The Devil is in the Details - COT Data, More than Just a Net Position

Last week's giant selloff put trader data among the top headlines, but simply looking at weekly net positions fails to tell the whole story.

Three Point Summary

1. This regime appears to be characterized by a structurally lower net long position and a more balanced mix of long and short bets. WTI has shown the most drastic change, with the L/S balance falling toward historical lows.

2. That means analyzing spec trading requires more consideration of the details and composition of the net position. We can no longer just track noncommercial changes over time. Under the surface, we see interesting positions with the number of traders holding longs historically high but with a smaller position. Conversely, for the size of the short position, the number of traders holding shorts appears small. This creates an interesting balance of risks, which probably contributes to volatility and muddies the message in directional moves.

3. The data in products is a lot noisier, with a selloff last week that gave a headline that matched crude sentiment, but underlying data that demands more scrutiny. Gasoline positioning has shifted wildly shorter after being record long, gasoil is as negative as crude, and ULSD’s selloff followed what looked like an improving seasonal pattern. Overall, though, watch how interest in products wanes as cracks tighten back toward what was considered normal before the pandemic.

Oil Positioning in a New Regime

In recent weeks, weekly CFTC reports have garnered increased attention, and with good reason. The financial/forward market has significantly influenced this regime, especially considering the longer transit times as we optimize supply chains for risk. It’s important to resist the temptation to draw direct connections to net futures and options positions changes and price changes. (I’ve written about how this data is often manipulated to support an analyst’s opinion, so I won’t rehash that here.) We can, however, look at the recent trends to see how they fit into the post-pandemic/post-war regime because they are starkly different and have important implications.

Last week alone, the net position shifted shorter by >123mm barrels Brent/WTI combined. That extended a run of 7/8 weeks of net selling that has dropped the net futures position from nearly in line with the average of the previous three years to making headlines for being the shortest of the last decade.

Overlaying the average position of the three years leading up to the pandemic shows the stark difference between this regime and the past. Traders have carried much less length in aggregate than before, even as open interest slowly rises toward historical highs.

That’s extended the long-term trend, which can be better seen in WTI. Starting in 2016 (when the US lifted export restrictions), WTI had been a structurally long play. With increased access to international markets, the net length expanded wildly, eventually peaking with the volatility of the pandemic (and record low prices). Since then, it’s been all downhill. While still structurally long and much longer biased than Brent, the trend has retreated toward historical lows. The ratio of open longs to shorts recently dipped below 2 for the first time in ~8 years, finding a low at 1.9, the lowest since August 2016.

WTI, by the way, was trading in the mid-$40s in August 2016.

Trader Concentration Defining the New Regime

The makeup of traders has changed even more than that aggregate number suggests, which may have implications for directional volatility that aren’t covered in the weekly net change. For example, 164 traders hold long futures positions in Brent, a top decile collection of traders. One would expect the total long holdings to be closer to 600k contracts than the current 417k at that level. Meanwhile, 103 traders held short positions last week, also toward the top of the range but not quite as extreme. What stands out in that position is that the short position of ~575k lots is about 140% of what you’d expect based on historical averages.

That creates an interesting dynamic. There are many long traders, so it’s possible that there aren’t many left to join the party. At the same time, the average position is small, so the traders at the part might be able to carry more risk. Conversely, there may be room for the number of short traders to rise, but there may not be much additional capacity to carry risk among those already short. The large average short position also creates the risk of lumpy liquidation. That forces us to dive into the details to interpret the weekly changes.

Last week’s record move saw a total of >44k contracts of long futures positions closed and just under 20k contracts of short futures opened, a combined change in futures of ~64k contracts. That aggregate move, though, was reached with longs closing positions despite new traders joining (+12 long futures holders), combined with a rising total short position with fewer traders holding shorts (-4 short futures traders).

There is no way to guess how this unwinds, but the more concentrated short position appears more sensitive to rising volatility. In other words, a few short traders with large positions may have a short-term outsized influence on the market if forced out quickly. More likely, they trim positions over time and contribute to a floor price as they cover their position. At the same time, with a historically high long trader club, an extended bear run has plenty of kindling to catch fire.

Refined Products Join the Club

With open interest rising in refined products trading, the contribution to sentiment is important to consider, and the recent negative sentiment seen in oil is mirrored in gasoil. Over the last four weeks, gasoil specs have sold 12.8k net futures contracts, extending the streak of rolling monthly short moves to eight weeks and moving the net position from short ~90k contracts to short ~168.5 contracts, driven entirely by the Other Reportables’ net short 189k contracts. This puts gasoil futures at the shortest net position on record.

Stateside, NYH ULSD futures have rebounded from their lowest net long position of the year, even though last week saw substantial selling. Despite last week’s >9k contract net-short move, the last month has seen the total ULSD position rise by just over 5k contracts in what looks like the beginning of a seasonal recovery.

That brought the position back in line with the average level of the last three years, though still well above the pre-pandemic norm. Additionally, traders were less committed to the move than oil traders. The net short move was reached through new short positions outnumbering new longs ~3:1, despite fewer traders on both sides of the ledger holding positions.

The most drastic move among the three major refined products contracts came in RBOB Gasoline. At the beginning of April, RBOB traders held more futures net long than at any point since January 2021. They’ve since cut the position in half, selling a net ~37k contracts to trim the position to just 46k net long. Like diesel, this brought the position back in line with recent norms, but doing so with much more aggressive selling.

Traders have closed longs for eight consecutive weeks, totaling >38k exited positions, reducing the ratio of long/short contracts from 3.6 to 2.2. Over the period, the number of traders holding long futures contracts has dropped from >100 to as low as 83 (recovering to 87 last week), while the number holding short futures contracts ticked down from a high of 45 to 31 in the most recent week, leaving the gasoline trading relatively under-concentrated, in contrast to Brent.

With all this activity, total open interest across these three products has recovered to >1.7mm contracts, consistently working higher toward the ~2mm contract high first set in 2018. That interest had fit nicely into the post-pandemic regime, as crack spreads were wildly volatile and set new highs consistently. It remains something to watch. Notably, as cracks have deteriorated, the focus (as measured by open interest) has shifted to oil. The combined open interest of Brent and WTI has risen faster than that of refined products, pulling the open interest ratio back in line with pre-pandemic levels.