I predicted Bitcoin falling to $49k this year and January delivered some very concerning red flags

My $49k Bitcoin bear thesis, a January check-in, the plumbing is flashing while price bleeds

I wrote my medium-term $49,000 bear thesis in late November with one simple idea, Bitcoin still moves in cycles, and the next real “this is the low” moment tends to arrive when miner economics and flows line up at the same time.

It is now Jan. 30, 2026, and the honest update is this, the variables I care about look more stressed than they did when I published, and the tape has not delivered the kind of panic price print that makes those variables matter to everyone at once.

Somewhat paradoxically, my ‘medium-term bear thesis’ was intended to be long-term bullish. The idea being that we could get a short, sharp bear market with max pain followed by a sustained, multi-year bull run. However, the price isn’t quite matching with the signals right now.

Bitcoin is hovering around the low $80,000s (after falling to $81,000 overnight) as I write this, which means my high-$40ks zone has not even come into view yet.

That disconnect is the story.

Because beneath the price, the parts of the system that pay for Bitcoin’s security, and the parts that move institutional size, are acting like winter already arrived.

The winter feeling is coming from fees, not the chart

Start with the security budget, because that was my original “fragility” claim.

On Jan. 29, miners earned about $37.22 million in daily revenue.

On the same date, total transaction fees paid per day were about $260,550.

Do the math and you get the mood music, fees are roughly 0.7% of miner revenue.

That is not “fees are weak,” that is “fees are basically absent,” in the sense that the fee market is contributing almost nothing to the cost of securing the chain on a day-to-day basis.

Even the live mempool picture looks sleepy. The projected next-block median fee rate is around 0.12 to 0.14 sat/vB right now.

So when people ask why I keep circling back to miner economics, it is because this is what a fee floor failing looks like in real time. The network leans on issuance, issuance steps down on schedule, and everything else has to pick up the slack later.

The ETF window has been a steady leak, with a few ugly gulps

The second leg of my framework was flow elasticity, the idea that the ETF era creates a clean, mechanical way to see risk appetite turn.

In January, that elasticity has been pointing in the wrong direction.

On Farside, the last few weeks show multiple heavy outflow prints, including -$708.7M on Jan. 21 and -$817.8M on Jan. 29.

Total net flows are also negative at -$1.095B year-to-date. That matters more than any single day because it changes the psychology of dips. In the soft-landing version of my thesis, the tape gets support from persistent dip buying through the ETF pipe. Right now, the pipe has been taking water out.

There were big green days earlier in the month too, Jan. 13 at +$753.8M and Jan. 14 at +$840.6M, and those are real, but the late-month flow prints have been the kind you feel on a desk.

If you trade for a living, you know this sensation, price holds up, the internals start to rot, and everyone keeps looking for the moment the chart finally reflects what the plumbing has been saying.

Hashrate is wobbling, miners are adapting, and that adaptation changes behavior

Another piece of the setup is miner elasticity.

Hashrate is still huge, but it has been swinging. On Jan. 29 the daily average is roughly 901 EH/s, down from earlier peaks this month.

That by itself does not equal capitulation, and I am not trying to force a dramatic story onto routine variance. It does fit the broader point, miners now have more knobs to turn.

The most important knob is the one nobody talked about in prior cycles, AI and HPC hosting.

When a miner signs long-duration compute deals, that business starts to look less like a pure BTC margin machine and more like a power, land, and infrastructure operator that happens to mine Bitcoin.

TeraWulf put that shift in bold print when it announced two 10-year HPC colocation agreements with Fluidstack for 200+ MW, with Google backstopping a large portion of obligations and receiving an equity stake, per the company’s own release.

Riot has been exploring the same direction, including a formal evaluation to potentially repurpose significant capacity for AI and HPC, according to DataCenterDynamics.

This matters for Bitcoin market structure because it changes the incentives around hashrate at the lows.

A miner with a second revenue stream can behave differently under stress. They might curtail or redirect capacity without immediate existential pressure, they might protect liquidity for buildouts, they might sell BTC more mechanically to fund capex, they might simply stop caring about marginal hashprice in the way a pure miner once did.

That is the elasticity I was pointing at, and it is starting to show up in the data’s tone even while price sits high.

So what is the “state of the thesis” right now

Here is the cleanest way I can say it in one breath.

The fee floor looks broken, ETF flows have been risk-off for weeks, and the miner business model is evolving in a way that can amplify reflexive behavior during drawdowns.

Those are the conditions I wrote about.

The missing ingredient is the part people remember, the chart dumping into the zone where panic turns into inventory transfer.

Bitcoin at $82k does not force anyone to make that decision. A print in the $40ks would.

That is why this update is less about price targets and more about tension. The system is building tension.

The human-interest angle people miss, miners are running two companies at once

When you reduce this to “fees are down,” it sounds like a chart note.

In real life it looks like operators trying to keep the lights on, negotiating power contracts, planning buildouts, courting AI customers, juggling shareholders, and still needing to compete in the most brutal hash race on earth.

A low-fee environment does not just weaken the security budget, it forces miners to get creative, and creativity introduces new behaviors into the market.

The base-case bear I described in November was always about that behavior showing up at the same time as flow pressure, and then price finally doing the thing it does when leverage and narrative crack together.

Right now, two of those levers are already pulled.

What would make me say the bear is resolving early

I am keeping my flip-level framework, and I am keeping it boring on purpose.

• Fees need to stop living in the mud, the YCharts fee line needs to rebuild a real floor relative to the YCharts revenue line.
• ETF flow behavior needs to change, the Farside table needs to show consistent dip buying again, not late-month air pockets.
• Mempool conditions need to feel alive again, fee pressure showing up in the mempool medians in a way that suggests real settlement demand.

If those happen while price stays elevated, the “shortest winter yet” framing starts to win.

If those stay weak and price eventually breaks, the $49k style print stays in play as a liquidity magnet, because that is where the buyer base tends to change character.

Where I stand today

I do not have the cathartic conclusion that every market story wants, because the market has not given it yet.

The infrastructure tells me winter conditions are already here.

The chart tells me the crowd has not felt them.

That gap is the thing to watch, because gaps like this do not usually persist forever.

And when they close, they close fast.