The Qualcomm Match Is Not a Percentage of Your Pay. It Is Four Tiers, and the Last One Pays Ten Cents.

Almost every 401(k) match you will read about is written the same way: the company puts in fifty cents for every dollar you defer, up to six percent of your pay. Your salary sets the ceiling, and the rate is flat all the way to it.

This plan is not built that way. The match is a tiered structure denominated in cumulative dollars of employee contribution: 100% of the first $1,500, 50% of the next $1,500, 33% of the next $7,500, and 10% of everything after that. There is no "X% of the first Y% of compensation" anywhere in it. Your salary does not set the tiers. The tiers are the same dollar figures for a new engineer and a vice president.

That formula is not published in a single tidy document. It was established by reconciling the tier arithmetic against the maximum match dollars Qualcomm discloses in its own proxy statements, and it reconciles for all four years on record, 2022, 2023, 2024 and 2025. The 2025 figure of $7,150 reconciles only if you use the age 60–63 catch-up, which matches the same proxy's changed wording about catch-up contributions "for employees at certain ages". Four independent years agreeing is why this is stated plainly rather than hedged.

What this assumes ↓

When RSUs vest, the value on the vesting date is taxed as ordinary income, not as a capital gain. The most common reason people hold after vesting is to reach long-term capital gains treatment, which is real, but it applies only to the growth from the vest date onward. The vest itself was already a taxable event at ordinary rates.

Which reframes the decision. If the company handed you a $50,000 cash bonus, would your first move be to buy $50,000 of company stock with it? For most people the answer is no. Holding a vested RSU position is making exactly that trade, every quarter, by default rather than by decision.

The concentration argument is sharper here than at most employers because your salary, your match, your unvested equity and your next promotion are already claims on the same company. A vested position on top of that is not diversification, it is the fourth helping. That is a risk decision, and it deserves to be argued as one rather than settled by a tax rule of thumb.

We have written about the vesting decision itself in more detail, and about what a long-term gain actually costs a San Diego household once the federal surtax and California are counted, the arithmetic there is usually the part that surprises people.

The first is health coverage. Leaving before 65 means covering the gap yourself, and COBRA costs 102% of the total premium, including the share the company had been paying, which most people have never seen as a number. For a family plan in San Diego that is a substantial monthly line item for however many years sit between your last day and Medicare.

The second is what happens to unvested equity. Whether the package accelerates vesting, vests pro rata, or forfeits the unvested balance outright is frequently the difference between a good offer and a poor one, and it is usually decided in the plan documents rather than negotiated.

The third is the tax year the money lands in. A large lump sum arriving in a year where you have already earned a full salary and bonus stacks on top of that income rather than starting fresh. Whether any of it can be deferred, offset, or shifted is worth asking before signing, not after.

Accessing a 401(k) before 59½ without the 10% penalty is possible through substantially equal periodic payments under IRC §72(t), but it is a rigid, multi-year commitment and breaking the schedule triggers retroactive penalties. It is a real option and a poor one to improvise.

Qualcomm has never filed a Form 11-K. Form 11-K is required only where plan interests or employer stock are registered on Form S-8, and every Qualcomm S-8 back to 2008 registers equity-incentive or acquisition plans rather than the 401(k). The plan holds no company stock fund requiring registration, so no filing obligation attaches. The practical consequence is that there is no public plan description to quote.

So: whether the match includes a true-up is not disclosed anywhere. A search of the 2026 proxy for "true-up" returns nothing, and there is no 11-K to check. We are not inferring it either way.

Nor is it public whether the tiers are applied per pay period or on cumulative year-to-date contributions. That distinction is the whole of the front-loading question: because these tiers are denominated in dollars rather than as a percentage of each paycheck, an employee who reaches the limit by June has still crossed every tier and, on a cumulative calculation, would receive the identical full-year match. A front-loading penalty would exist only if the bands are prorated per period, which is unusual for a dollar-tiered design and is unverified. If you have been told you lost match by front-loading, that is worth checking against your own statements rather than against an article.

Vesting is reported as graded, 50% after one year of service, 100% after two. That comes from two independent secondary sources that agree, with no primary document behind them, so treat it as probably right rather than settled.

One thing the record does say clearly: the plan offers a self-directed brokerage account. That is stated in Qualcomm's own 2025 proxy, in the board's response to a shareholder proposal, and it is the best-sourced plan detail available.

All of this resolves in about thirty seconds from inside a Fidelity NetBenefits account, or from the summary plan description. If you have either, bring them and we can stop guessing.

The arithmetic behind $47,500 is the 2026 annual additions limit of $72,000 less the $24,500 elective deferral limit. That is correct as far as it goes, but it assumes the employer contributes nothing. Employer contributions count toward the same $72,000 ceiling, so for anyone actually receiving the match at a full deferral the figure is $72,000 − $24,500 − $6,125, which is $41,375. Roughly six thousand dollars less than the number usually quoted.

The second correction runs the other way, and it is the one people more often get wrong to their own cost: catch-up contributions do not count as annual additions and do not consume that headroom. They stack on top. Someone 50 or older does not lose after-tax room by making catch-ups, the 2026 ceiling becomes $80,000, or $83,250 for ages 60 to 63.

Both figures assume something we cannot confirm: that this plan permits after-tax non-Roth contributions at all. We could not verify that it does. The 2026 proxy describes the match as applying to "Pretax and Roth" contributions, and the available employer benefit descriptions list Roth deferrals rather than an after-tax bucket. Neither confirms nor denies it. Until it is confirmed from the plan documents or a NetBenefits contribution-elections screen, treat the mega backdoor Roth as an open question here rather than a strategy on the shelf.

One further constraint that arithmetic alone will not show: after-tax contributions are subject to ACP nondiscrimination testing. In a plan with heavy highly-compensated participation and no safe harbour on the after-tax source, contributions can be refunded after year end. That is the most common real-world failure of an otherwise sound mega backdoor plan.