The Swap : How your PEA buys America without ever owning it

The PEA, France’s tax-advantaged equity savings plan, was built for a precise mission: to steer French savings toward European companies. That is its entire reason for being, written into the law and hammered home in every brochure. And yet millions of account holders now use it to bet on the S&P 500, the Nasdaq 100, or the MSCI World. How does a vehicle designed to finance Sanofi, LVMH, and TotalEnergies end up exposing its owner to Apple, Nvidia, and Microsoft?

The answer lies in a contract whose very existence escapes almost every investor. Not a loophole, not a sleight of hand: a perfectly legal, tightly regulated mechanism, spelled out in plain language in the regulatory documentation of the funds involved. It is called the swap. And for anyone willing to look at what sits underneath it, it tells a far more interesting story than the simple question of how to fit the S&P 500 into a PEA.

What You Think You Own, and What the Fund Actually Holds

Start with the founding misunderstanding. When you buy a PEA-eligible S&P 500 ETF, you reasonably assume that, through your shares, you own a sliver of the five hundred largest American companies. You don’t. The fund does not hold a single American stock.

What it holds is a basket of liquid European equities: LVMH, Sanofi, SAP, ASML, names chosen to satisfy the regulatory quota that requires at least 75 percent of assets to be domiciled in the European Union or the European Economic Area. This basket is the fund’s visible skeleton, its compliance facade. But its purpose is not to give you European exposure. Its purpose is something else entirely.

On top of that basket, the fund manager enters into a performance swap with an investment bank, usually its own parent company. This contract, the total return swap, works on a disarmingly simple principle: the fund pays the bank the performance of its European basket, and in return the bank pays the fund the exact performance of the S&P 500. The swap is reset very frequently, sometimes daily. When the American index rises faster than the European basket, the bank pays the difference to the fund; when the reverse happens, the fund pays the bank.

For you, the result is identical to a direct investment in the index. But the nature of what you hold has changed completely. You do not own America. You hold a claim on a bank, backed by a contract, secured by a collateral basket of European shares you never chose and whose performance is of no concern to you whatsoever. You are aiming at the S&P 500; the fund holds a basket of CAC stocks and a signed piece of paper. It is the same reversal at work in structured products, where the saver believes he is investing when he is in fact selling his bank an insurance policy: the label promises one thing, the structure delivers another.

This distinction is not a plumbing detail. It is the whole point.

The False Obvious: “European Stocks Don’t Perform as Well”

Here is the objection that springs naturally to mind, and that turns out to be a dead end. If European markets structurally underperform American ones, why on earth would an ETF provider have any interest in offering this arrangement? Wouldn’t it lose money promising S&P 500 returns while holding only sleepier European stocks?

This intuition rests on a confusion. It assumes that the European basket’s performance benefits, or harms, someone. In fact it benefits no one and harms no one, because the swap neutralizes it the moment it is recorded. The European basket is not an investment. It is a regulatory extra, doubling as collateral. Its only job is to tick the “75 percent eligible assets” box and to serve as a guarantee should the counterparty default. Whether that basket rises or falls has no bearing on what the end investor receives, which is, by construction, American performance.

The provider, for its part, does not earn its keep on the relative performance of two baskets. It earns it elsewhere, and far more handsomely than you might imagine.

How They Make Their Living

The swap’s value chain has several links, and each takes its tithe.

The first, and most visible, is the annual management fee. It looks trivial, often between 0.05 and 0.15 percent, but it applies to colossal pools of assets. When a single PEA ETF holds several billion euros, a tenth of a percent becomes a comfortable, perfectly recurring income stream.

The second link is the bank counterparty itself. The bank that signs the swap collects a margin on the contract. It can also lend out the substitute basket’s securities to other market participants, with securities lending generating additional income on assets that were, in any case, only sitting there as collateral. Since this counterparty is usually the manager’s own parent company, most of the value stays within the same financial group. The loop is elegantly closed.

But the third lever is the most interesting, and it is the one that decisively overturns the opening objection. Far from being a degraded product tolerated for lack of anything better, the synthetic ETF enjoys a structural tax advantage that frequently lets it beat its physical equivalent.

The explanation lies in a provision of the American tax code, the well-known IRS rule 871(m). A physical ETF that actually holds American shares is subject, on the dividends it collects, to a withholding tax of 15 percent where a tax treaty applies, and up to 30 percent where one does not. That levy mechanically eats into its net return. The synthetic ETF, by contrast, holds no American shares at all: it receives the index’s performance through the swap, gross dividends included, and rule 871(m) specifically exempts swaps that track certain qualified major American indices from this withholding. The dividend is paid in full, untouched.

One caveat is in order: this tax grace is not universal. It applies only to indices the IRS recognizes as qualified, the broad, liquid headline indices such as the S&P 500, the Nasdaq 100, or the MSCI USA. On sector or more obscure indices, the machinery seizes up: the bank counterparty then bears the withholding and passes it through into the swap’s price, so the advantage evaporates. The trick is therefore nothing magical; it depends on the exact scope of the index being tracked, and it is precisely on the most popular indices that it delivers its full effect.

The upshot: on the S&P 500, a swap-based ETF holds a structural return advantage over a physical ETF tracking the same index. The arrangement many take for a regulatory makeshift is, after tax, often superior to “real” ownership. The French investor who parks the S&P 500 inside a PEA is therefore not making a degrading compromise. He is exploiting, most often without knowing it, an American tax quirk that European financial engineers have methodically industrialized.

No one in this chain is doing charity. And it is precisely because everyone profits that the product exists in such abundance.

And What About European Companies in All This?

One question deserves to be asked, if only to answer it honestly: all this money flowing through baskets of LVMH, Sanofi, SAP, and ASML, does it artificially inflate the valuation of these European giants?

The intuition is tempting. If tens of billions of euros of French savings feed ETFs that, to qualify for the PEA, must buy European stocks, then a mechanical demand is created for those shares, regardless of their merits. On paper, the effect is real.

But one must resist the temptation of the spectacular conclusion. Several factors sharply limit its reach. First, scale: the sums involved, however impressive in absolute terms, remain modest against the market capitalization of these behemoths, whose free float runs into the hundreds of billions. Second, the nature of the basket: it is collateral that is frequently reset, not a long-term holding that piles up year after year. Finally, and above all, the counterparty carrying these shares generally hedges its position through offsetting trades, which largely cancels out the net buying pressure.

So the effect on European companies’ valuations does exist, but it is marginal and indirect, very far from being a driver of prices. To claim otherwise would flatter the narrative while playing fast and loose with the numbers. The European basket remains what it is: a regulatory alibi, not a support fund for the continent’s industry. For a vehicle meant to finance the European economy, the irony is worth underlining.

Paper Gold and the Bullion Bar: a Useful Analogy, So Long as You Fence It In

To grasp the deeper nature of what one holds, a comparison imposes itself: gold. There is physical gold, the bar in the vault, which you own with no one able to default on you. And there is paper gold, gold ETFs, certificates, contracts pegged to the price of the metal without your ever touching a single gram. In one case, a good; in the other, a promise.

The analogy perfectly illuminates the central axis of our subject. The synthetic ETF in a PEA is the paper gold of American exposure: not the metal, but the claim on its performance, hanging on the solvency of whoever issues it. It also illuminates the systemic fragility, for paper gold carries a famous criticism, that of stacked promises: the volume of paper claims vastly exceeds the gold that could actually be delivered, and the system would not hold if everyone demanded their due at once. This is exactly the kind of cascading fragility the swap conceals beneath its apparent solidity.

Still, the analogy must be handled with care, or it will turn against whoever wields it. On two points precisely.

First point: a share held in an ordinary securities account is not a bar in a vault either. It is dematerialized, recorded along a custody chain (your broker, then Euroclear). You never hold a company in the palm of your hand. So the right mapping is not physical versus paper, but rather this: the synthetic PEA is akin to paper gold, while the ordinary securities account, the CTO, is akin to allocated gold or a registered security, real ownership, to be sure, but still inscribed in a ledger, never an object you can clutch in your pocket. To suggest that the CTO is the bullion bar would flatter it unduly.

Second point, more dangerous still: paper gold is almost always an inferior product to physical gold for an equivalent exposure, because it piles on fees and counterparty risk while offering nothing in return. Yet we have just established the opposite for the synthetic ETF: after tax, thanks to the American tax quirk, it often beats its physical equivalent. The two arrangements share the mechanics, a claim rather than a good, but not the motivation. One buys paper gold out of convenience, accepting a degradation; one parks the S&P 500 in a synthetic PEA because it is, net of tax, the best vehicle available. The analogy holds for the nature of ownership, not for the quality of the product.

The Fragility No One Looks At

Here is the heart of the matter, the part that fee comparisons and TER rankings never mention. The arrangement is elegant, efficient, often superior. But it is conditional. It rests on three pillars, not one of which depends on you.

The first pillar is an American tax provision. The synthetic ETF’s entire return advantage on the S&P 500 flows from the withholding exemption granted by rule 871(m). Yet that rule was created precisely to close optimization loopholes, and it has already been tightened in the past. Nothing guarantees that Washington will indefinitely tolerate European savers capturing the gross dividends of American companies through swaps. The day the American legislator closes that door, the structural advantage will evaporate, and the product will lose part of its reason for being. Your wealth strategy rests, unbeknownst to you, on the tax goodwill of a foreign country.

The second pillar is the solvency of the bank counterparty, a soundness never quite as settled as it looks. The UCITS framework caps counterparty risk exposure, in principle, at 10 percent of the fund’s net assets, and the frequent resetting of the swap along with the posting of collateral reduce that risk further. To be precise, and to avoid scaremongering: the substitute basket is often over-collateralized, with the counterparty depositing 105 or 110 percent of the fund’s value in European shares with an independent custodian. If the bank fails, the fund does not vanish: it recovers that basket of European shares. The risk, then, is not losing everything, but an execution risk. In the time it takes to liquidate the European basket and rebuild the exposure, the S&P 500 may have moved, and the gap widens at precisely the worst moment, in the thick of the storm, when markets are sliding and liquidity is drying up. Contained, then, but not nil. In normal times, this risk is theoretical. In a systemic crisis, when bank counterparties become the very thing on everyone’s mind, it is the days of maximum stress that count, not the calm years.

The third pillar is the regulatory framework itself, UCITS and the PEA’s eligibility rules, liable to evolve with European and national policy trade-offs.

Line up these three pillars, and the true nature of the arrangement appears. Your American exposure passes through the balance sheet of a French bank, through the tolerance of the American tax authorities, and through the stability of a European regulatory framework. Three dependencies in a cascade, not one of which exists when you hold the security directly. For anyone interested in questions of sovereignty and the fragility of financial edifices, this is a textbook case: a mass-market product, sold as simple and robust, that is in reality nothing more than a delicate balance suspended on decisions made elsewhere.

So No, It Is Not the Same as a CTO

Let us be fair to the end. The replication the swap provides is excellent. The tracking error is often minuscule, the performance faithful, and the tax optimization on dividends even confers an advantage that physical ownership does not. On the strict measure of net return, the synthetic PEA ETF has nothing to be ashamed of, quite the opposite.

But fidelity of replication and the nature of ownership are two distinct things, and it is the second that radically separates the synthetic PEA from an ordinary securities account.

In a CTO, when you buy a physical S&P 500 ETF, the fund genuinely holds the American shares. Your exposure depends neither on a counterparty that might default, nor on a foreign tax loophole that might close, nor on a regulatory quota that forces a substitute basket. You own, in the full sense and through your shares, a piece of the American economy. In a synthetic PEA, you hold a claim backed by a derivative, secured by collateral that does not interest you, and whose advantage rests on the tolerance of a tax authority that is not your own.

The real reason for accepting this compromise is not, in fact, the product’s intrinsic quality. It is the PEA’s tax wrapper, its reduced social charges and its exemption from income tax after five years, in a country where the taxation of wealth weighs heavily enough to justify no end of contortions. One does not choose the swap because it would be better in absolute terms; one chooses it because it is the only key that opens the PEA’s tax door to America. It is a perfectly rational trade-off, provided you know what you are trading.

To own real shares is ownership. To house the S&P 500 in a PEA is exposure on lease: brilliantly engineered, tax-efficient, but revocable, and hanging on decisions you will never make. The distinction means nothing as long as all goes well. It will mean everything the day something goes wrong.