# Synchrony: earnings down 18%, EPS down 6%, and a hold at $73.64 SYF, NYSE. Published October 26, 2026. Price as of the October 23, 2026 regular session close. Values in US dollars. ## Verdict: HOLD at $73.64 Synchrony trades 3.7% above a Base value of $71.00 and about 1.55 times book value. The price needs a 17.84% terminal return on equity against my 17.0%, a modest gap, so there is no margin of safety. A 30% margin of safety puts the buy range at $41.50 to $62.26, well below the price, so this is a hold for holders and a watchlist name for new money. - Bear value: $43.70 - Base value: $71.00 - Bull value: $91.60 - Upside to base value: -3.6% - Maximum buy price: $41.49 (range $41.50 to $62.26 by method; a conditional re-entry trigger, not an unconditional order) - Required margin of safety: 30% - Earnings multiple cross-check: $74.83 (1.6% above the price) - Price to common book: 1.55x ($47.64 reported book value a share) - Price to earnings: 8.1x on normalized no-growth earnings of $9.05, 8.7x on year-one Base EPS - Return the price implies: terminal return on equity of 17.84% (Base 17.0%) - Return the Base implies: 11.66% a year over five years, 12.94% at an unchanged price to book - Capital: CET1 ratio 12.9% against a 7.0% requirement - Dividend: $0.34 a quarter declared, $1.36 annualized (1.85% yield) - Business quality: 60 to 75 of 100 (growth and runway not assessable) - Valuation score: 35/100 - Confidence: 21.25/25, grade B In the third quarter Synchrony earned $864 million for common shareholders, down from $1,057 million a year earlier. Earnings per share fell only from $2.86 to $2.68. The gap is the share count: in nine months the company retired 38.3 million shares, taking the total from 347.4 million to 311.4 million. Buybacks are a real part of this story, and they are also the reason "EPS growth" at a card lender needs a second look at the aggregate earnings and the capital behind them. At $73.64 on October 23, 2026, the shares trade 3.7% above my Base value of $71.00, so the Base is 3.6% below the price. The Bear value of $43.70 is 40.7% below the price and the Bull value of $91.60 is 24.4% above it. A deep credit stress gives $26.44, which is a stress test and not a floor. I assign no probabilities to the cases. The required margin of safety of 30% gives a buy range of $41.50 to $62.26 and a conservative limit of $41.49. For existing holders this is a hold. For new money it is a watchlist name. Valuation route: Banks, a deposit-funded consumer credit bank. Economic classification: partial, 8 to 10 of 12, with Quality Growth as the supported floor and Compounder not established. Primary method: five years of common-equity residual income with funded capital distributions, the sector method for a bank earning above its cost of equity. Cross-check: normalized year-five earnings at 9 times. Values are present fair-value estimates, not 12-month price targets. ## What Synchrony is Synchrony is a bank holding company that runs private-label and co-branded credit cards, installment loans and savings products through merchant partners. Five sales platforms report as one operating segment, so a sum-of-the-parts valuation is not justified. Deposits of $83.0 billion fund about 83% of the balance sheet, and loans of $103.3 billion carry an allowance of $10.3 billion, or 10.01% of loans. The business depends on a short list of partners. Amazon, Lowe's, PayPal including Venmo, Sam's Club and TJX generated 54% of 2025 interest and fees and held 52% of receivables. Their contracts run to 2030 to 2035, and 22 of the 25 largest relationships expire in 2028 or later. Long contracts are evidence of a durable franchise, not a promise that the economics stay the same at renewal, because the merchant owns the customer relationship at the checkout. Newer programs (OnePay with Walmart, Amazon Pay Later, the Lowe's commercial portfolio, Versatile Credit) extend the same activity and do not change the risk class. They are also too young to show their eventual returns, so I model them and do not give them credit for results that do not exist yet. ## The latest quarter | Third quarter 2026 | Result | Comment | | --- | --- | --- | | Common net income | $864 million | vs $1,057 million a year earlier | | Earnings per share | $2.68 | vs $2.86; buybacks softened the decline | | Net charge-offs | 4.79% | nine months 5.21%; long-run range 5.5% to 6.0% | | 30+ day delinquencies | 4.34% | 4.39% a year ago, 4.16% in Q2 | | 90+ day delinquencies | 2.06% | 2.12% a year ago, 2.01% in Q2 | | Purchase volume / loans | +8% / about +3% | average active accounts about unchanged | | CET1 ratio | 12.9% | requirement including buffer 7.0% | Credit looks better than the long-run range, but delinquencies rose from the second quarter, and young loans have not finished seasoning. Quarterly return on common equity of 22.9% (on a simple four-times convention) is a snapshot, not a sustainable level, and I do not use it as a valuation input. ## What the earnings are made of I use twelve months to September 30 (common income of $3,242 million), not the third quarter times four. A static normalization with loan losses at 5.75% of average loans, the midpoint of management's normal range, gives lower earnings than the reported figure: | $ million | Twelve months reported | Static normalized | | --- | --- | --- | | Net interest income | 18,813 | 18,813 | | Retailer share expense | -4,287 | -4,161 | | Other income | 501 | 469 | | Operating costs | -5,417 | -5,342 | | Net charge-offs | -5,313 | -5,817 | | Reserve release effect | 69 | 0 | | Income tax | -1,033 | -991 | | Preferred distributions | -91 | -120 | | Common net income | 3,242 | 2,852 | The normalized $2,852 million is $9.05 a share on 315 million economic shares. It assumes no growth. Adding $300 million a year of reserve growth, which a lender needs to grow loans, lowers it to $2,627 million. The explicit Base forecast gives $2,679 million in the first year. Each 50 basis points of loan losses is worth $505.8 million of losses at the twelve-month loan scale. After 25% retailer sharing (my sensitivity, not a disclosed contract percentage) and 25% tax, it reduces common income by $284.5 million, or $0.90 a share. Reported earnings also include one-offs I removed: 75 million of operating cost adjustments and 32 million of net other income, mainly a Visa gain. Cash flow measures from a corporate model do not apply. Operating cash flow of $9,555 million is dominated by banking balance sheet flows, and interest on deposits is already inside net interest income. The relevant questions are what a share earns after credit costs, and how much capital the loan book needs before anything is left to distribute. ## Capital, preferred stock and shares | September 30, 2026 ($ million) | Amount | | --- | --- | | Total equity | 16,550 | | Less preferred carrying equity | -1,716 | | Common book equity | 14,834 | | Company tangible common equity | 13,324 | | CET1 capital / risk-weighted assets | 13,584 / 105,704 (12.9%) | | Common book per share on 315 million shares | $47.09 (reported $47.64) | There are 311.4 million common shares outstanding (October 16) and I use 315 million economic shares, which adds the average award dilution of 3.4 million. The current award inventory is not in the latest filing, so I test 312 to 318 million. The effect is small: the difference between the ends of that range is about $1.36 a share. Three preferred series (A, B and C) carry a combined principal of $1,750 million and annual coupons of $119.7 million, which I deduct from common earnings every year. Series B resets in May 2029 and series C in August 2031. I hold the coupon constant, which is an estimate after those dates. In nine months common book fell from $15,544 million to $14,834 million after $2,878 million of treasury purchases: aggregate common equity fell while book value per share rose, so future per-share gains need funded buybacks, not wishful share counts. Funding is mostly direct certificates of deposit ($43.6 billion) and savings ($29.1 billion). Uninsured deposits are an estimated $7.1 billion (8.6%), the Q3 cost of interest-bearing deposits is 3.69% (4.00% a year earlier) and unsecured notes of $1,600 million mature in 2027. A static test in which 20% of uninsured and 5% of other deposits leave leaves $9.8 billion of liquidity. The severe test (all uninsured and 20% of other deposits) leaves a $7.3 billion gap that would need collections, collateralized borrowing or asset actions. That does not signal distress, but it is why I treat distributions as the first thing to go in a stress. ## Primary valuation: residual income The method values the common equity from its book value plus the present value of profits above the required return, with the cash that can be paid out under a capital rule. I forecast five years ending October 26 of each year to 2031 and discount at a 12.5% cost of equity: a 4.96% ten-year Treasury yield (Federal Reserve, October 22) plus an estimated 5% equity premium times a risk loading of 1.5 gives 12.46%, rounded to 12.5%. The 1.5 is my loading, not a measured regression beta. Base assumptions, all estimates: - Loan growth of 3% a year, net charge-offs of 5.75%, an allowance of 10% of loans and operating costs of 5.35% of average loans. - Retained capital at the current ratio: the higher of $13,584 million and 12.851% of risk-weighted assets, so growth is funded before anything is distributed. - A 25% tax rate, a 1% tax on buybacks, and a constant preferred coupon of $119.7 million. - A terminal return on common equity of 17% and growth of 3% after year five, which gives a justified price of 1.47 times book. | Base case ($ million) | Year 1 | Year 3 | Year 5 | | --- | --- | --- | --- | | Average loans | 104,896 | 111,284 | 118,062 | | Net interest income | 19,506 | 20,694 | 21,955 | | Net charge-offs | 6,032 | 6,399 | 6,789 | | Pretax income | 3,731 | 3,952 | 4,192 | | Common income | 2,679 | 2,844 | 3,025 | | Dividends / buybacks | 428 / 1,825 | 367 / 2,025 | 311 / 2,232 | | Ending common book | 15,242 | 16,094 | 16,998 | | Ending economic shares (million) | 291.8 | 248.7 | 210.1 | | Base value bridge ($ million) | Amount | | --- | --- | | Present value of five years of dividends and buybacks | 8,466 | | Terminal common equity in 2031 | 25,049 | | Present value of terminal equity | 13,900 | | Common equity today | 22,366 | | Per economic share (315 million) | $71.00 | This gives **$71.00** a share. The terminal value is 62.2% of the total, which is material and is why I test return on equity, growth and the discount rate below. Buybacks inside the model are priced at the model's own future value, so a continuing holder gets only dividends and the remaining claim, not a free gain from shrinking shares. If the company pays 20% above or below that value, the five-year return changes from 11.66% to 10.14% or 14.05%. ## Secondary valuation: earnings multiple The cross-check takes the same five-year path and values year-five common earnings of $3,025 million at 9 times, discounted at 12.5%. The result is $74.83 a share. Nine times is a modest alternative to the primary method's own implied 8.3 times trailing earnings in year five, and not a premium for quality. | Method | Today equity ($ million) | Per share | Price to common book | | --- | --- | --- | --- | | Residual income (primary) | 22,366 | $71.00 | 1.51x | | 9x year-five earnings | 23,571 | $74.83 | 1.59x | The two are 5.25% apart, and the whole $3.83 difference is the terminal value, because the interim cash is identical. The methods share credit costs, reserves, tax and the opening share count, so their agreement is not independent proof and I did not raise confidence for it. I do not average them. The residual income value is the central estimate. ## Historical cross-check and growth diagnostic | Year | Common income ($ million) | Diluted EPS | Net charge-off rate | Year-end price to book | | --- | --- | --- | --- | --- | | 2021 | 4,179 | $7.34 | 2.92% | 1.89x | | 2022 | 2,974 | $6.15 | 3.00% | 1.19x | | 2023 | 2,196 | $5.19 | 4.87% | 1.18x | | 2024 | 3,427 | $8.55 | 6.31% | 1.64x | | 2025 | 3,469 | $9.28 | 5.65% | 1.86x | The history is a reminder that earnings at a card lender swing with reserves. 2021 included a $1,578 million reserve release and 2024 a $802 million after-tax gain on the Pets Best sale, so an EPS growth rate over this period measures accounting more than compounding. Common book per share rose from $20.49 at the end of 2020 to $44.74 at the end of 2025, a 16.90% annual rate, or 19.35% with $4.89 of dividends held in cash. That is an endpoint measure, not a market return, and repurchases helped. At $73.64 the price is 1.55 times reported book, a higher multiple than Capital One (1.15x), Ally (0.90x) and Bread Financial (1.17x) on the same date. On growth, Base common earnings grow 3.08% a year from year one to year five, but earnings per share grow 11.66% a year ($8.50 to $13.22), because the share count falls 7.68% a year. The forward price to earnings on year-one EPS is 8.66 times, and the growth-adjusted ratio of 0.74 is a diagnostic, not a fair value, since most of the per-share growth comes from repurchases. Peers have no clean forward growth data, so I do not publish a peer ranking. The classification of growth is mixed: funded operating growth plus accretion from a smaller share count. | Year-five earnings stress | Year-five EPS | Residual income value | 9x earnings value | | --- | --- | --- | --- | | Base | $13.22 | $71.00 | $74.83 | | Year-five income 20% lower, temporary | $10.64 | $69.95 | $64.18 | | Year-five income 40% lower, temporary | $8.03 | $68.89 | $53.54 | | Terminal income 20% lower, lasting | | $60.29 | | | Terminal income 40% lower, lasting | | $49.57 | | The first two stress rows are temporary and the primary method barely moves because the terminal return is unchanged. The last two are lasting damage to the terminal return, which is the case that matters. ## Valuation triangulation I keep the residual income result of $71.00 as the central value and show the multiple result of $74.83 beside it. Three ranges answer three different questions and none is a probability interval: - **Operating cases:** $43.70 to $91.60 (deep credit stress $26.44). - **Method range at Base inputs:** $71.00 to $74.83. - **Selected input and method set:** $59.28 to $88.95 for residual income and $66.36 to $84.08 for the multiple method, from six combinations of discount rate, share count and terminal return. As an outside check, the Federal Reserve's 2026 severe test shows cumulative loan losses of 20.7% over nine quarters for Synchrony, with its CET1 ratio falling to a minimum of 12.5% from 12.6%. It starts from different data and does not include continued dividends and buybacks, so it supports the idea that pre-provision earnings and retailer sharing absorb losses, and nothing more. I found no usable consensus model to blend in, so no analyst target enters this analysis. ## What is the market already pricing in? Solving the Base model backwards for $73.64, one driver at a time: | Solved for $73.64 | Implied | Base | | --- | --- | --- | | Terminal return on common equity | 17.84% | 17.00% | | Net charge-offs in all five years | 5.37% | 5.75% | | Exit price to earnings in the cross-check | 8.78x | 9.00x | These are alternatives, not a market forecast. None needs a heroic assumption, which is why the price is close to my value rather than far from it. The stock is not pricing in a Bull outcome, and it is not pricing a normal credit cycle at a large discount either. For a holder who buys now and receives only dividends, with buybacks at fair prices, the five-year return is: | Five-year holder return | Bear | Base | Bull | | --- | --- | --- | --- | | Share price in 2031 | $72.81 | $119.22 | $155.52 | | Dividends per share, cumulative | $4.90 | $6.80 | $7.53 | | Annual return (IRR) | 1.10% | 11.66% | 17.66% | | Annual return at unchanged price to book (1.56x) | 11.14% | 12.94% | 13.54% | Base returns 11.66% a year, a little below the 12.5% cost of equity, because the price today is slightly above my value. Keeping today's price to book at the exit raises it to 12.94%. The Bear column shows how a weaker franchise and a lower justified multiple can erase the benefit of retained earnings and buybacks: 1.10% against 11.14% at an unchanged multiple. These are conditional outcomes of each case and not a forecast. ## Bear, base and bull scenarios | Assumption | Bear | Base | Bull | | --- | --- | --- | --- | | Loan growth, years 1 to 5 | 0 / -2 / 0 / 1 / 2% | 3% each year | 5% each year | | Net charge-off rate | 7 / 7 / 6.5 / 6.25 / 6% | 5.75% each year | 5.25 / 5.25 / 5.25 / 5.5 / 5.5% | | Allowance to loans | 11.5% to 10.5% | 10% | 9.75% | | Terminal return on equity / growth | 12% / 1% | 17% / 3% | 20% / 4% | | Year-five common income ($ million) | 2,729 | 3,025 | 3,674 | | Ending shares (million) | 196.6 | 210.1 | 225.0 | | Minimum CET1 ratio | 12.80% | 12.85% | 12.85% | | Residual income value | $43.70 | $71.00 | $91.60 | | Versus $73.64 | -40.7% | -3.6% | +24.4% | **Bear:** credit contraction with lasting partner, funding and regulatory damage. Year-five income of $2,729 million includes a $222 million reserve release that does not repeat, and a further $711 million of annual after-tax earnings is lost permanently, so terminal income is $1,796 million, 34.2% lower. **Base:** normal losses, modest funded growth and margin discipline. **Bull:** stronger program economics, better efficiency and credit, with growth of 5% a year. The cost of equity is 12.5% in all three, so the scenarios differ only in operating performance. **Deep credit stress:** a recession with losses of 9.5% in year one, reserves of 14% of loans and a $400 million after-tax loss on securities gives $26.44. Common income is -$4,305 million in year one and the minimum CET1 ratio is 7.85%, just above the 7.0% requirement, so there is no modeled breach in a simplified test. Distributions stop after the declared dividend. The minimum common book is $9,258 million. It is a stress test and not a floor, and it leaves out deposit runs, deferred tax limits and supervisory restrictions. ### Sensitivity: residual income value per share | Terminal ROE \ cost of equity | 11.5% | 12.5% | 13.5% | | --- | --- | --- | --- | | 15.0% | $71.77 | $64.70 | $58.95 | | 17.0% | $79.14 | $71.00 | $64.41 | | 19.0% | $86.50 | $77.31 | $69.87 | ### Sensitivity: loan losses and retailer sharing | Net charge-offs \ share of extra losses absorbed | 0% | 25% | 50% | | --- | --- | --- | --- | | 5.50% | $72.17 | $72.74 | $73.32 | | 5.75% | $69.85 | $71.00 | $72.16 | | 6.00% | $67.53 | $69.26 | $71.00 | The discount rate is the largest single input: moving the cost of equity from 11.5% to 13.5% moves the value from $79.14 to $64.41. A 50 basis point shift in the yield on loans moves it from $66.36 to $75.64. The loan loss rate matters less than many readers expect, because the model counts losses once and retailer sharing offsets part of them, but the lasting tests above show what a permanent loss of returns does. ## How the buy price is set Confidence B starts at a 30% margin of safety. I add nothing for credit cycles, concentration or method dependence, because each is already in an operating case, the discount rate or the evidence score, and charging it again would count it twice. The buy range comes from six combinations on the Base operating path, each multiplied by 70%: | Input set | Method | Value | Buy limit | | --- | --- | --- | --- | | Cautious: 13.5% cost of equity, 318 million shares, terminal ROE 15.5%, growth 2% | Residual income | $59.28 | $41.50 | | Cautious (8x earnings) | Earnings multiple | $66.36 | $46.45 | | Central: 12.5%, 315 million shares, 17%, 3% | Residual income | $71.00 | $49.70 | | Central (9x earnings) | Earnings multiple | $74.83 | $52.38 | | Favorable: 11.5%, 312 million shares, 18.5%, 4% | Residual income | $88.95 | $62.26 | | Favorable (10x earnings) | Earnings multiple | $84.08 | $58.86 | The buy range is **$41.50 to $62.26**, and the conservative limit, rounded down, is **$41.49**. Reaching it takes a price decline of 43.7%. That is a policy outcome of stacked cautious assumptions, not a forecast that the stock will fall, and not an order: it is a conditional re-entry trigger. Below it, the credit, funding and partner thesis still has to hold up. If the drop came from a broken thesis, the limit is void. Above $62.26, none of the six combinations supports a new purchase. ## Dividend and buybacks The board declared a dividend of $0.34 a share, with a record date of November 5 and payment on November 16, 2026. Annualized, $1.36 is a 1.85% yield. Only the declared quarter is a commitment, and future dividends depend on the board and on regulators. At $428 million in year one, the Base dividend is covered 6.25 times by common earnings, with a payout of 16%. Do not read the buyback pace as an income yield. Over twelve months the company paid $420 million in dividends and $3,840 million for treasury stock, a total of $4,260 million against common earnings of $3,242 million. The excess released capital that had built up, and it will not repeat indefinitely. I add no buyback yield to the holder return in the model. ## Reasons to own SYF - **An embedded distribution advantage.** Long merchant relationships, underwriting data and servicing scale at the point of purchase are hard to copy, even though merchants keep bargaining power. - **A positive return spread.** After normalized credit costs, growth reserves, operating costs and preferred dividends, the Base still earns about 18% on common book against a 12.5% cost of equity. - **Capital and funding capacity.** A 12.9% CET1 ratio against a 7.0% requirement and $83 billion of deposits allow for growth and distributions when earnings exceed the capital retained. - **Per-share record.** Common book per share compounded at 16.9% a year over five years, and 38.3 million shares were retired in nine months. ## What could go wrong? - **Credit and program economics.** New cohorts could season into higher losses while reserves rise, and retailer sharing may offset less than the 25% I assume. Lower normalized profitability cuts both distributions and the multiple. - **Partner and regulatory economics.** A concentrated partner group can reprice at renewal, and fee or interest rules can change. The earlier late-fee rule was vacated, and an inquiry or proposal is not an enacted cap. - **Capital and funding.** Deposit repricing or credit stress can force earnings retention and halt buybacks at the moment shares get cheaper. - **Execution and allocation.** Fraud losses doubled in the quarter ($91 million against $43 million a year earlier), and buybacks above intrinsic value can destroy value while EPS still grows. Incentives tied to EPS and return on equity can reward a smaller denominator. ## Management execution Management credibility scores **10 of the 14 points I can assess, within a possible range of 10 to 16 out of 20**. Three dimensions (guidance versus outcomes, capital conversion commitments and acquisition returns) cannot be reconstructed from public data and are not counted as zero. Regulatory capital discipline, metric definitions and per-share outcomes earned the top marks. - Q3 common income fell 18% while EPS fell 6%: buybacks softened the decline without reversing the aggregate fall. - The 2025 incentive plan recorded a miss on receivables growth and a zero payout for that component, though the overall annual award paid 115.4% of target. - Half of the long-term award depends on cumulative EPS and average return on equity, which is why I check aggregate income and capital next to the per-share figures. - The June 2026 reclassification of internal-use software raised reported tangible equity, and common book did not change. I use common book to avoid rewarding a presentation change. ## Stock Analyza scorecard | Business quality | Score | | --- | --- | | Moat and pricing power | 15/20 | | Economic return over cost of capital | 15/20 | | Balance sheet and resilience | 11.25/15 | | Earnings quality | 7.5/15 | | Growth and runway | not assessable, 0 to 15 | | Management and capital allocation | 7.5/10 | | Dilution and governance | 3.75/5 | | Total (85% of the weight assessed) | 60 known, up to 75 of 100 | I do not rescale the known part: the missing runway score is missing evidence, not a poor result. The valuation score is **35 of 100**. The discount to the Base is negative, the Bear is just beyond 40% below the price, and the reverse valuation fits the central range. A small boundary crossing can change a heuristic score without changing the model. | Economic classification (banks) | Points | | --- | --- | | Sustainable return over cost of equity (about 17.7% against 12.5%) | 2/2 | | Book value plus dividends per share growth (19.35% a year) | 2/2 | | Deposit and funding quality | 1/2 | | Capital strength | 1/1 | | Credit quality (loss rate 2.92% to 6.31% in five years) | 0/1 | | Recurring customer economics | 2/2 | | Reinvestment runway | no data, 0 to 2 | | Known score | 8, range 8 to 10 of 12 | The supported class is Quality Growth. Compounder is not established, because the returns on new programs are not disclosed. The class does not select a multiple or change the value. ## Confidence score | Confidence | Score | | --- | --- | | Historical data and sources | 5/5 | | Accounting and ownership | 3.75/5 | | Forward assumptions | 3.75/5 | | Model construction | 5/5 | | Independent challenge | 3.75/5 | | Total | 21.25/25, grade B | The missing current award inventory caps an otherwise higher result at B. The grade measures how well the model is identified, not the chance of a loss, and wide credit scenarios do not lower it. ## What I would watch from here ### Green - Seasoned loan growth with stable or falling year-on-year delinquencies (30+ day today: 4.34%) - Allowance at or above 9.5% of loans without reliance on releases (today: 10.01%) - Aggregate normalized common income growing with a funded capital bridge - Fraud losses falling from $91 million a quarter without cutting essential spending - OnePay and co-brand growth disclosing usable returns ### Yellow - Net charge-offs drifting toward 6% or 30+ day delinquencies up 50 basis points year on year - EPS rising while aggregate common income keeps falling - Fraud staying near twice the prior year, or costs above 5.7% of loans - Deposit cost rising faster than loan yields - Buybacks at prices well above the model's fair values ### Red - Net charge-offs above 6.5% for two quarters without pricing or sharing compensation - CET1 below 11%, or falling more than 100 basis points alongside large buybacks - Rapid deposit runoff or a rollover problem - Loss or permanent repricing of a major partner - A binding fee or capital rule that invalidates normal returns Nothing is red today. The next official earnings date is not yet announced. The dividend record date is November 5 and the payment date November 16. The monitoring points are loss and delinquency development, growth reserves, fraud and servicing expense, after-sharing profit before provisions, and the actual capital retained. ## The three most important thesis breakers 1. Persistent credit deterioration that retailer sharing does not compensate. 2. Capital or funding erosion that stops the funded operating plan. 3. Loss or repricing of a major program that permanently cuts common returns. ## Stress test and independent check I recomputed every value in this analysis separately, with 50-digit decimal arithmetic and a different solution path for the buyback and tax loop: 2,958 checks, no discrepancies. That checks the arithmetic, not the forecasts. A second review challenged the claims, normalization, retailer sharing assumption, capital policy, peer definitions and the strongest opposing view. The main changes it forced: | Issue | Resolution | | --- | --- | | Unrealized securities loss and regulatory capital | Deep stress loss reduces book but not CET1 (opt-out), reconciled in later years | | Buyback tax | 1% tax included in the capital budget | | Bear and Deep terminal income | Reserve release reversal and permanent impairment quantified separately | | Capital target sensitivity | Relabelled as a joint change; no free capital release counted | | Secondary method independence | Shared drivers disclosed; no confidence reward | | EPS growth versus aggregate growth | Separated; no buyback yield added on top | | Preferred resets | B in 2029, C in 2031 flagged as an assumption | The inputs that matter most are the cost of equity, the terminal return and the yield on loans, with loan losses and retailer sharing next. The strongest bull argument is that new programs season well, fraud and servicing costs fall, and returns above 17% last, so the 17.84% return the price needs is feasible and my value may be too low. The strongest bear argument is that sharing absorbs less than assumed, young books need both higher losses and higher reserves, and merchants take more of the economics, while buybacks above intrinsic value destroy value even as EPS grows. My $43.70 and $26.44 are not floors, and a low price to earnings is not a margin of safety on its own. ## Final verdict: hold at $73.64, watchlist for new money Synchrony is a profitable, cyclical consumer bank priced close to what I think it is worth. A low headline price to earnings does not offer a margin of safety once normal credit losses, growth reserves and the capital that supports receivables are counted. The price is above all six entry cases. **Verdict: HOLD at $73.64, watchlist for new money. Base value $71.00, 3.6% below the price, cross-check $74.83, Confidence B (21.25/25). Required margin of safety 30%, buy range $41.50 to $62.26, conservative buy limit $41.49.** For an existing holder, the original cost of the shares is irrelevant, and nothing here says to sell: the price sits near my value, and a review is warranted if it approaches the Bull value of $91.60 with unchanged fundamentals, or if the thesis breaks at any price. New money waits for a price at or below $62.26, and a clean credit and funding picture. If $41.49 ever appears, check why before treating it as a bargain. The 11.66% a year that a buyer at $73.64 can earn in the Base depends on a smaller share count as much as on a bigger business, and the shares only shrink for as long as the credit book lets the company keep releasing capital. ## More analyses on Patreon This analysis is free. Premium members get every in-depth analysis, the full archive and a vote on which stocks get analyzed next: https://www.patreon.com/stockanalyza ## Sources Primary sources are Synchrony's [third quarter 2026 Form 10-Q](https://investors.synchrony.com/filings-regulatory/sec-filings/all-sec-filings/content/0001601712-26-000043/syf-20260930.htm), the [earnings release](https://investors.synchrony.com/filings-regulatory/sec-filings/all-sec-filings/content/0001601712-26-000040/earningsrelease3q26.htm) with its [financial tables](https://investors.synchrony.com/filings-regulatory/sec-filings/all-sec-filings/content/0001601712-26-000040/financialtables3q26.htm), the [quarterly presentation](https://www.sec.gov/Archives/edgar/data/1601712/000160171226000040/a3q26earningspresentatio.htm) and the [2025 annual report](https://www.sec.gov/Archives/edgar/data/1601712/000160171226000006/syf-20251231.htm). The dividend comes from the [company announcement](https://www.synchrony.com/contenthub/newsroom/synchrony-reports-third-quarter-2026-results.html) and the earnings calendar from the [investor relations site](https://investors.synchrony.com/news-events/calendar). Peer data comes from [Capital One](https://investor.capitalone.com/news-releases/news-release-details/capital-one-reports-third-quarter-2026-net-income-35-billion-or/), [Ally](https://mmx.prnewswire.com/media/MS2016505/Ally-3Q-2026-Press-Release.pdf?id=OA3027784&p=original) and [Bread Financial](https://investor.breadfinancial.com/static-files/239f6c71-99fe-4889-850d-a1594a62a331). The supervisory stress result is from the [Federal Reserve 2026 stress test tables](https://www.federalreserve.gov/publications/2026-june-dodd-frank-act-stress-test-appendix-a.htm) and the Treasury yield from the [H.15 release](https://www.federalreserve.gov/releases/h15/). The market price is from [WSJ](https://www.wsj.com/market-data/stock/syf). The cost of equity, the economic share count, the normalized earnings, the forecasts, the scenario figures and the scores are my own calculations or estimates, labeled as such. The reference market price is $73.64, the October 23, 2026 regular session close. Balance sheet figures are as of September 30, 2026. **Disclaimer.** This analysis is provided solely for research and educational purposes. It is not personalized investment, financial, legal or tax advice. Values are estimates, not forecasts or trading instructions. Synchrony is exposed to consumer credit losses, merchant partner concentration and renewals, funding and deposit costs, capital requirements, fee and interest rate regulation, fraud and valuation risk. Investors can lose part or all of their invested capital. Method: Master Investment Analysis Framework v4.2. Approach: five-year common-equity residual income with funded capital distributions, cross-checked by a 9x year-five earnings multiple. Confidence: B/21.25. Data checks: passed with warnings. Open data gaps: current award inventory, returns on newer programs, retailer sharing protection, post-September capital movements. Conditional research only.