In its first full month, Kalshi's 15-minute gold market outtraded Ether by 70%

If gold were drawing traders out of Ether, Ether's count should have sagged after the gold series opened in August ^[11]. It rose. Kalshi's 15-minute Ether contracts grew from 6.1 million in January to 233 million in July ^[12], and to 318 million in September ^[1]. Ether gained about 36% between July and September, the same stretch in which gold went from no contracts to 542 million ^[16]. The sources do not include an August Ether figure, or any data on whether the same accounts trade both series.

On Predict Charts' estimates, gold also earned slightly more per contract. About $5 million over 542 million contracts is roughly 0.92 cents each, while Ether's $2.6 million over 318 million is about 0.82 cents ^[17].

One plausible source of that gap is the fee curve. Kalshi says its charges depend on a contract's expected earnings, and Predict Charts says the effective rate peaks for contracts near 50 cents and falls as prices move toward either extreme ^[6]. InGame found 15-minute markets were 13% of Kalshi's volume in the week to Oct. 5 but 20% of its fees ^[13], so each dollar traded in them paid about 1.5 times the platform's average rate ^[18]. Daniel O'Boyle of InGame wrote that the difference arises "because Kalshi uses a fee formula that depends on the odds of a contract," and that "fees are higher as a share of volume on contracts priced at close to 50/50 odds than they would be on the biggest favorites or longshots" ^[7]. Short-duration price markets often trade near even odds, since traders are only deciding whether an asset rises or falls over a few minutes ^[8].

If the $20.4 million that 15-minute markets earned in that week was 20% of all fees ^{[5] [13]}, Kalshi's total estimated take for the seven days was roughly $100 million ^[19]. All non-sports markets together produced about $25.1 million of it ^[5], leaving roughly three-quarters to sports ^[20].

Kalshi's own commodities figure uses different units. The company said on Sept. 8 that commodities had reached $400 million in cumulative trading volume within seven months, roughly half the time its crypto category needed ^[9]. It said commodities also had more than four times crypto's volume at the same stage ^[9]. The $400 million is a dollar total for a whole category through early September; the gold figure counts contracts in a single month, so neither can be checked against the other. "Crypto markets demonstrated the potential for new categories on Kalshi to scale from tens of millions to billions in monthly volume," the company said ^[10].

One possibility is that gold is additive, and the 15-minute format becomes a template Kalshi repeats on other commodities. Another is that gold and Ether compete for the same short-horizon traders, in which case Ether's count flattens or falls while gold keeps climbing. A third is that September was a launch-month peak and gold settles lower. I think the first fits the record best, because Ether kept growing through the months in which gold launched. An October in which Ether trades fewer than 318 million contracts while gold holds above 542 million ^[1] would support the cannibalization case and show the additive view to be wrong.