The numbers behind the warningFor months, business groups have warned of a "Pymes Power Cliff" using owner surveys. Now the utility's own filing with the regulator puts hard numbers under it: average electricity bills for the more than 75,000 commercial accounts LUMA uses as a proxy for Puerto Rico's small and midsize businesses jumped 19% in a single month, from $579 in September to $691 in October.^[1] Consumption barely moved - up 2.5%, from 145.6 million to 149.3 million kilowatt-hours - while total billings rose from $43.5 million to $51.9 million, an increase of more than $8 million in one month on essentially the same amount of power.^[2] That gap points to price, not usage: the Energy Bureau's temporary 5.28-cents-per-kilowatt-hour fourth-quarter adjustment alone accounts for roughly $7.9 million of the jump, implying an all-in cost of about 34.8 cents per kilowatt-hour in October, up from 29.9 cents in September. The so-what for the 75,000Three numbers frame the affordability question the Energy Bureau now faces. First, $4.6 million of these accounts' balances are already more than 90 days past due - arrears building before the winter quarter plays out.^[3] Second, the Centro Unido de Detallistas' warning that up to 12,000 small firms could close, putting about 70,000 jobs at risk, now has an official billing dataset behind it rather than member anecdotes alone.^[4] Third, the governor's announced $3 million in energy aid would reach roughly 1,000 businesses at $3,000 each - about 1.3% of the 75,000 accounts, and less than five months of the average October bill. The math argues for treating bill reduction as the primary policy lever rather than after-the-fact aid: a quarter-point cut in the adjustment reaches every account automatically, while a $3,000 grant covers a fraction of the firms most exposed. V2A Consulting's estimate that a 20% further increase in power costs would remove about $935 million a year from the economy puts the stakes in context of an economy already running flat.^[4] What the filing cannot yet say matters for the next decision: LUMA lacks the revenue and employment data that legally define a Pyme, so regulators cannot yet see which of the 75,000 accounts are the firms actually at risk - or whether aid is reaching them. Account-level data tied to size and sector would show where the cliff is steepest.^[1]