#8月CPI数据出炉

#每周来晒
August US CPI landed on September 11 and looked like a non-event: headline inflation up 0.4 percent month over month and 3.4 percent year over year, both in line with consensus. The detail underneath is what matters, and I think most people are reading this print backwards.

The question here is whether the data changes the Fed's rate cut path. My answer is blunt. There is no cut path to change. The market is pricing a hike, and that is the story of this week.

Question one, will CPI change the Fed's path

Yes, but not in the direction the question implies. August core CPI came in at 0.3 percent month over month against a 0.2 percent consensus, while core year over year eased to 2.4 percent from 2.5 percent in July. Energy did most of the damage: gasoline jumped 3.9 percent on the month and accounted for more than one third of the whole increase, with energy up 16.3 percent year over year and gasoline up 27.4 percent year over year. Shelter, the component that decides whether core inflation is truly sticky, rose just 0.3 percent on the month and 3.0 percent year over year, with owners equivalent rent up only 0.2 percent. Core goods rose 0.1 percent on the month and 0.7 percent year over year, so tariff pass through into goods is still nowhere near the nightmare scenario priced a few months ago.

The monthly sequence: March plus 0.9 percent, April plus 0.6, May plus 0.5, June minus 0.4, July plus 0.1, August plus 0.4. June and July looked like disinflation was winning. August reopened the argument.

That is why rate expectations moved so violently. The federal funds target range is 3.50 to 3.75 percent, and before this print the market assigned roughly 68 percent odds to a hike at the September 16 meeting. After the release that jumped to about 85.6 percent on CME FedWatch pricing, with Reuters around 82 percent and an intraday peak near 90 percent. If it happens it will be the first rate hike since 2023, not the first cut of a new cycle. The market is now pricing a possible second hike in December, and at least one major bank's base case is three hikes across September, December and March. The US 10 year Treasury yield touched 4.97 percent on September 11, the highest since late 2023, against roughly 4.20 percent at the start of the year, about 77 basis points of repricing. That number, more than the CPI print itself, is what is moving every risk asset on the board.

My read is that the report confirmed a regime rather than creating one. The Fed fears inflation re accelerating more than a growth slowdown, and a 0.3 percent core print gives it cover. What matters now is whether September is framed as one hike with a long pause, or the start of a sequence.

Question two, how crypto and stocks react short term

Equities reacted unusually. On September 11 the Dow closed at 52,573.29, up 0.98 percent, the S and P 500 at 7,656.98, up 0.86 percent, the Nasdaq Composite at 26,333.04, up 0.96 percent, and the Russell 2000 at 2,903.94, up 0.45 percent, ending a four day losing streak. Zoom out one week and it flips: the Dow lost about 1.6 percent, the S and P 500 about 0.8 percent, the Nasdaq about 0.7 percent. That was a bounce inside a downtrend, not a breakout. Oil helped, with Brent touching 109.97 dollars a barrel before closing down 2.8 percent at 104.61 and WTI off about 3 percent at 99.28. Gold finished nearly flat at 4,408.90 an ounce, a market that cannot decide whether inflation or a growth hit is the bigger risk.

Crypto is behaving more coherently than equities so far.

Bitcoin trades near 77,327 dollars, down about 0.7 percent over 24 hours and about 2.8 percent over seven days. The September 11 candle opened at 76,577 and closed at 77,228, up 0.85 percent, but the intraday range ran from 76,023 to 79,874, a swing of just over 5 percent. From the September 6 close of 80,341, Bitcoin is down about 3.9 percent. Price sits roughly 1.9 percent below its 200 day average near 78,830, and the hourly Bollinger bands have compressed into a 76,967 to 77,649 pocket. That is a market holding its breath before the Fed.

The internals tell more than price. Bitcoin open interest is around 51.9 billion dollars, down about 1.9 percent in 24 hours, with perpetual funding still mildly positive at roughly 0.003 percent per eight hour window. Slightly positive funding with falling open interest reads to me as de leveraging, not a fresh short attack, and the long short ratio near 1.165 shows positioning that is net long without being euphoric.

Ethereum is the standout. ETH trades near 2,535 dollars, and on September 11 it opened at 2,437.97 and closed at 2,516.66, up 3.23 percent, with an intraday range from 2,433.85 to 2,666, a 9.5 percent swing. Over seven days ETH is up about 3.4 percent while Bitcoin is down about 2.8 percent, a gap of more than six percentage points. The ETH to BTC ratio moved from about 0.0313 on September 6 to roughly 0.0328 now. ETH is about 2 percent above its 200 day average near 2,487 while Bitcoin sits below its own. That is a leadership change, not noise.

Flows explain why. Spot Ethereum ETFs took in 216.4 million dollars net on September 11, BlackRock's ETHA alone pulling 149 million, and the four day net from September 8 to 11 is positive 197.1 million. Total Ethereum ETF assets stand near 16.31 billion dollars with 13.39 billion of cumulative net inflows. Bitcoin ETFs went the other way: minus 13.3 million on September 11, minus 282.6 million on September 10, minus 120.2 million on September 9, roughly 416 million of three day outflow against about 97.6 billion in total ETF assets.

Solana trades near 102 dollars, up 3.84 percent on September 11 but still about 4 percent below its September 6 close, with open interest around 6.2 billion dollars. Total crypto market cap is 2.74 trillion dollars, up 1 percent in 24 hours on 102.8 billion of volume, with Bitcoin dominance at 58.65 percent, Ethereum dominance at 11.68 percent and the altcoin season index at 42, still Bitcoin led territory. The fear and greed gauge reads 68, far from panic.

There is also a security thread under the tape. About 4,000 BTC, roughly 320 million dollars, was pulled from Liquid Network's federation wallet on September 6 and the network paused operations, while Coldcard exploit losses have reached about 1,806 BTC, roughly 143.9 million dollars.

Question three, where I see the opportunities

I will be straight about my bias, not advice.

First, Ethereum's relative strength is the cleanest expression of this setup. When institutional flows rotate toward ETH while BTC ETFs bleed, allocators are usually hunting higher beta inside the same asset class. The 200 day average near 2,487 is the line I care about. Holding it keeps the thesis alive; losing it invalidates it fast.

Second, Bitcoin's 76,000 to 77,000 zone is the decision area. It held as the September 11 low at 76,023 and is the range floor. Above, 79,874 and the September 6 high at 80,560 need to fall for this to be more than consolidation.

Third, on equities, a hiking regime has clear internal logic. Financials benefit from wider net interest margins, energy benefits from the oil bid that caused part of this inflation, while long duration growth, semiconductors, REITs and utilities carry the most valuation risk with the 10 year knocking on 5 percent. Small caps get squeezed hardest, which is why I trust the Russell 2000 bounce least.

Fourth, gold at 4,408 dollars with a 10 year near 5 percent is not normal. That combination should pressure gold. Being merely flat tells me there is persistent demand for insurance that rate math alone does not explain.

Where the market goes from here

I will give conditions rather than predictions, because next week is binary.

If the Fed hikes 25 basis points on September 16 but frames it as a pause with a data dependent path, my base case is a relief rally. I would expect Bitcoin to reclaim 79,000 to 80,500 and Ethereum to attack the 2,666 September 11 high, with ETH ETF inflows continuing. That is the world where this CPI print was the peak of the fear.

If the Fed hikes and the updated projections signal more to come, pushing December pricing into the base case, the 10 year likely clears 5 percent. I would then expect Bitcoin to lose 76,000 and trade into 72,000 to 74,000, with small caps leading equities lower and altcoins falling harder than Bitcoin, because dominance at 58.65 percent says money is already hiding in the largest asset.

The middle scenario, a hike with no clear signal, is the one I assign the highest probability. I would expect a knee jerk spike both ways within 48 hours, a failed break of either 76,000 or 80,500, and a range bound market into the October 14 CPI release and the Q3 earnings window opening in mid October. Falling open interest into the event says leverage has already been trimmed.

Three things would change my mind fastest: core CPI back at 0.2 percent or lower, a decisive drop in the 10 year toward 4.5 percent, or Bitcoin ETF flows turning sustainably positive. Until one appears, the path of least resistance looks sideways with a downward tilt, punctuated by sharp moves around macro releases.

Final thought

The consensus read is that this CPI report was a nothingburger because the headline matched. I think that is exactly wrong. Headline matched, core beat to the upside, gasoline did a third of the damage, hike odds moved from 68 percent to above 85 percent in one session and the 10 year touched 4.97 percent. That is the moment the market stopped pretending cuts were coming. Crypto absorbed it better than equities, Ethereum is leading, and Bitcoin is compressing below its 200 day average waiting for a decision that arrives in four days.

#weeklyshare