Bernstein's $10 trillion prediction-market call rests on half a percent of a $900 trillion pool

The upgrade is mostly a longer runway. In April the same firm had volume going from $51 billion in 2025 to $1 trillion by 2030 ^[5]. That compounds at about 81% a year; run the rate five years further and 2035 lands near $19.6 trillion ^[16]. The new number is $10 trillion ^[1].

Bernstein puts the addressable market for financial-asset contracts at about $700 trillion in 2025 and $900 trillion by 2035 ^[11]. Capturing just 0.5% of the 2035 pool, it says, yields roughly $4.7 trillion of annual volume from financial contracts alone ^[12]. Gross that up by the 49% mix assumption ^[2] and the total is about $9.6 trillion ^[18].

In dollars, the category that is supposedly being overtaken grows hard. Sports at 61% of roughly $50 billion in 2025 is about $30 billion; sports at 38% of $10 trillion is $3.8 trillion, or about 125 times as much ^[20]. Financial contracts go from about $6 billion to $4.9 trillion over the same span, near 800 times ^[26].

The 2026 book is narrower than the forecast implies. Industry volume ran about $300 billion through August against about $50 billion for all of 2025, driven by short-dated crypto and commodity products including 15-minute Bitcoin contracts ^[6]. Kalshi held nearly 60% of that ^[7], or roughly $180 billion ^[21]. Its commodity business, the one that has to carry the thesis into oil and metals, was about $590 million year to date with $410 million of it printed in August ^[9]. That is a third of a percent of Kalshi's volume, and about 70% of its commodity total in one month ^[21][22]. The $410 billion full-year estimate for 2026 implies $27.5 billion a month from September through December against $37.5 billion a month for January through August ^[23].

Bernstein said the product pipeline does some of the lifting, citing "KPI markets, where users can trade single metrics for a company, such as production, deliveries or subscriptions, rather than its share price" ^[14]. It also expects perpetual futures to spread out of crypto into commodities and individual stocks ^[15]. That is a plausible description of what a venue can list; who trades it is a separate question.

So the number to test is the 0.5%. The mix, at least in part, the 2026 flow supports. Prediction markets would be taking that share from venues that already intermediate the same risk, and 15-minute contracts turn notional over fast enough that a venue can report large volume off a thin set of end users. If the short-dated crypto contracts are genuine retail substitution for perps, the penetration figure is conservative and the $900 trillion pool is the right denominator. If they are market-maker churn, the $300 billion is turnover and the denominator is wrong by an order of magnitude.

The regulatory line sets the timing either way. Bernstein expects US sports prediction markets to get firm rules no earlier than 2027-2028, because courts disagree on whether the contracts are federal derivatives or state-regulated gaming ^[13].