This is a forecast with no price target in it. Bitcoin traded at 79,108 dollars on 1 September 2026, down 9.62 percent on the year, while August pulled 3.52 billion dollars into the spot ETFs. Those two facts point in opposite directions. What follows is how I read them, and why I could be wrong.
The Bull Case, In Three Numbers
First, the flows. Spot Bitcoin ETFs took in 3.52 billion dollars in August 2026, per SoSoValue. That money arrived while the price was falling. A buyer who allocates into a down month is not chasing a chart, and that is the cleanest signal this asset class produces. Price tells you what the marginal seller wants today. Flows tell you what the patient buyer is doing all month.
Second, turnover has not left. Solana settled roughly 1.178 billion dollars of DEX volume a day and about 50.3 billion over thirty days, and memecoins were 42 percent of it. Pump.fun accounted for about 492 million of that thirty day figure, ran 13.68 million dollars of weekly fees, and did 2.4 million dollars on 25 August 2026 alone. Annualised that is somewhere near 460 to 500 million dollars. A dead market does not have a fee run rate.
Third, the deposits fell but the rails did not. DeFi total value locked went from about 115 billion dollars in January 2026 to about 70 billion, a 39 percent decline. Aave alone went from 26.4 billion to 14.3 billion. Nothing broke to cause that. Yields compressed, rates elsewhere got competitive, and money left. The contracts kept settling the whole way down.
Why That Is A Forecast And Not A Price
A price target is a guess with a decimal point on it. What I will say instead is directional and testable. Capital that enters an asset through a regulated wrapper during a drawdown behaves differently than capital that enters through an exchange during a rally. It is slower to leave. If ETF inflows stay positive through a second consecutive negative quarter, the holder base has changed composition, and that change outlives any single move. That is the thing worth watching. Not the number on the screen.
The Case Against Everything Above
Security is the honest counterweight. The second quarter of 2026 saw 85 separate incidents and 775 million dollars in losses. The year to date total is 942 million. Drift accounted for 295 million and KelpDAO for 293 million, so two events are about 62 percent of the entire year. That is not a tail sitting quietly in a risk model. That is the risk model.
Fee revenue is also not adoption. Memecoin turnover measures risk appetite, and risk appetite is the first thing to leave when conditions tighten. A fee run rate built on speculation is reflexive by construction, so it does not forecast anything. It confirms what already happened, which is a different job.
And ETF flows reverse. The same wrapper that made buying easy makes selling easy, and a monthly inflow figure is a snapshot, not a commitment. Anyone reading August as proof of anything should ask what they would say if September prints negative. If the answer changes the thesis, the thesis was the flow, not the asset.
What I Do With This
I hold, I use no leverage, and I size positions so a total loss changes nothing about my month. I do not trade forecasts, including my own. The point of writing one down is that it can be checked later against what happened, which is the only thing that separates a view from a feeling.
Disclosure
I am not a licensed financial advisor and nothing here is a recommendation. Every figure carries its source and the date it was read. Prices, flows and fee run rates move, and this page is not updated when they do. I hold Solana and a token from one Solana NFT collection. The full disclaimer is at codywise.io/legal/disclaimer.
