# Positioning Fell 77%, Prices Only 10%: Notes on MacroVoices #559 Source: Realpha Blog (blog.getrealpha.com) Original article and charts: https://blog.getrealpha.com/en/blog/macrovoices-2026-11-19-macrovoices-559-ole-hansen-what-s-next-after-the-s/ > Ole Hansen of Saxo Bank on the speculative washout in commodities, from MacroVoices episode 559 recorded November 19, 2026. These are personal listening notes for educational discussion — not investment advice, and no security is recommended here. Published: 2026-11-20 Locale: en Tags: commodities, crude oil, gold, copper, positioning, inflation TL;DR: Combined net longs across 25 major commodity futures fell 77% in two months while the index retraced 10% — that gap says the selling came from positioning, not from supply and demand. ![An emptying trading floor at dusk, long rows of desks receding into depth, a single screen still lit with a falling curve](/covers/macrovoices-2026-11-19-macrovoices-559-ole-hansen-what-s-next-after-the-s-cover.png) > Fortune in war is never ours to time; > to swallow shame and bear it is a man. > East of the river there are gifted sons — > who says they could not ride back on the dust? > > —— Du Mu, "At the Wu River Pavilion" (Tang, c. 9th century; my own translation) On MacroVoices episode 559, recorded November 19, 2026, Saxo Bank's commodity chief Ole Hansen laid out two numbers that refuse to agree. In the US futures positioning data through the week of November 10, combined net longs across 25 major commodity futures had collapsed 77% in two months, while the Bloomberg Commodity Index retraced only about 10% over the same stretch. His read is that the selling came from hedge funds cutting exposure to follow the trend, with physical supply and demand holding up — which makes this a coiled spring rather than a broken market. He named the condition too: it takes a shift in the technical or the fundamental setup to turn, and he does not expect that before year end. For 2027 he still looks for gains in commodities, below this year's 25%. ## A Danish Boy and His Sundays Asked about the 1970s commodity super cycle, Hansen said he was too young to follow the end of Bretton Woods. What he remembers is narrower and sharper: in Denmark, Sundays came without cars, because fuel prices were surging and the country had to save. Losing a day of movement each week is the kind of memory that lives in the body, not in a chart. The second cycle he counts is the first decade of this century, when China and India arrived as demand forces at the same time. The third is now. He admitted the start date is a little convenient — measured from the pandemic recession low in 2020, prices are up roughly 150%, and that figure already includes the recent correction. His name for this one is the energy transition and electrification. Cycles are easy to label afterwards, so he added a reason that does not depend on demand at all: commodity pricing is half a supply story. Companies have spent several years moving from just-in-time to just-in-case, having learned what a missing component does to a production line. Deglobalization is changing how firms and nations hold inventory, and that runs on its own clock. ## What a London Decade Taught Him to Watch Hansen spent a decade at a hedge fund in London, so he knows the mechanics from inside: they buy into strength and sell into weakness. Short, medium, and in some cases long-term moving averages now point lower, so funds trim longs and add shorts. They are doing exactly what the manual says — and this time it dragged the market's aggregate position back to where it started the year. Hence the gap. Positioning down 77%, the index down 10%; no single "demand is deteriorating" story covers both. What does cover both is this: the sellers were not selling because they turned bearish on copper or wheat, they were selling because the line was going down. Hansen calls that a potential reverse indicator, and then says the most honest thing in the episode — be patient, it will not happen before year end, next year is the setup. ![Two lines start from the same point: net long positioning drops steeply to barely a fifth of where it began, while the index runs almost flat, leaving a wide gap in height at the right-hand end.](/figures/positioning-vs-index-gap-en.svg) He kept the conditions visible. A reversal needs the technical or the fundamental to move first: the technical is price drawing its own pattern, the fundamental could be inflation rolling over and attention to rate hikes fading, especially if growth takes a hit and the Fed has to look at its dual mandate. None of that has shown up. He is describing a spring that is loaded, not one that has released. ## He Has to Fall Out of Love Before Gold Bottoms Turning to gold, Erik Townsend joked about the Fibonacci tool: back when the charting packages shipped it, did anyone imagine drawing a range $3,800 tall on gold? Hansen's answer was that finding major support meant going back that far — back to the low of 2023, the last floor before central banks, ETF investors and Chinese buyers all decided together that they wanted gold. Then came the line I keep returning to: precious metals is a market you have to fall out of love with before it finds a bottom and rebounds. He traded his first gold below $300 an ounce around 1998 or 1999, so the sentence grew out of two full tides. On the numbers: precious metals gained 80% in 2025 and are down 14% this year on his measure, which covers gold and silver only. Judged on this year alone it looks bad; stretched across two years it is a correction after a run. ![A single line rises sharply, then eases back a little, and even after that pullback it sits far above where it started.](/figures/precious-metals-two-years-en.svg) Silver is worse — down a third this year, with the gold-silver ratio back above 81 after trading in the low 70s earlier in the year. He is lukewarm on it, and the reason sits away from supply. The deficit case has been argued for years, he said, while price gets set by the level of financial investor interest, and that interest has faded while high prices challenged industrial demand. Volatility past a certain point forces you out even when your view holds. Platinum's break above a thousand dollars last year he called one of the most beautiful breakouts of his career, taking out a seventeen-year downtrend; over five years the annualized returns are 13% for silver and 15% for platinum. ## The Week Copper Lost Its 200-Day Days before the recording, copper broke below its 200-day moving average and dropped like a rock, falling close to 20% over two months to around $5.74 per pound. The story the market told itself was that the AI capex boom was about to bust and data center power demand would evaporate. Hansen does not accept that chain. Short term, he said, this is psychological: the Nasdaq was falling, they recorded ahead of Nvidia's results, and pressure travelled from the tech theme into copper, squeezing out speculative longs that had built up. The world is still electrifying, and high energy prices from the Middle East war have accelerated EV adoption. More to the point, supply was already struggling to meet copper demand, and mining input costs keep climbing. The detail I found more useful than the conclusion runs the other way. Exchange-monitored copper stockpiles now exceed a million tons, and 67% of that sits in COMEX warehouses inside the US; a sizeable chunk of the 30% monitored by the LME is also held stateside. That copper moved there to get ahead of tariff risk. If no tariff arrives, some of it can find its way back into the global market and relieve the tightness showing up in China, where inventories are very low. He lists AI and tariffs as the two risks inside an otherwise bullish view, and this is the less-counted of the two. ![A single stacked bar of stocks: the largest segment on the left sits in COMEX warehouses, part of the London segment is also held in the US, a brace above groups these as inside the US, and an arrow below points to metal flowing back to world markets.](/figures/copper-stocks-moved-to-us-en.svg) ## Diesel Is the Line That Matters Energy delivered a 74% total return this year, while flat price movement accounts for about 45% — the nearly 30% in between came from backwardation, the premium in nearby contracts that pays you each time you roll forward. Hansen calls backwardation a long commodity trader's best friend, though the curve is less steep than when he last appeared. The counterexample is natural gas: down about 25% this year, with prices a year out 24% above spot, a contango that makes it brutal to hold long. ![Two futures curves side by side: the left one slopes down from the front month to the later months, and the right one slopes up from the front month to the later months.](/figures/backwardation-vs-contango-en.svg) The line he watches is diesel. His reasoning is direct: consumers do not pay for crude, they pay for diesel and jet fuel, so the difference between crude at 110 and crude at 90 matters less than what distillates cost. Diesel cracks still hover above $160, while gasoline cracks have collapsed back to pre-war levels in both Europe and the US, since this is the low demand season. He made the refining constraint concrete. One barrel of crude in does not give you one barrel of distillate out; the split is fixed, and the bulk of the barrel is currently not earning what a refiner needs. At that point a refiner considers cutting runs, however pretty the diesel crack looks. Maintenance has been deferred because every plant that could run has been running flat out, which makes ramping up harder later. Roughly 10% of Middle East refining capacity is missing, Russia has losses too, and recovery is slow. Europe let old refineries reach the end of their lives without the appetite or the permits to build new ones, grew reliant on Middle East supply, and then watched that supply get cut. Hansen's estimate: with Brent trading 60 to 70 before the war, 70 to 90 is the likely new normal afterwards, still carrying some backwardation. When Erik pushed on whether crack spreads have a new permanent regime — blown-up refineries take a decade to replace and nobody wants to fund one — Hansen agreed with the worry and said plainly that he has not seen anyone work the analysis through. ![A barrel is cut into two parts: the smaller part on top is diesel and jet fuel, the larger part below is the bulk of the barrel that is not earning the money you want, and an arrow on the right points to run rates coming down.](/figures/one-barrel-fixed-ratio-en.svg) ## A Million Contracts, Come and Gone Agriculture in July and August was the same story in another costume. Black Sea supply risk plus heat waves across several regions lifted corn, soybeans, wheat, rice and sugar, and speculative positioning in six soy and grain contracts jumped from flat to a million contracts. It is now back to roughly flat. Corn flipped from a record long to a net short, and wheat has been net short for the best part of four and a half years. ![A curve rises high from near the zero line and then falls back to it, like a hill that came and went.](/figures/grain-positioning-round-trip-en.svg) This year's El Niño is near record strength and has not produced a grain price spike, which Hansen credits to global stockpiles entering the period in robust shape, rice included. He offered one view that sits outside the price screen: a 100% rise in crude is something you discuss in terms of the economic outlook, while a 100% rise in rice and wheat is a food security matter. The vulnerable link he named is sugar, dependent on Brazil, India and Thailand, and already up strongly earlier in the year. ## The News Is Still Good — So Why Does My Position Keep Falling? I have been in this spot. Nothing in the case has broken, the story still holds, and the account shrinks every day until you start hunting for what you missed. This episode gave me a usable split: break a price decline into two separate questions — who is selling, and has supply and demand changed? Positioning data answers the first one. Combined net longs down 77% against an index down 10% says the sellers were adjusting exposure, rather than marking down a tightness that eased. The other side looks different: inventories building, physical premiums flipping to discounts, downstream buyers pushing prices down. Hansen read copper exactly this way — mine supply struggling and input costs rising as his supply-demand anchor, elevated speculative longs and a falling Nasdaq as his positioning anchor, and short-term pressure assigned to the second. ![On the left a falling price line splits into two arrows pointing to two cards, one asking who is selling and the other asking whether supply and demand changed, with small bars inside pointing down in one card and up in the other.](/figures/who-is-selling-vs-supply-demand-en.svg) What I took from it is a habit: next time something I hold is falling and I cannot name the seller, I should admit I am looking at an answer to the first question without the material to read it. When the material is missing, doing less beats guessing right. ## It Fell 20% — Do I Step In? The other familiar bind: something is down nearly a fifth, it looks cheap, and you have no idea whether another fifth is waiting. Hansen gave no answer here. He gave three conditions, and all three can be written down in advance. The first is the trigger. He wants a change in the technical or the fundamental setup, and he put "not before year end" in front of it. That sentence earns its keep by being falsifiable — a rebound before year end would mean part of his read on the drivers is wrong. The second is leverage. He split the same judgment across two kinds of holder: the unleveraged can look through to the longer term, while the leveraged have to address and reduce positions, and those forced sales are themselves weighing on price. The third is the time window. Patrick Ceresna's copper spread in the same episode carries timing as its principal risk, in his own words — copper can stay oversold while the option clock runs out. ![Two horizontal bars of very different length: the upper one extends into an arrow on the right to show that oversold conditions can persist, the lower one is much shorter and shows the time left on the option, and a dashed line marks where time runs out first.](/figures/oversold-outlasts-the-clock-en.svg) My own clumsy version is to write, before buying, "if X has not happened in three months, I accept this reason does not hold," where X is something observable rather than a feeling. Hansen's X is a technical or fundamental turn; Ceresna's X is a rebound inside six weeks. Both men put the failure condition up front, separately from whether they were bullish or bearish. ## The One Thing Worth Taking With You The idea I am keeping from this episode: when something is falling, ask who is selling before asking whether the world changed. Those two answers can point opposite ways, and anyone who blends them will read someone else's forced selling as evidence that their own judgment was wrong. Here is something I have tried, if you want it. Pick one thing that has recently pulled your mood down — money need not be involved; maybe someone has gone quiet, a group chat has cooled, a project stopped coming up. Take a sheet of paper and write two versions of the cause. On the left: their circumstances changed (busy, tired, occupied by something else). On the right: the thing itself changed (they stopped wanting it, it is over). Then hold off on the conclusion and go looking for one concrete sign that separates the two: are they like this with everyone? Were they like this the last time they were exhausted? If you cannot find that sign, what you have in hand is not enough to answer the right-hand column.