The mechanism, stated precisely
The standard picture of a Roth IRA is index funds and slow compounding, capped at a few thousand dollars of contributions a year. What Peter Thiel did was different in one specific way that does all the work.
In 1999, before PayPal was worth anything, Thiel used his Roth IRA to buy 1.7 million founder shares at $0.001 each, a tenth of a penny, for a total of $1,700. When eBay bought PayPal in 2002, those shares were worth tens of millions, and by the end of that year his Roth held $28.5 million. He then reinvested inside the account, into an early Facebook stake and into Palantir, and the balance compounded to a reported $5 billion, all of which can eventually come out tax-free.
Here is the part that matters, because it is the whole game. The tax advantage did not come from the Roth's tax-free growth alone. It came from what went into the Roth: an asset priced at essentially zero that was about to become worth a fortune. The annual contribution limit was $2,000 in 1999, and 1.7 million shares at a tenth of a penny cost $1,700, so the entire future fortune fit inside a single year's contribution while it was still nominally worthless.
That is the trick, and it is worth being blunt about it. The Roth is tax-free by design; everyone gets that. The extraordinary result came from putting an asset with almost unlimited upside into the account while its official price was almost zero, so the explosive growth happened entirely inside the tax-free wrapper. It is not really an IRA strategy. It is a valuation strategy that uses an IRA.
Why almost nobody can do this
The reflexive lesson people draw is that they too could turn a small Roth contribution into a fortune. For nearly everyone, that is wrong, and the reason is precise rather than vague.
The strategy requires a specific and rare position: access to an asset that is legitimately priced at almost nothing but has enormous expected upside. That describes founder shares in a startup at formation, and virtually nothing else available to an ordinary investor. When you buy a public stock, its price already reflects the market's view of its future, so there is no gap to exploit. When you buy shares in an established private company, they are priced by real valuations. The tenth-of-a-penny share exists only at the very beginning of a company's life, and it is available essentially only to founders and the earliest employees and investors.
So the honest framing is that this is not a strategy most people can copy. It is a benefit that accrues to people who are already founding companies, which is a narrow and already-advantaged group. The $1,700 headline makes it sound democratic. The access required makes it the opposite.
The rule that is supposed to prevent it
There is a legal constraint here, and it is genuinely contested, which is part of why the story is more than a curiosity.
IRAs are not allowed to buy assets for less than fair market value, a rule designed precisely to stop people from stuffing artificially cheap assets into a tax shelter. The whole strategy lives or dies on whether a tenth of a penny was the genuine fair market value of those founder shares in 1999, or an artificially low number set by the founders themselves.
The defenders' case is real: at formation, a startup share often genuinely is worth almost nothing, because the company is an idea with no revenue and a high probability of failure. Par value of a tenth of a penny is a normal, defensible price for founder stock, and most startups do fail, which is what justifies the low valuation. On this view Thiel took a real risk on a worthless-looking asset that happened to succeed spectacularly, which is exactly what the Roth is supposed to reward.
The critics' case is equally real: the founders set that price themselves, on shares in their own company, with full knowledge of the prospects, and the valuation was low precisely because they controlled it. When the person buying the asset for their own Roth is also the person setting its price, the fair-market-value rule has very little to constrain. ProPublica noted the purchase risked running afoul of the rules designed to prevent IRAs from becoming illegal tax shelters. Both descriptions fit the same facts, and the tension between them is the unresolved heart of the issue.
It is not one person
This reads as a story about one unusually aggressive billionaire, but the data says it is a category. Using 2011 tax data, the GAO identified roughly 8,000 taxpayers with Roth balances above $5 million. By 2019 that number exceeded 28,000, and 497 individuals held an average of about $150 million in their IRAs.
You cannot reach $150 million in an account with a few-thousand-dollar annual contribution limit through public-market investing and compounding. The math does not work in a lifetime. Balances at that level essentially require the same move Thiel made: putting privately-priced, pre-explosion assets into the account. The concentration of these mega-Roths among a small number of very wealthy people tells you the strategy is systematic, not a one-off, and that is what turned it into a policy question rather than a piece of trivia.
The fairness question, honestly on both sides
Whether this should be allowed is a genuine values disagreement, and it deserves both sides rather than a verdict.
The case that it is fine: nothing about it is illegal, Thiel followed the rules as written, he took real risk on assets that could easily have gone to zero, and the Roth's entire purpose is to reward people who pay tax up front and let investments grow untaxed. He paid the tax he owed on the contribution. Punishing a successful legal strategy after the fact, the argument goes, is changing the rules on someone who played by them, and the risk he took was real even if the outcome looks inevitable in hindsight.
The case that it is a problem: the Roth IRA was created to give middle-class families retirement security, and Senator Ron Wyden's objection is that it was designed to provide retirement security to middle-class families, not to allow mega-millionaires and billionaires to avoid paying taxes. A vehicle capped at a few thousand dollars a year, aimed at ordinary savers, is being used to shelter billions by a group with access no ordinary saver has, and the cost is untold tax revenue the public treasury never collects. Following the letter of a rule while defeating its clear purpose is the definition of a loophole, on this view, even when no law is broken.
Both of these are serious arguments. The disagreement is not really about the facts, which are not in dispute. It is about whether legality settles the question or whether purpose does, and reasonable people land in different places on that.
What Congress has tried to do
The policy fight is live, not hypothetical. Lawmakers have proposed forcing distributions from oversized retirement accounts: under one House proposal, holders would have to withdraw 50% of the amount in accounts over $10 million, and 100% of amounts over $20 million, with income thresholds exempting ordinary savers, single filers under $400,000 and couples under $450,000. Thiel could be forced to draw down all but roughly $20 million and owe income tax on the growth.
Notably, even the proposed fix has a gap. One estate planner observed that a taxpayer clever enough to keep adjusted gross income below the threshold could avoid the new rule altogether, which is a reminder that people with sophisticated advisors tend to have more room to maneuver around targeted rules than the rules' authors expect. That is a recurring pattern in tax policy: a rule aimed at a specific abuse often catches the less sophisticated while the most sophisticated route around it.
The useful takeaway
For an ordinary investor, the practical lesson is smaller and more honest than the headline. The Roth IRA is a genuinely powerful, genuinely valuable tool, precisely because tax-free growth compounds. Contributing early and consistently, and holding assets with real growth potential inside the Roth rather than in a taxable account, is sound and available to everyone. That is the replicable part, and it matters.
The part that is not replicable is the one that produced the billions, and pretending otherwise does readers a disservice. Turning $1,700 into $5 billion tax-free required buying founder shares at a tenth of a penny, which required founding a company that became PayPal. The account was ordinary. The asset was not, and the access to that asset is what almost nobody has.
The larger story is one the tax code keeps producing: a benefit written for one purpose, used at scale for another by people positioned to exploit the gap, followed by a slow and imperfect legislative attempt to close it. Whether you see Thiel's Roth as brilliant tax planning or a middle-class program captured by billionaires depends on a question the facts alone cannot answer, which is whether following a rule to a place its authors never intended is cleverness or abuse. That question is still open, and the money is still tax-free while it stays that way.