crude, inflation expectations climb in a way no central bank can dismiss as transitory a second time. Hormuz transits have fallen sharply as vessels are targeted, and Brent's move above $88 flows straight into the price level. That is why the cut disappeared and hike odds sit near 78%.
"War means easing" was written for conflicts that arrive as demand shocks, fear freezes spending, growth stalls, the central bank cuts to catch the economy. It quietly assumed the war made the economy weak. This war makes energy expensive instead: a supply shock, where the policy that cushions growth makes inflation worse. The same headline flips from a reason to cut into a reason to hike.
In a single week, the odds of a hike at the July meeting swung from 45% before the CPI print, to 10% after it, and back to the mid-30s. That volatility isn't noise around a known answer, it is the answer. The committee is split, several officials projecting hikes while others argue the opposite, so every data release becomes a referendum on which faction wins.
The July 28-29 FOMC meeting, oil's behavior around Hormuz transit volumes, and the next inflation print. If crude falls back, the chain unwinds and the cut returns to the table. If it doesn't, the Fed hikes into a war, and every asset priced off "rates go down eventually" reprices.
Silver just printed a fresh 2026 low, during an escalating war that, by the textbook, should be lifting it. It doesn't, because silver gets hit twice: rising rate-hike odds punish the monetary half, and growth fear kills the industrial half. Two exits at once, in the same tape.
And I know this metal personally. 359 of my 387 trades were silver, it cost me $1,360. The market didn't beat me; my own patterns did, and the journal caught every one. That's the whole pitch: drop your export, read the debrief, see your silver.
Silver carries a split identity: roughly half its demand is monetary (a precious-metal store of value) and half is industrial (electronics, solar, medical). In a risk-off tape driven by rate fears, the monetary half bleeds like gold, a non-yielding asset loses to rising real yields. But silver also absorbs an industrial blow that gold never feels: if war-driven inflation forces tighter policy and slower growth, factory demand for silver falls too. Gold reliably outruns silver in these panics, which is why the gold-to-silver ratio climbs when the move is about rates rather than industry.
The reason this note is personal: on my own funded account, 359 of 387 trades were silver, for −$1,360 concentrated in that one symbol. The damage wasn't the market, it was concentration and averaging into losers, the exact patterns a written debrief surfaces and a P&L chart hides. That's what the instant analyzer is built to catch: symbol concentration, session edge, no-stop trades, and revenge re-entries, generated from one broker export in seconds, the file never leaving your browser.
The same three dials that govern gold govern silver's monetary half, September hike odds, real yields, and the dollar, plus one silver-only tell: the gold/silver ratio, which reads whether the next move is monetary (ratio up) or a genuine industrial recovery (ratio down).
On July 1, more than 430 tokenized U.S. stocks and ETFs, Nvidia, Tesla, Apple, SPY, QQQ, went live on Uniswap across Ethereum and BNB Chain, KYC-gated and closed to U.S. persons. The venue for tokenized value is being assembled in public, and underneath it, a toll now runs.
On Christmas Day 2025, governance switched on the fee: up to a quarter of every pool fee now buys and burns UNI. A retroactive burn destroyed 100 million tokens (~$600M) on day one. In February, BlackRock listed its BUIDL fund on UniswapX and bought UNI, a $14 trillion manager's first DeFi deal of its kind. The wild part: most of the approved toll isn't switched on yet.
Tokenized assets arrive in an assembly line: someone issues the asset legally (Securitize, Ondo), someone provides liquidity, then the asset needs a venue and plumbing to route each order. On Ethereum and the EVM world, the venue and the plumbing are increasingly the same thing, and this month, the assets started listing on it. Ondo's tokenized-equities platform, live since September 2025, has crossed $1 billion in TVL with over $20 billion in cumulative volume.
The UNIfication vote passed with 99.9%: a share of every pool's fees, 25% on low-fee pools, 16.7% on high-volatility pools, flows into a contract that can only be unlocked by burning UNI. Interface fees went to zero in the same stroke. January's early data implied a ~$26M annualized run-rate with only the first deployment wave live; the rollout has since phased across chains, with Unichain's sequencer revenue burning UNI too.
Per the desk-source memo whose milestones we verified independently: the protocol's effective take of LP fees has climbed from ~2.3% in January to ~7% today, while the approved structure reaches 17 to 20% fully deployed. What isn't live: v4 fees (hosting the tokenized-stock pools), UniswapX fees (the layer the Ondo integration routes through), fee-discount auctions, and aggregator hooks. The fee sources most exposed to the tokenization thesis are the ones still switched off, the gap between what runs and what's approved is the entire bull argument.
Aerodrome is expanding to Ethereum mainnet with a stated goal of 10 to 15% of onchain exchange volume, a direct attack on the venue share every scenario depends on. Fee capture can push liquidity toward venues that don't tax it. Two-thirds of the approved take rate is execution risk until each piece passes governance. And access restrictions cap near-term flow. The market's own verdict so far: UNI spiked as much as 40% on the BlackRock news and faded within days, it still prices current capture, not approved capture. The full 2030 scenario map is in the desk's Substack note.
On July 16, gold fell toward $4,000 an ounce, its lowest level since November 2025, while US forces struck Iranian targets and Tehran retaliated against American bases. Down roughly 24% from the January peak of $5,300, capping the worst quarter in 13 years. Everything the textbook teaches says this cannot happen.
It happened because gold's real enemy in 2026 isn't peace, it's a 5% yield. The same war that should feed the safe-haven bid is feeding it to bonds instead.
The chain runs through oil. Fresh escalation around the Strait of Hormuz, the corridor for roughly a fifth of global crude, pushed oil to one-month highs. Higher energy prices feed directly into inflation expectations, and markets responded by pricing about a 51% chance of a September rate hike. Rising Treasury yields and a firmer dollar do the rest: gold pays no coupon, so when the risk-free rate climbs, the opportunity cost of holding it climbs with it. Every dollar parked in bullion is a dollar not earning 5% in Treasuries.
"War = buy gold" was written for a world where the Fed answers conflict with easing. That was true in 1990, in 2001, in 2020. It quietly assumes falling rates, and in 2026 the assumption is inverted: the Fed's next move is more likely a hike than a cut, because the war itself is inflationary through oil. The fear channel (buy the haven) and the rates channel (sell the non-yielder) are both real forces; right now the rates channel is simply the heavier one. Same war, opposite trade.
History rhymes with the fade, not the spike. Gold popped on the 1990 Gulf invasion and gave it back within months; it jumped on Russia's 2022 invasion of Ukraine and retraced once the shock was priced. "Buy the rumor, sell the invasion" is one of the oldest adages on the metals desk. What's different this cycle is the starting point: gold entered the conflict already stretched after a parabolic run to $5,300, so there was a full year of momentum positioning waiting to unwind into the first sustained selling.
Three dials decide whether $4,000 is a bottom or a waypoint: September hike odds (a collapse toward cuts re-arms the gold bid instantly), real yields (gold historically bottoms when inflation-adjusted rates roll over), and the dollar (a softer DXY is mechanical support, since gold is priced in it). Peace headlines, counterintuitively, matter less than a single soft CPI print.
Same trader, same strategy. On scheduled high-impact news days (FOMC, CPI, NFP): −$1,496 across the month. Every other day: +$596. The strategy was never the problem, the discipline on days the economic calendar announced in advance was.
The chain, from one broker export: news day hits → volatility spikes → stops get run → revenge trades follow → the account bleeds. Five days a month, quietly wiping out four weeks of green. I didn't find this pattern manually, my journal did. 387 trades, zero notes typed by hand.
In the minutes before an FOMC decision, a CPI print or a payrolls release, market makers pull quotes and liquidity thins. Spreads widen. When the number hits, the entire rate path gets repriced in seconds, a move that would normally take a session happens in two candles. Stops placed at "sensible" technical levels sit exactly where that repricing sweeps, so they fill with slippage, at the worst prints of the day.
The financial damage is only half the mechanism. The stop-out triggers the emotional sequence the debrief flagged across this account: 39 revenge re-entries, jumping back into the same symbol within minutes of a loss, and 173 trades carried with no stop at all, the signature of a trader who has stopped managing risk and started managing feelings. On May 14 that cascade compounded into 13 averaged-down silver longs closed together for −$2,291. None of that shows on a P&L chart; all of it shows in the written record.
Roughly five scheduled high-impact days a month produced −$1,496, while the other ~15 trading days produced +$596. The edge was real and positive on ordinary days, and a handful of calendar-announced sessions consumed it entirely. The fix costs nothing: the economic calendar is public, the dates are known weeks ahead, and the highest-return trade available was being flat through them.
A special forces soldier turned $32,000 into over $400,000 in 30 days on prediction markets, 13 bets on the capture of Maduro, every one timed around classified military intelligence he saw at work. The DOJ indicted him in April 2026.
The detail that matters: unlike insider stock trades you have to subpoena, every bet was public on the blockchain, permanently. The evidence published itself, that's how he was caught. And he's not alone: a French whale netted ~$47M on the 2024 election, 77 wallets have been flagged around OpenAI launches, and the House has an open probe into both Polymarket and Kalshi.
Insider trading in equities is proven through subpoenaed brokerage records, phone logs and cooperating witnesses, slow, adversarial work. On a blockchain-settled prediction market, the equivalent evidence publishes itself in real time: wallet, size, timestamp, market, all permanent and public. Investigators didn't need to reconstruct the soldier's trail; they needed only to read it. Thirteen bets, each clustered around intelligence he encountered at work, sitting on a public ledger forever.
Event contracts occupy a genuinely gray zone: classic securities-law insider doctrine is built around corporate information and fiduciary duty, neither of which maps cleanly onto "will a foreign leader be captured." Prosecutors in this case reached for the tools that do apply, misuse of classified information, but the broader question of what counts as an unfair informational edge on an event market is largely unwritten law. That vacuum is exactly what the House Oversight probe into both Polymarket and Kalshi exists to examine.
The soldier is the indicted case, not the only case. A French trader known as "Théo" cleared roughly $47M on the 2024 US election using proprietary polling. Analysts have flagged 77 positions across ~60 wallets placed suspiciously ahead of OpenAI product announcements. And a WSJ investigation found a marketing campaign showing creators "winning" ~$900k when the underlying positions would have lost money, the trust problems run in both directions. Radical transparency cuts every way: it catches cheaters, and it exposes the platforms too.
The Big Short investor disclosed a full-sized position: ~60% Flutter (FanDuel), ~40% DraftKings. His thesis: prediction markets exploit a loophole, nationwide event contracts under CFTC oversight, paying zero state gaming taxes while sportsbooks pay heavily. His words: "the political climate will not tolerate this."
That loophole competition already cost Flutter ~65% and DraftKings ~45% from their peaks. The hedge inside the bet: both sportsbooks are building their own prediction markets, so if the loophole dies, he wins, and if it survives, he still wins.
A sportsbook operating state-by-state pays gaming taxes that run as high as ~51% of revenue in the most aggressive states, plus licensing costs in every jurisdiction. A CFTC-regulated event-contract exchange offers a near-identical product, a priced bet on an outcome, nationwide, under one federal umbrella, at effectively zero state gaming tax. That is not a marginal cost edge; it is a structurally different business. It's also precisely why the incumbents bled: Flutter roughly −65% and DraftKings −45% from their peaks as the untaxed competitor scaled.
What makes this a Burry trade rather than a simple dip-buy is the branch structure. Branch one: regulators close the loophole, "the political climate will not tolerate this," in his words, and the taxed incumbents' biggest competitive threat is neutralized. Branch two: the loophole survives, and both Flutter and DraftKings are already building their own prediction-market products, so they inherit the same tax treatment they currently envy. Heads the thesis wins, tails the hedge wins. The bet is on who ends up owning the customer, not on which regulatory outcome occurs.
Regulated incumbents absorbing their unregulated disruptors is one of the older patterns in American market structure, it is how offshore poker gave way to licensed operators and how crypto exchanges are being pulled into the regulated perimeter now. Burry is betting the pattern repeats. The open variable is timing: CFTC-versus-state litigation is live, Congress is circling, and until one of them moves, the loophole keeps compounding in the challengers' favor.
The June FOMC minutes revealed the most divided Fed in years: rates held for a fourth straight meeting, 9 of 18 policymakers wanting a hike, others pushing cuts, and forward guidance deleted entirely. The new Chair's own words for his committee: "a family fight."
Buried in the minutes: the Fed flagged the A.I. datacenter buildout as an inflation driver, electricity and tech demand keeping inflation sticky and rates high. Your chart isn't moving on technicals; it's pricing 18 people arguing.
Markets can price a hawkish Fed and they can price a dovish Fed; what they cannot price is a committee that doesn't know which it is. With 9 of 18 policymakers penciling at least one 2026 hike while others argue for cuts, every data release becomes a referendum on which faction gains ground, which is why single prints now move rates markets the way full meetings used to. Volatility isn't a side effect of the split; it is the split, expressed in price.
Deleting guidance, the statement shrank to ~130 words, a third of its usual length, is a deliberate return to pre-2008 central banking, when the Fed reserved the right to surprise. Guidance was invented at the zero bound to substitute words for ammunition; abandoning it says the committee wants optionality more than predictability. For traders the practical translation is brutal: the reaction function must now be inferred from data, not read from a script, and every anchor you used for the last decade is gone.
Buried in the minutes: ongoing demand for AI infrastructure is expected to sustain upward pressure on technology prices and electricity, and AI-driven investment strong enough to push growth above potential could make inflation more persistent. In plain terms, the committee named the datacenter buildout as a reason rates stay high, while the Chair himself argues AI is eventually disinflationary. Both can be true on different clocks: inflationary while it's being built, deflationary once it runs. The build phase is the one your borrowing costs live in.
Everyone learns "war = buy gold." Then Iran struck 85 US sites, oil jumped 5%, and gold fell ~2% with silver breaking below $60. Why: the rates channel beat the fear channel. War → oil spike → inflation fear → Fed hike odds up → higher rates kill non-yielding metals.
"War = buy gold" quietly assumes the Fed is cutting. It isn't, and that one assumption is the difference between the textbook and the tape.
Every geopolitical shock transmits to gold through two competing channels. The fear channel: uncertainty pushes capital toward havens, gold, the dollar, Treasuries. The rates channel: war lifts oil, oil lifts inflation expectations, inflation expectations lift the odds of tighter policy, and higher yields punish an asset that pays nothing. Both fired on the Iran escalation. The tape showed which one is heavier in 2026: gold fell 2.2% to $4,066 with silver breaking $60, on a day the 30-year traded above 5% and the Treasury revoked Iran's oil waiver.
Silver amplified the move because it is only half a monetary metal, roughly half of demand is industrial. A war that threatens growth threatens factory demand, so in a risk-off tape silver absorbs the haven selling and the industrial fear, which is why gold reliably outruns it in panics and the gold/silver ratio climbs. Watching that ratio is one of the cleanest reads on whether a metals move is monetary or cyclical.
"War = buy gold" worked for fifty years because conflict historically arrived alongside easing central banks. The adage never stated its own precondition. In a hiking regime the identical shock produces the opposite trade, and traders running the old playbook are, functionally, trading a different market than the one on their screen. That gap between the remembered rule and the live mechanism is where accounts quietly bleed.
New episodes daily, find @evalyticsx1 on Instagram, @evalytics0x100 on TikTok, and Evalytics on YouTube.
The big journals are powerful, and heavy. Connect your broker, learn the dashboards, type your own notes. Evalytics does one thing, and does it without any of that.
They link your live broker login and sync everything. Evalytics just reads a file you export, it never sees your login, your funds, or your open positions.
They hand you six tools and fifty reports. Evalytics hands you one thing: a written debrief you'll actually read, in seconds.
Not a generalist stock platform with a prop feature bolted on. Every debrief speaks your language, evals, drawdown, rules, payouts.
I'm Finn. I'm 18, I trade a funded account, and last quarter I gave back $900 across 387 trades, not because the strategy was broken, but because 359 of those trades were silver, 173 had no stop, and 39 were revenge re-entries taken minutes after a loss. On one day in May I averaged into 13 silver longs and closed them together for −$2,291.
None of that showed up on a P&L chart. It only showed up when something read the export line by line and wrote down what actually happened. I couldn't find a tool that did it, so I built one, first for me, now for anyone who's tired of being told to journal and never doing it.
Start free. Upgrade when you can't trade without it.
Send your trade export and I'll write your first debrief by hand, free, while the app is in beta. No card, no catch, I just want it in front of real funded traders.