Three different people asked me some version of the same question last week: is it back? In June the messages I was getting asked whether to sell. That flip, from “should I get out” to “am I already too late,” took about nine days.

Bitcoin settled Friday at $78,344, after tagging $79,526 intraday. The Friday before, it closed at $63,048. That is a 24% week, and CoinDesk called it the best since 2023. I’ve been trading and brokering Bitcoin out of Panama since 2011, so I’m not going to pretend a week like that is nothing.

I’m writing this on Sunday, with price at $77,653 and the weekend behind us. That extra couple of days turns out to matter, because what has happened since Friday tells you more than the spike did.

My read is still that this looks like a bear-market bounce that may well be building a higher floor, rather than a bear market that has ended. Here’s the reasoning, including the part that has moved against me.

What actually moved it

Four things stacked up inside a week, and only one of them is what I’d call organic demand.

The U.S. Treasury doubled its long-end buybacks, and then the bond market shrugged. On August 19 Treasury announced it would raise liquidity-support buybacks in the 10-to-20-year and 20-to-30-year sectors from a $2 billion maximum per operation to at least $4 billion, effective September 9 and running through November 4. Long yields fell on the news, with the 30-year giving up 9 basis points to 5.196% and the 10-year 6 basis points to 4.647%, per CNBC.

The relief lasted about a day. By August 20 that decline had been wiped out, and on Friday the 30-year climbed again to 5.273% with the 10-year back around 4.7%. Treasury Secretary Scott Bessent had to say publicly that the $4 billion figure was a floor rather than a ceiling. So the macro spark that lit this move has already round-tripped, and Bitcoin kept the gains anyway. That is the most interesting fact of the past week and I’ll come back to it.

Shorts got run over. Traders who had been positioned against Bitcoin for six weeks were forced to buy it back, which pushed price up, which liquidated more shorts. CoinGlass counted roughly $2.7 billion in crypto short liquidations over 24 hours around August 19-20, as Bloomberg reported, with shorts making up about 92% of the total. That is forced buying, and forced buying stops when the shorts run out.

ETF demand genuinely came back. This is the real one. U.S. spot Bitcoin ETFs have now taken in money for five straight sessions: $297.5M on August 17, $189.3M on the 18th, $517.2M on the 19th, $606.3M on the 20th and $307.5M on Friday, with BlackRock’s IBIT taking $239.3M of the last one (Farside). That is about $1.92 billion in five days. Spot buying by funds that have to actually hold the coin is a different animal from a derivatives squeeze, and it deserves the weight.

Washington made encouraging noises. The Senate has a cloture vote on the CLARITY Act, the market-structure bill that would split SEC and CFTC jurisdiction, scheduled for September 15, and President Trump voiced support at a White House meeting on the 19th. Worth being clear what that vote is: a procedural motion needing 60 votes just to begin debate, with unresolved fights over government-ethics provisions, illicit-finance rules, and bank objections to letting exchanges pay yield on stablecoin balances. Congress returns September 14 with 14 working days before the October election recess. That is a sentiment tailwind, not a law.

The sentiment flip is the part that bothers me

Here is the move in one picture. The solid line is price; the dashed line is the Crypto Fear & Greed Index.

BTC price vs the Crypto Fear & Greed Index, Aug 16 – Aug 23 (Aug 23 intraday)$60k$70k$80k255075Aug 16Aug 17Aug 18Aug 19Aug 20Aug 21Aug 22Aug 23BTC priceFear & Greed

BTC price against the Crypto Fear & Greed Index, August 16 to August 23. The final point is an intraday read as of 22:05 UTC, not a settled daily close. Sources: Crypto.com Exchange 1D candles (BTC/USDT) and alternative.me.

The index read 27 on August 12 and 34 on the 16th, squarely in Fear. It hit 72 on Friday and has eased back to 66 as I write. Nine days from Fear to Greed, and the mood has now come slightly off the boil without price doing the same.

I don’t treat Fear & Greed as a timing tool, and neither should you. What it measures is the mood of the crowd, and a crowd that reprices its own mood by 45 points in nine days is reacting to the chart, not to anything it learned. That is chasing, and it tends to arrive at the point in a move where the newest buyer is least prepared to sit through a 15% retrace.

What I said in June, and what actually happened

In June I wrote a piece about the “$48,000 Bitcoin” chart that was going around, and I put a stake in the ground: this was a confirmed bear by the 50-week rule, the 200-week average near $62,000 was the level that mattered, and my base case was a shallower bottom nearer the mid-$50,000s than a clean flush to $48K.

The low so far was $57,803, set in July on Crypto.com’s BTC/USDT tape. June bottomed at $58,100. So that call was roughly right, and I’d rather say that plainly than pretend I don’t keep score.

Which is exactly why I’m not flipping now. The reasoning that got me the shallower bottom, that ETFs and corporate treasuries absorb supply in a way the 2018 and 2022 markets couldn’t, is the same reasoning that says the recovery should be a grind rather than a rocket. A 24% week driven by liquidations is not what that thesis predicts. It’s what the old market did.

The honest bull case

I don’t want to argue one side of this, so here is what genuinely favours the optimists, and it got stronger over the weekend.

Price held after the macro catalyst went away. The yield relief that started this was fully unwound by Friday, and Bitcoin still settled at its best level since May and spent two days between roughly $75,550 and $78,830. A pure squeeze usually gives back most of the move once the forced buying stops and the story that justified it evaporates. This one hasn’t, so far.

The drawdown from the October 6, 2025 peak of $126,198 bottomed around 54%. Past Bitcoin bears took roughly 75% to 85% off the top. If the institutional bid really has compressed both the euphoria and the panic, a shallower and shorter bear is precisely what you’d expect, and a higher floor at each ratchet follows from it.

The $62,000 to $65,000 shelf that contained price for most of August broke on real volume with real ETF flow behind it, not on a thin overnight wick. That is the kind of break that tends to become support.

The case that this is still a bounce

Glassnode published a read on August 20 that lines up with my gut, and it’s the most concrete counterweight I’ve seen. Their Realized Profit/Loss Ratio sits at 0.75. It has historically needed to fall below 0.5 to mark genuine seller exhaustion at a cycle bottom, and to climb toward 2 to confirm a regime change. At 0.75 it is in neither place. Glassnode’s own conclusion, reported here, is to treat the recovery “as a local rally rather than a full regime change.”

They also flag that the Coinbase Premium remains negative, meaning U.S. spot buyers have not meaningfully re-engaged even after a 24% week. Put that beside the liquidation data and the picture is consistent: this move was made in the derivatives market, not the spot market.

That distinction lands differently depending on where you’re sitting. Most retail buying across Latin America is spot, on local exchanges, in whatever size a paycheck allows. Very little of it is leveraged, and almost none of it can touch a U.S. spot ETF directly. So a week like this one happens to our market rather than with it: you get the price without having been part of the mechanism that made it, and the temptation is to buy in right after the leverage has finished doing its work. That is the worst seat at the table.

And the shallow-drawdown argument cuts both ways, which the bulls tend to skip. Yes, 54% is mild against a 75% to 85% history. It’s also entirely possible that means this bear simply isn’t finished. Total crypto market capitalization is about $2.63 trillion today against a record near $4.3 trillion last October, still roughly 40% below. Bitcoin itself is about 38% under its all-time high. None of that is euphoria territory. It’s just less cheap than it was two weeks ago.

What I’m actually watching

Not the highs. The retest, which has not happened yet.

Two quiet days in a $3,000 range over a weekend, with the ETF desks shut, is not a test of anything. The real one comes when a proper wave of selling arrives and we find out where it stops. If the low-to-mid $70,000s absorb it, the higher-floor thesis has evidence behind it and I’ll say so. If price slides back under $70,000 and keeps going toward the old $62,000 to $65,000 shelf, then what we watched was a squeeze that ran out of shorts.

Beyond that: whether ETF creations hold up now that price has stopped going vertical, whether long yields keep climbing after September 9 or the buybacks eventually bite, whether the Coinbase Premium turns positive, and what the Senate actually does on September 15 rather than what the market currently assumes it will do.

The takeaway

If you’re sitting on a decision this weekend, the useful question isn’t whether Bitcoin goes up from here. It’s whether you’d be making the same decision if the last two weeks had gone the other way. If the answer is no, then what you’re actually trading is the mood in that dashed line.

The things that help in a week like this are unglamorous. Know why you own what you own. Size to survive a 20% move against you, because you just watched one happen in the other direction. Hold your own keys, especially now, when volatile weeks bring out thin local liquidity, wobbly platforms, and the recovery-scam crowd. And be suspicious of any argument, mine included, that got dramatically more confident in nine days.

This is analysis, not advice: I have no idea what price does next, and neither does anyone quoting a target at you.

Sources