Every other category in the CS2 skin market moves together, to a meaningful degree. We measured the day-to-day correlation between daily percentage price changes across six categories, knives, gloves, stickers, weapons, cases, and everything else, over roughly 1,250 trading days. Knives and stickers move together with a correlation of 0.48. Weapons and stickers, 0.58. Knives and weapons, 0.46. Knives and gloves, 0.37. These are moderate-to-strong relationships: when the broader market has a good or bad day, most categories feel it together. Cases don't. The correlation between cases and every single other category, knives, gloves, stickers, weapons, everything, rounds to 0.00. Not low. Zero.
Why that looks interesting
A category with genuinely zero correlation to the rest of a market is, in finance terms, exactly what a portfolio wants: something that doesn't crash when everything else crashes, and doesn't need the same conditions to hold value. If cases really behaved independently of the broader CS2 market's ups and downs, that would be a real, useful, differentiator, a category worth holding specifically because it doesn't share the market's risk.
Why it isn't
The catch shows up the moment you look at how often the price actually moves at all. Across every other category, the case market is flat, literally zero percent change, on about 1% of days. Cases are flat on 11.7% of days, nearly twelve times more often. The median daily price move for cases is also the smallest of any category, 0.5%, smaller even than gloves. Put those two facts together and the "zero correlation" stops looking like independence and starts looking like something else: a market that mostly isn't moving because there's barely anyone left to move it. You can't correlate with something that isn't happening. A category trading this thinly will show close to zero correlation with everything else almost by construction, not because it's hedged against market-wide moves, but because there usually isn't a price move to correlate in the first place. This lines up with something we've written about before: the Cases category has already lost 99.7% of its value from its own historical peak (May 2023, when it was worth $708K, down to roughly $2.3K now). A category that's lost nearly everything and trades on a handful of scattered listings a day isn't sitting outside the market's ups and downs by design. It's a market that has mostly stopped functioning, and what little price data survives looks statistically disconnected from everything else simply because there's so little of it.
The general lesson
This is a useful trap to know about beyond cases specifically: an illiquid category will always look "uncorrelated" with everything else, right up until it isn't a category you can actually buy or sell in any real size. Zero correlation is genuinely valuable when it comes from an asset that trades actively and simply responds to different underlying drivers. It's not valuable, and can be actively misleading, when it comes from an asset that's barely trading at all. The math can't tell the difference on its own; you have to check the liquidity behind the number before deciding what it means. If you're looking at correlation data anywhere in the CS2 market, or anywhere else, the volume and flat-day rate behind that number matters as much as the correlation figure itself. A 0.00 correlation on a category with normal trading activity is a real signal. A 0.00 correlation on a category that's flat one day in nine is usually just a market that's gone quiet.
