Southern Co. Posts Q3 Beat, But the Street Stays Cautious

SO reported Q3 results on November 5 with the stock essentially flat on the week against peers that have moved slightly more, a gap worth examining given the 12% run the stock put in over the prior month.

Among its closest peers, the positioning picture varies. LNT added nearly 3% on the week, EXC rose 1.8%, and DUK gained 1.4%. SO's 1.2% weekly gain trails that peer median of roughly 1.4%. None of those names carry meaningful short pressure either, but SO's borrow market is the loosest of the group by a wide margin. Availability is at 2,609% of short interest, meaning there are more than 26 shares available to lend for every one already borrowed. That compares to a 52-week low availability of 783%, itself still exceptionally loose. Short interest is 2.7% of free float, edged down about 1% on the week, and the cost to borrow is 0.44%. The lending market carries no tension whatsoever.

Options positioning tells a different story. The put/call ratio has remained elevated at 0.81 through the week, well above its 20-day average of 0.72 and close to the 52-week high of 0.89. That is roughly 1.4 standard deviations above the recent mean, the most defensive options configuration SO has seen in a year. The timing matters: the Q3 print landed on November 5, and the hedging cluster built directly around that date. With the event now past, it will be worth watching whether the put/call ratio normalises or stays bid.

The analyst picture is settled but not comfortable. The consensus is Hold, with 12 analysts at that rating. The key fault line is between Morgan Stanley, which holds an Underweight with an $84 target, and HSBC, which initiated with a Buy and a $99.50 target on October 22. Between those poles, most of the Street spent October trimming targets without changing ratings. JP Morgan, Barclays, Truist Securities, and Jefferies all cut, with Barclays lowering to $92, now below the current $93.75 close. The mean target sits in the mid-$90s, implying limited upside from here. EPS surprise ranks in the 60th percentile, so the company does beat estimates reasonably consistently, and forward EPS growth for 2026 to 2030 is forecast at a 7.6% CAGR, which the bull case leans on heavily. The bear case is simpler: a PE of 19.3x and EV/EBITDA of 12.1x represent a relatively full valuation for a regulated utility facing potential regulatory tightening in Georgia if the PSC shifts Democratic.

Insider activity is routine rather than directional. All recent open-market sales are small, the most recent being 100 shares by the Comptroller on November 2 under a 10b5-1 plan. Net insider selling over the past 90 days is about $842,000 in value, but nearly all of it reflects pre-scheduled plan sales. There is no discretionary buying or selling at scale.

The reactor output signal from ORTEX Alt Data is worth a note of context, not a prediction. Nuclear output at SO's units registered a z-score of minus 10 against its own trailing history as of October 28, the deepest reading in the dataset, reflecting a significant outage or curtailment period. The dataset has not yet accumulated enough history to test whether it leads reported revenue, so no causal inference is warranted. It is an operational fact that the Q3 print will either clarify or contradict.

The question the November 5 results will answer is whether the demand story from data centres and industrial customers is translating into earnings growth fast enough to justify the multiple, or whether the Street's cautious target-trimming over October was the right read.