The Levanter Signal. A weekly read of volatility, valuation and the week ahead across crypto, FX and commodities. This is the subscriber tier, and it is free while we build the list. We will tell you before that changes. Subscribe at read.levantermarkets.com.
Data captured at 07:33 GST on 16 September 2026. Every figure below is stamped to a period. This is the accountable read behind the free weekly: the changes since last week, the levels to watch, and a claim we will score in the next issue.
Editor’s line. My read of the week: every bit of the noise is parked in metals and energy while crypto and FX sit quiet, and a bitcoin 44 percent under its own trend line is a valuation worth noting, not a dip that needs buying.
The seven-day volatility map
This is the part with measurable skill, and it is why it leads the issue. The model tags each market turbulent or calm for the week ahead. In the five-year point-in-time backtest it classified the seven-day regime correctly about 65 percent of the time across 5,118 calls, 95 percent interval 63 to 66, 13 percentage points above its naïve baseline, and 73 percent at thirty days, 21 points above baseline. That is a backtest, not a live forward record: the live scoreboard is only now starting to fill.
Accuracy alone can flatter a model that never commits, so we also score the confidence behind each call. The Brier score is 0.228 against 0.25 for always guessing the base rate, a skill score of 0.088 over 2,559 scored calls. Positive but small. Read the calls as a lean, not a conviction.
For the coming week the model reads 15 markets turbulent: silver, oil, copper, platinum, palladium, Brent crude, wheat, corn, soybeans, coffee, sugar, cotton, gasoline, heating oil and USDJPY. The rest of the displayed set is calm, including bitcoin, ether, solana, six of the seven FX pairs, the S&P 500 and natural gas. The average market is therefore contained even though a few names are carrying wide ranges.
Worth naming: 14 of the 15 turbulent markets are commodities, with only the yen outside that class, and that name sits close to the line. Treat this as turbulence concentrated in one corner of the board rather than a market that is nervous everywhere. The two call for different responses.
Average cross-asset correlation is near 0.61. That is high enough that diversification is thin: position count is not the same as risk spread this week.
The full board, market by market
This is the model’s working rather than its conclusion. Volatility is annualised. The median column is each market’s own long-run median, so every row is judged against itself and never against a common threshold: a 6 percent reading in FX can be stretched while a 35 percent reading in crypto is quiet. Percentile is where the current reading sits in that market’s own history. The thirty-day column is there so you can see whether a call is a one-week disturbance or a settled regime.
The most stretched reading on the board is platinum, running 2.38 times its own median, and the quietest is the euro at 0.38 times. A market can be called turbulent while still sitting below another market’s calm reading, which is the whole point of judging each one against itself.
Treat the rows near the line with less confidence than the rest. Gold at 0.99x is close enough to their own median that the call could go either way. We would rather flag that than present every row as equally settled. The rows worth acting on are the stretched readings at the top and the quiet ones at the bottom.
Where the two horizons disagree. Turbulent this week but calm at thirty days, soybeans. Calm this week but turbulent at thirty days, gold and ether. A split like that flags a near-term move without telling you whether it lasts, so it is the set to watch rather than the set to act on.
Out-of-range flag: the Aussie and sterling are trading outside the volatility range the model was fitted on. The call still prints, but it is an extrapolation and we would discount it accordingly.
The big crypto reads. Bitcoin’s one-week volatility is near 26 percent against a 39 median, 0.67 times its own normal, and ether is near 39 percent against a 53 median, 0.74 times its own normal. Bitcoin reads calm at both seven and thirty days, and ether reads calm at seven days and turbulent at thirty. The two horizons disagree, which flags a near-term move without telling us whether it lasts. In the metals, copper, palladium, platinum and silver read turbulent at both seven and thirty days, so that is not just a one-week disturbance.
On direction the model is close to a coin flip: 50 percent over 320 backtested crypto calls, a 95 percent interval of 45 to 55 percent that straddles the coin-flip line. We forecast volatility, not direction. Anyone selling you the second thing at these sample sizes is selling you noise.
What this map is not: it says nothing about which way a price goes, it cannot tell you why a market is stretched, and a turbulent call is not a reason to trade. It is a statement about the width of the range, which is the input to position size, not to direction.
What changed since the last Signal
Week on week: Newly turbulent, copper. Calmed back to normal, gold. Bitcoin is about 2 points cheaper against its fitted value.
The one chart: bitcoin against its long-run trend
Bitcoin is near 76,000 dollars. The valuation fit models price against how long the network has existed, on a log-log scale. Fair value on that fit lands near 135,000, about 44 percent below the line, and the fitted floor sits near 58,000. Bitcoin has closed above that floor line for roughly 95 percent of the historical sample. That is an in-sample observation, not a tested probability and not a guaranteed level of support.
Our cycle gauge reports a second number, and it is worth being precise about what it is. It fits the same shape of curve, price against network age, but on a different price history and with a different band, then adds halving timing to classify the phase. It reads bitcoin as post-peak cooldown, about 46 percent below its own trend line. Do not read the two figures as confirming each other. They are the same kind of fit run over overlapping data, so close agreement is close to guaranteed and tells you nothing the first number did not. Both are long-horizon context. Where price sits against a multi-year fit says nothing about the next five days, so read it as valuation, not a reason to act on the week.
What the model can and cannot do
It is a statistical fit of price to time. It has no hard economic mechanism behind it, cannot call tops, and may fail outside the historical sample. It is a valuation anchor, not a timing tool. Treat the fair value and the floor as distant reference points, never as targets and never as a reason to size up.
The week behind, and what it rhymes with
Over the past seven days crypto was narrow and speculative-led: 3 of 35 coins higher, cap-weighted about -4 percent on the week and +22 percent over thirty days, led by the speculative end, XMR +2 percent and WLFI +2 percent, with a best-to-worst spread near 24 points. Dominance held near 69 percent and the stablecoins we track kept their pegs. In foreign exchange the biggest seven-day move was USDSEK at +2.0 percent, ranges otherwise tight. In commodities the metals fell across the board, gold -1 percent, silver -1 percent and palladium -2 percent. The gains were narrow, but the largest moves stayed further out on the risk curve, and the dollar and most FX ranges were comparatively quiet.
Read across the three asset classes, the turbulence is concentrated in silver, oil, copper, platinum, palladium, Brent crude, wheat, corn, soybeans, coffee, sugar, cotton, gasoline, heating oil and USDJPY while the rest stays quiet. That is specific pockets of risk rather than a broad regime shift, and the tape alone does not tell us why they are the loud ones this week.
Subscriber watchlist, with levels
• Bitcoin. Fitted floor near 58,000, fair value near 135,000. A weekly close below the fitted floor would be historically unusual and would challenge the model, rather than automatically creating a buying opportunity.
• Ether volatility. Current annualised volatility is near 39 percent against a historical median around 53. Watch whether the thirty-day classification also flips from calm to turbulent.
• The commodity complex. Whether the turbulent bid broadens beyond silver, oil, copper, platinum, palladium, Brent crude, wheat, corn, soybeans, coffee, sugar, cotton, gasoline and heating oil or fades back to calm.
• Pegs and dominance. Stablecoins are holding and bitcoin dominance is near 69 percent. A tracked peg below 0.995 would trigger Levanter’s wobble alert. A sharp dominance move would show the balance within crypto changing.
To score next week: the model calls silver, oil, copper, platinum, palladium, Brent crude, wheat, corn, soybeans, coffee, sugar, cotton, gasoline, heating oil and USDJPY turbulent and the rest calm. In the next issue we score each call the way the model does, whether realised volatility over the week came in above or below the asset’s running-median volatility, and show the hits and misses. That is the claim you can hold this Signal to. One call sits close enough to the line that we would not defend it hard: gold at 0.99x, called calm. We score it anyway. Quietly dropping the calls that look shaky is how a scoreboard gets flattered, and a scoreboard you cannot trust is worth nothing to you.
*This is the Levanter Signal, the weekly subscriber note, free for now while we build the list. We will tell you before that changes. Subscribe at read.levantermarkets.com. The daily, weekly and monthly reviews stay free at levantermarkets.com. Educational market analysis, not financial advice.*