The Month in Markets, and the Bigger Picture
A longer look across crypto, FX and commodities for September: what moved, the macro picture the tape is painting, where we sit in the cycle, and a proper opinion on where the balance of risk lies from here.
The month in review
Figures cover the thirty days to 27 September 2026, a trailing window rather than the calendar month, and are captured at 21:08 GST on that day. Another Levanter piece drawn on a different day will differ for that reason alone.
The month read risk-on. Across every market we cover, the strongest performers were QNT (+222.7%), NEAR (+200.5%), BTW (+184.5%), and the weakest single market was COFFEE at -22%. The spread between them, and where each sits by asset class, is the month's story in one line.
Crypto carried the risk appetite, with 34 of 35 coins higher on the month (bitcoin +9% and ether +10%) and TRX the notable faller at -2%. Bitcoin dominance is near 68%.
In currencies the dollar was stronger. USDMXN was the standout pair (+4.5%) and NZDUSD the weakest (-5.0%).
Commodities averaged +0.0%, led by NAT GAS (+13%) with COFFEE the laggard (-22%).
Rotation and leadership
Step back from the individual names and the rotation is clearest at the asset-class level. On average crypto did the most work this month (+47.0%) and FX the least (-0.4%).
Which class leads tells you what the market is paying up for. Risk and liquidity, or safety and hard assets. That is worth more than any single ticker.
Inside crypto, the equal-weighted basket returned +47% against +13% cap-weighted. The average coin beat the heavyweights, so the move broadened into smaller names. Historically that signals healthy appetite, and also a later, frothier stage where the quality bar quietly drops. Dominance near 68% fits the picture.
Rotation is worth tracking because it turns before prices do. Leadership passing from the majors to the small caps, from crypto to gold, or from growth-sensitive metals to defensive ones, is the market rehearsing its next mood while the index still looks calm. We would rather catch the rehearsal than wait for the show.
The macro picture
Start with the dollar, because it prices everything else. It was stronger on the month.
A firmer dollar tends to tighten global financial conditions and to lean against commodities and risk assets priced in it. We take that from the tape, not from any headline. It is the single most important number here.
Gold was -7.1% on the month, and it is the market's quiet barometer of real rates and fear at once. Its strength giving way points to the opposite, a market comfortable enough to leave the safety trade. We read it as a sentiment gauge and leave it at that.
The industrial complex is the reality check on the narrative.
Copper, the metal with a PhD in economics, was +2.6% and oil +12.4%. Read side by side, that points to a genuine growth impulse. When the paper markets and the physical economy disagree, the physical economy is usually the one worth believing.
Put it on one canvas and the month's macro tell is this: crypto trading risk-on while gold slipped is about as honest a risk-on signal as markets produce. We offer that as context. Context is what stops you reading a single market in a vacuum.
Everything else is downstream of the dollar.
Where we are in the cycle
Here is where we stand. The cycle clock reads post-peak cooldown. We are roughly 890 days past the 2024 halving. Bitcoin sits about 41% below the cycle gauge's trend line. Our separate valuation fit, run on a different price history, puts fair value near $135k and its floor around $58k. Both fit price against network age, so read their agreement as overlap rather than confirmation. Ether trades about 24% below its own trend. Solana sits about 31% below its own trend.
A word on how we value the network, because it is fashionable to quote Metcalfe's Law, the idea that a network is worth the square of its users. It is a good idea that has stopped working for bitcoin.
Fit against active addresses since 2019 it has an R-squared of about 0.01 and the wrong sign, because exchange batching, layer-two activity and post-ETF custody now hide real users from the on-chain count.
What still holds is the valuation fit, log price on log network age, and that is the fair value and floor we quote. We would rather tell you which model broke than quote you a number that sounds clever and means nothing.
The ether-to-bitcoin ratio is 0.0318. Leadership inside crypto rotates, and the majors do not move as one, which is why a single 'crypto' number hides more than it reveals.
Risks, and what would change our mind
No honest monthly skips the other side of the argument, so here is ours, plainly.
The bull case. The cross-market tape is risk-on, crypto breadth is positive. Taken together that is an environment where risk has been rewarded and the path of least resistance has been up.
The bear case. The cycle reads post-peak cooldown, the halving math points to diminishing returns, the volatility model leans turbulent. Set against the bull case, that leaves the easy gains possibly behind us and the margin for error thinner than it feels.
What would change our mind, either way. A decisive break in the dollar, gold rolling over or accelerating, a spike in cross-asset correlation, or a flip in the volatility regime. Those are the signals we watch. A loud headline is not one of them.
Opinion: the Levanter thesis
Every cycle produces the same conversation at roughly the same point. The early move is dismissed, the middle is doubted, the top is celebrated as a new paradigm, and the cooldown is explained away as a healthy pause right up until it is not.
We appear to be somewhere in the second half of that arc. Humility serves better there than a target.
The uncomfortable fact the halving math keeps repeating is diminishing returns. Each era has delivered a smaller multiple than the one before, because a market cannot keep compounding at the same rate as its base grows without eventually swallowing the entire world. That is not bearishness. It is arithmetic.
The people who lose the most in this phase are the ones who size their expectations to the last cycle rather than the trend of cycles.
There is a subtler trap in a cooldown. It can last far longer, and feel far more constructive, than a crash.
A market that grinds within a wide range for months trains people out of their discipline. It rewards the sellers of options and the takers of borrowed money, then reminds everyone at once why those trades carried a premium in the first place. Risk accumulates quietly in exactly these stretches, which is what makes them dangerous.
None of that tells you what price does next month, and we will not pretend it does. What it tells you is how to hold whatever you hold: with position sizes that assume the drawdowns of this asset class are real and recurring, not theoretical, and with a plan that survives being wrong.
The broader point is that structure beats prediction.
Where an asset sits against its own long history, how its volatility is behaving, and whether the whole board is moving as one are all knowable. The next candle is not. Our own volatility read backs this up, landing near 72% at a month and 71% at a quarter, while our direction calls sit where theory says they should, close to a coin flip.
If that all sounds like a counsel of modesty, it is, and deliberately so.
The single most expensive belief in this business is that someone, somewhere, can tell you what happens next. The entire architecture of financial media exists to sell you that belief on a monthly subscription. We are trying to sell you the opposite: a clear-eyed read of what is knowable, and an honest label on what is not.
So take from this what the data actually supports and leave the rest. Watch the volatility, respect the cycle, and let the process carry the weight. We will be back next month with the same discipline and, in all likelihood, a different-looking market to apply it to.
The month ahead
For the month ahead we hold the same discipline. We will not tell you where prices are going, because we cannot and neither can anyone selling you the opposite. We will tell you where turbulence is likely to sit, where each market stands against its own history, and what would change the picture.
As it stands, the volatility model frames the coming weeks as turbulent across the board and the cross-market backdrop remains risk-on. If the dollar or gold breaks its recent character, or correlations spike, that is the signal to revisit the whole read. We will, as the data does.