Last week, we said BTC was still in a range and that $85K was the level to watch. It never got back there. BTC started this week near $82K, slipped below $80K, bounced after the Fed, and then gave the bounce back. Now it sits near $80K again.
Two months ago, we said BTC had entered a long consolidation. We still read the market that way. The price has gone nowhere, but a lot is happening inside the range. Coins keep changing hands at the same prices, while open interest and funding rates come down. If this goes on without a breakdown, the market will have a stronger cost base for the next move.
Volatility is getting tighter too. That doesn’t tell us which way the next move goes, or when it comes. It tells us a bigger move is building. We still think the range breaks higher. Until then, we trade less and let the range play out.
What Changed
Macro. The Fed paused, but it did not pivot. Rates stay at 3.75%-4.00%, and two officials voted for another 25bp hike. The statement was firm. Growth is strong, inflation is still high, and the Fed wants more data before its next step. The press conference sounded softer, so the market cut the odds of a December hike. Both things can be true. The Fed still leans toward tightening. It is just not in a hurry.
The GDP data tell the same story. Q3 GDP tracking fell to 3.2%, but net exports took about three points off. Part of that comes from US companies buying AI equipment made in Taiwan, Korea and Mexico. Demand at home is still strong. Consumer spending is growing about 3.5%, equipment investment about 20%, and domestic final sales about 4.3%. AI spending is still going strong. Some of it just shows up in other countries’ numbers.
Inflation is the harder part. For years, memory chips and other tech products helped push goods prices down. Now they are starting to push them up. Market inflation expectations are still calm, but the near-term risk is going up. So we don’t buy the simple story that slower GDP means quick rate cuts.
AI. The capex downturn still hasn’t come. If anything, the cycle is getting longer. Alphabet and Microsoft spent about $97 billion in the September quarter, up 64% from a year ago. Alphabet, Microsoft and Meta together point to about $174 billion of capex in Q4, close to double last year. Alphabet and Meta both say spending goes up again in 2027.
Microsoft gives the clearest read on demand. Azure grew 48% in constant currency, and Microsoft guided 49%-50% for next quarter. Capex came in below plan for two reasons. Microsoft got more out of each chip, and some outside capacity arrived late. Customers did not ask for less compute. The spending is also spreading out, into HBM, chip equipment, networking, power and cooling.
There are still real risks:
Financing and customer concentration are still the weak spots. Oracle expects to raise about $20 billion, and OpenAI looks like more than half of its contracted backlog. Supplier financing and prepayments help, but they mostly pass the risk to someone else.
HBM looks sold out for years. But when customers order two or three times to be safe, demand looks bigger than it really is.
AI software is making more money, but the picture is messier than in hardware. ServiceNow raised its year-end AI contract value target to $1.7 billion, while its near-term backlog is still growing slowly.
Positioning is still odd. Leveraged equity ETFs saw record inflows, and QQQ short interest is the highest since 2023. But trading in regular stocks is still near the low end of its history. There is fuel for another squeeze. We just don’t see broad FOMO.
Our view stays the same. The AI cycle holds, and the evidence now reaches further into 2027. We stay long companies with real orders, strong balance sheets and clear returns. We don’t chase the most leveraged parts of the trade.
Crypto. In futures, aggressive traders were net sellers. In spot, buyers kept stepping in. We think these are two different groups. Leveraged traders cut risk, while cash buyers bought the dip. Open interest and funding both fell, so this looks like old longs leaving. We don’t see new shorts piling in.
ETF flows got better, but they are not steady yet. Spot BTC ETFs took in about $424 million over the last three US trading days. Over five days, there was still a net outflow of about $350 million. That’s an improvement after heavy selling. It is not a new wave of demand yet.
Altcoins still lag BTC, as we expected. What’s new is that alt positioning is cooling. Part of the drop in open interest came from lower prices, but alt open interest also fell faster than BTC’s. Coins outside BTC and ETH now make up about 32% of crypto open interest. Over the last seven days, only 26 of the top 80 alts beat BTC, and only 11 went up at all. This doesn’t mean alts are turning. It is one thing that has to happen first. We still need wider breadth and real spot buying. BTC stays our main crypto position.
Signal Scorecard
