Who is right about Bank of America at $51: Wall Street's $67 or the model's $36?

The $1.2 billion fine hid the real number: $8.3 billion out, $6.9 billion in

Bank of America ($BAC) reported on 14 October 2026 a profit of $0.91 a share for the quarter to September 2026. Strip out the $1.2 billion regulatory settlement charge ($0.17 a share) and adjusted earnings were $1.08 a share on revenue of $30.0 billion, against estimates of $1.10 and $30.58 billion. The first miss after four straight beats sent the stock down 4.8% the next day.

The charge is a closed chapter. On 13 October 2026 the U.S. Treasury announced penalties of more than $1.21 billion, including the largest fine in OCC history, for failing to report over $5 billion of suspicious transactions. It is not tax-deductible, so every dollar came straight out of capital.

The structural news sits one line lower. The bank paid $2.2 billion in dividends and bought back $6.0 billion of stock, $8.3 billion in total, while net income was $6.9 billion. CET1 capital - the core equity regulators count - fell $3 billion to $198 billion, including a $1.8 billion hit from unrealised losses on bond holdings as long-term rates rose 0.85 percentage points in the quarter. FICC trading revenue fell 17% to $2.6 billion, and Global Markets earned a 9% return on allocated capital, down from 20% in the quarter to June 2026.

From 1 January 2027 the capital cushion shrinks to 0.4 percentage points

The CET1 ratio ended the quarter at a preliminary 10.9%, down from 11.2% in June 2026 and 11.6% in September 2025. The regulatory minimum is 10.0% today. Under current rules the G-SIB surcharge - the extra capital the largest banks must hold - rises by half a percentage point on 1 January 2027, lifting the minimum to 10.5%, as the 10-Q for the quarter to June 2026 sets out. That leaves 0.4 percentage points of room, roughly $7 billion, for a buyback that spends $6 billion a quarter.

That buyback is the engine of per-share growth. Shares outstanding fell from 7.33 billion to 6.92 billion in twelve months, a 5.6% reduction, while adjusted EPS rose from $1.06 to $1.08 - 2%. The whole per-share gain came from the share count, not from profit. The dividend, raised 14% to $0.32 a share on 24 July 2026, takes another $2.2 billion a quarter.

Management's answer is rates: on the 14 October 2026 call it guided net interest income growth of about 8% for 2026 and 8% or more for 2027, assuming three further quarter-point hikes in December 2026, March 2027 and June 2027 after the Fed's first increase since 2023 on 16 September 2026. Rates one percentage point above the curve add $1.4 billion of NII over twelve months. But the same hikes marked down the bond book by $1.8 billion and hurt mortgage and municipal trading. It is a landlord raising rents while the appraiser cuts the value of the building.

Why the model stops at $36.18 while the price is $51.22

The FPI fair value is $36.18, which puts the price of $51.22 29.4% above fair value. The blend is 50% discounted cash flow, 30% peer multiples and 20% analyst consensus. Since the consensus leg sits at $65.92, the cash flow and peer legs must land well below $36 for the blend to end up where it does.

A bank earns a multiple of its book value in proportion to its return on equity, and $BAC's ROE of 10.7% is slightly below the industry median of 10.9%. A peer-level return earns a peer-level multiple: the industry trades at 10.8x earnings, $BAC at 11.2x. The price is 1.3x book value of $39.26 and 1.76x tangible book of $29.14; the fair value sits just under book, where a bank earning roughly its cost of equity normally trades. The 26% free cash flow yield in the metrics is a bank accounting quirk - deposit inflows count as cash - and not a buy signal, as the bank valuation guide explains.

Street vs model: a $67 median, a $36 model, and a price sitting in the middle

Wall Street counts 35 buys, 18 holds and 1 sell. Targets run from $59 to $75 with a median of $67, 31% above the price - while the model sits 23% below it.

The targets are drifting down. Targets issued in the quarter to 15 October 2026 averaged $65.50; those of the last month average $64.86. On 15 October 2026 Morgan Stanley set $61, RBC $60, Wells Fargo $64 and Barclays $68, after Redburn's $72 on 6 October and UBS's $67 on 5 October. TD Cowen cut to $56 with a hold the same day, trimming EPS estimates to $4.49 for 2026 and $5.05 for 2027. Even that lowest fresh target is 55% above the model.

The gap has one source. The Street capitalises 2027 earnings near $5.05 a share - $67 is 13.3x that figure - and assumes the $24 billion-a-year buyback keeps lifting EPS. The model anchors on today's 10.7% ROE and $39.26 of book value and gives no credit to buybacks the 10.5% floor may not fund. Insiders made 0 purchases and 7 sales worth $31.1 million in the twelve months to October 2026.

Three numbers before 15 January 2027: 10.5%, $6 billion, $55.20

The stock lost 16.4% in the three months to 16 October 2026 against a 2.7% gain for the S&P 500, and an RSI(14) of 24 says the selling is stretched.

Re-rate up: the Fed delivers the December 2026 hike management assumes, fourth-quarter NII rises sequentially, CET1 holds at or above 10.9% with the buyback at $6 billion, and the price reclaims the 200-day average of $55.20. Re-rate down: the buyback drops below $6 billion in the quarter to December 2026, 2027 NII guidance slips under 8%, FICC falls again, and the price revisits the 52-week low of $46.12.

The checkpoints are dated: FOMC decisions on 28 October and 9 December 2026, the G-SIB step-up on 1 January 2027, and fourth-quarter results on 15 January 2027, where the Street expects EPS of $1.14 on revenue of $30.59 billion. A $6 billion buyback with CET1 still at 10.9% would show the capital math works; anything less and the $67 targets lose their engine.