NerdWallet: $119.6 million of cash flow shrinks to $62.0 million once taxes return

NerdWallet generated $119.6 million of owner free cash flow over the last twelve months, after stock pay and all capital spending. That is 16.9% of its $708 million market value, and it looks cheap until you ask where the cash came from. Cash taxes were far below the tax charge, earnings are falling, and management says normal tax payments return in 2027. On a full-tax, guidance-based footing the figure is $62.0 million, 8.8% of the market value.

At the $10.91 close my triangulated Base value of $11.63 is 6.6% above the price, well inside the 15% band I treat as fair value. The Bear value is $5.95 and the Bull value $19.52. Weighting the three cases 30%, 50% and 20% gives $11.50. Confidence is grade C (14/25), the required margin of safety (MOS) is 45%, and the maximum buy price is $6.40, so the price is 70.5% above it. The two valuation methods give $12.55 and $10.24, and neither one brings the maximum buy price within reach of the share price.

Valuation route: an internet marketplace that is paid per lead or per funded transaction. Primary method: a five-year discounted cash flow on normalized owner cash. Secondary method: enterprise value to cash earnings (adjusted EBITDA minus stock pay and capitalized spending). Net cash of $56.9 million is added once, and the buyback is not credited separately.

What NerdWallet is

NerdWallet runs a free personal finance site and app where people compare credit cards, loans, insurance and savings products. It is paid when a reader applies for, or is approved for, a partner's product. About 650 employees support it, the gross margin is 93.5%, and the founders hold 53.8% of the votes through a dual-class structure.

The weak point is how readers arrive. Management has pointed to organic search as the reason credit card revenue fell $8.6 million year on year and small-business revenue fell 11%, and it describes its largest auto insurance partner relationship as stabilized, not recovered. I could not find Google's share of traffic, so it stays an unknown that I price through the Bear case and the margin of safety.

The latest quarter

Revenue grew while profit shrank, because management has switched to buying customers against a target return on the money spent. The full-year guide includes a $15 million to $20 million hit to non-GAAP operating income from that approach, and the payback comes after 2026. I do not treat it as a mistake, but it means near-term earnings are lower than the twelve months I would otherwise capitalize.

Which cash to value

I value owner cash: operating cash flow minus all capital spending, including capitalized software and intangibles, minus stock pay. The last twelve months look like this.

I do not use it as the base, for four reasons. First, NerdWallet is not a federal cash taxpayer in 2026: it paid $1.8 million of cash tax in the first half against a $10.6 million charge and received $9.5 million of refunds, and management expects normal payments in the second or third quarter of 2027. Second, earnings are falling, and the guide of $65 million to $80 million of GAAP operating income compares with $90.8 million over the last twelve months. Third, the marketing step-up. Fourth, reinvestment: amortization runs about $40 million a year against $17 million of capitalized spending, and the gap is filled by small acquisitions ($24.2 million of cash acquisitions in the last twelve months), so I charge $12 million a year for them.

Cash taxes were far below the tax charge, so I used a full tax rate. As a cross-check, operating income plus interest of $75 million after 26% tax is $55.5 million, which charges all amortization and so sits below my figure. Because the normalization takes effect within twelve months, I value today's run rate and the 2027 tax step is already in year one.

Shares and buybacks

There were 63.88 million shares at June 30, plus 1.0 million of dilution from stock awards, so I use 64.88 million. The 4.45 million options carry an $11.67 strike, above the price, and I count only the in-the-money effect. NerdWallet spent $162.2 million on buybacks in the last twelve months, which cut the diluted count by 14% in a year, and $67 million of authorization is left. The average price paid was $8.36 in the second quarter and $10.25 for the half, so at $10.91 new repurchases are being made above the second-quarter level. I do not credit the buyback separately, because the cash it uses is already in the cash flow I value.

The primary valuation: discounted owner cash

I discount five years of owner cash, then a final value. Year one is the normalized $62.0 million, and all inputs are my assumptions.

The Bear is a business where search keeps eroding cards and small-business revenue and the marketing spend does not pay back, with adjusted EBITDA of about $105 million. The Bull is adjusted EBITDA of about $165 million, which needs personal loans, deposits, student loans and the College Finance acquisition to compound for several years. Base is $12.55 a share.

Moving only the starting cash, $50 million gives $10.29, $62 million gives $12.55 and $75 million gives $15.00.

Where the value sits

The value after year five is 57% of the Bear, 67% of the Base and 74% of the Bull primary value, all below the 85% level at which I would stop and check a five-year model. Most of the Base value therefore comes from the first five years of cash, and the final value assumes 2% growth, which is no more than inflation.

The secondary valuation: a multiple of cash earnings

The second method takes adjusted EBITDA minus stock pay and capitalized spending, which is $93.5 million in the Base case, and applies a multiple. I use 5.0 times in the Bear, 6.5 times in the Base and 8.0 times in the Bull, then add net cash of $56.9 million and divide by 64.88 million shares. Base is $10.24, Bear $5.46 and Bull $15.61.

A multiple needs a reason and not just a range. These multiples are before tax. After 26% tax, 6.5 times is 8.8 times earnings, which at an 11% required return means a business whose earnings stay flat, about -0.3% a year forever. That is the view I hold for the Base: demand for the product is real, but the search channel and the falling latest quarter argue against paying for growth in this method. The Bear multiple of 5.0 times is 6.8 times after tax, or earnings shrinking about 2.4% a year at 12%. The Bull multiple of 8.0 times is 10.8 times after tax, only +0.7% a year at 10%, and is deliberately modest because the Bull case already carries a much larger earnings base of $165 million. The multiples are my assumptions, and I have no peer table to check them against.

Same business, two lenses

The two methods start from different measures, so I put them on the same footing in the Base case.

The gap of 20% between $12.55 and $10.24 comes from one assumption: the discounted model grows owner cash 5% a year for four years after year one, and the multiple assumes none. The market price sits between the two, which fits a market that expects some growth but not as much as I model. I do not pick a side. The weighted Base of $11.63 (60% model, 40% multiple) is a point inside a range of $10.24 to $12.55, and the range is the honest message. Because both methods start from the same guided EBITDA, their agreement does not raise my Confidence.

The choice of method does not change the decision. The maximum buy price is $6.90 on the owner cash model alone and $5.63 on the multiple alone, against a share price of $10.91.

Bear, Base and Bull

The probability-weighted value is $11.50. Divergence between the two methods is 14% in the Bear, 20% in the Base, right at my warning level, and 35% in the Bull, which I flag. The Bull needs adjusted EBITDA of $165 million against a guide of $131 million to $147 million, which is possible only over several years, and that is why it carries 20%.

What the price already assumes

At $10.91 the market value is $708 million and the enterprise value is $651 million: 4.7 times adjusted EBITDA and 10.5 times my normalized owner cash. Put differently, owner cash of $53.3 million, 14% below my $62.0 million, justifies today's price at 5% growth and an 11% discount rate. The market has already priced in the tax normalization and some of the earnings decline. It has not priced in a failure of the search channel, which is what the Bear case describes.

The stock is up 33% in the last month, but it is 19% lower this year and well below its 52-week high of $16.24, so the recent bounce has not erased the decline.

How the buy price is set

Confidence C calls for a margin of safety of 35% to 50%. I use 45%: 40% as the base for this grade, plus 5 points for concentration in the search channel and a few card and insurance partners. The search risk also sits in the Bear scenario and in the 11% discount rate, and the tax normalization sits in the cash flow, so I do not charge them again here. The maximum buy price is $11.63 times (1 minus 45%) = $6.40. At a 35% margin it would be $7.56. It is a decision reference, not an order to place.

Reasons to own

• Net cash and a 93.5% gross margin. The balance sheet carries no real debt, and each extra dollar of revenue is highly profitable if the marketing pays back.
• Return on capital near 20%. A marketplace that needs little equipment.
• Buybacks at modest prices. The diluted share count fell 14% in a year.
• Growth in the pipeline. The third-quarter revenue guide is about +17%, helped by personal loans, deposits and the College Finance acquisition.

These reasons justify research. They do not close the gap between $10.91 and the $6.40 I would want to pay.

What could go wrong?

• Organic search. The channel management blames for the card and small-business declines. Google's share of traffic is not disclosed.
• Marketing that does not pay back. The $15 million to $20 million hit from return-based spending is a bet, and the payback comes after 2026.
• Tax normalization. My owner cash assumes full tax from 2027. A larger cash tax bill would lower it further.
• Partner concentration. The largest auto insurance relationship is stabilized, not recovered, and a few card issuers matter a lot.
• Control and selling. Founders hold 53.8% of the votes, and a large holder, Topline, sold about 2.0 million shares in June at $8.95 to $9.27. Short interest is 17.7% of the float.

Management execution

Management is candid about the weak spots, which I count in its favor, and it has bought back stock well below today's price. The change to return-based marketing lowers near-term profit on purpose, and the same kind of acquisition spending keeps showing up in cash flow. I give management credibility 6 of 10 and capital allocation 6 of 10.

Stock Analyza scorecard

Business quality is a judgment, not an audited figure. The valuation score is 38 of 100: 8 of 40 for the discount to Base, 5 of 20 for protection against the Bear (45% below the price), 7 of 15 for agreement between methods, 6 of 10 for price history, 8 of 10 for what the price implies and 4 of 5 for data reliability. Business quality and valuation are separate numbers and are never blended.

Confidence score

Predictability scores lowest because of the search channel. Two inputs I could not find are the net operating loss balance, which could keep cash taxes low for longer, and revenue by product line.

What I would watch from here

The thresholds are my monitoring triggers, not company guidance. The next report is the third-quarter release on November 5, 2026.

The three most important thesis breakers

• Search traffic falls again. Both methods assume the base of revenue survives.
• Marketing spend does not pay back. Revenue grows but profit does not, and owner cash falls toward my Bear $42 million.
• Taxes normalize sooner or higher. This removes the part of the cash flow that is already priced in my Base.

Stress test and independent check

I rebuilt every number in a separate script: owner cash, normalized cash, the three scenarios, the sensitivity grids, the multiple conversions, the terminal value shares, the triangulation, the buy prices and the reverse valuation. They match the article. Facts come from NerdWallet's second-quarter 2026 release and its 10-Q for the period ended June 30, 2026. Quarterly figures and prices come from stockanalysis.com and Finviz. The $10.91 close of October 22 comes from Finviz, which showed a stale $11.05 intraday figure elsewhere, a gap of 1.3%.

The estimated inputs are the 26% tax rate, interest income of $2.5 million, sustaining acquisitions of $12 million, run-rate stock pay of $28 million, the Bear and Bull earnings, and all multiples and discount rates. Not available: the loss carryforward balance, revenue by vertical, Google's traffic share and partner concentration. Corrections to the first look all went down, from $119.6 million to $62.0 million, so I tested the opposite direction as well.

Even the generous combination leaves the price above the entry level, and only the Bull case reaches it, which is why I cap the Bull at 20%. The strongest argument against the Hold is that search stabilizes and the marketing pays back, which gives the Bull. The strongest argument for caution is that the most recent quarter shows profit falling 30% to 40% while the stock has bounced 33% in a month.

Final verdict: hold at $10.91

NerdWallet is a profitable, net-cash marketplace whose headline cash flow overstates what it will earn once taxes return and marketing spending pays for itself. The shares are fairly priced against my Base value, but at 70.5% above the maximum buy price they do not offer the protection I require.

Verdict: HOLD at $10.91. Triangulated Base value $11.63 (range of the two methods $10.24 to $12.55), Confidence C (14/25), required margin of safety 45%, maximum buy price $6.40. Main thesis breaker: a further decline in organic search traffic.

For holders this is a review, not an instruction to sell. For new money it means waiting for a price near $6.40 to $7.56, or for evidence that the search channel is stable.

Sources

Primary sources are NerdWallet's second-quarter 2026 results release and its Form 10-Q for the period ended June 30, 2026, both on SEC EDGAR. Quarterly figures come from stockanalysis.com. The $10.91 price is the October 22, 2026 regular session close. Per share values use 64.88 million diluted shares. The normalized tax rate, acquisition charge, multiples and discount rates are analytical assumptions, not company guidance.