The Royal Dutch/Shell Parity Mispricing
Two share classes of the same underlying company, permanently entitled to a fixed 60/40 split of every dollar the group earned, traded at prices that violated that ratio for decades — one of the cleanest textbook arbitrages ever documented, and one of the hardest to actually profit from.
Prerequisites: Limits to Arbitrage
In 1907, Royal Dutch Petroleum and Shell Transport and Trading merged their operations but kept two separate, publicly traded parent companies. The merger agreement fixed, forever, how the combined profits would be split between them: Royal Dutch shareholders got 60%, Shell shareholders got 40%. Every dividend, every asset sale, every dollar of earnings followed that exact ratio, no exceptions, for the rest of the entity's life.
That means the two stocks should have traded at a price ratio of exactly 60:40 — Royal Dutch worth 1.5 times Shell, always. It didn't. For most of the period academics studied it, the ratio wandered as far as 35% away from parity, and it stayed away from fair value for years at a stretch rather than snapping back in days.
When a company's economics are legally fixed in a ratio, but its two share classes trade in separate markets to separate investor bases, price and fair value can drift apart because the investors who could arbitrage the gap face limits that have nothing to do with the merger contract.
Why an obvious arbitrage didn't close
The mechanical trade is simple: if Royal Dutch trades rich relative to the 60:40 ratio, sell it and buy Shell, and collect the gap when the ratio corrects. LTCM ran exactly this trade in the late 1990s. The problem was that the two stocks were listed on different exchanges (Amsterdam and London primarily, with US listings too), held by different regional investor bases with different index memberships, and subject to different withholding-tax treatments on dividends. A Dutch pension fund mandated to hold Dutch-listed shares can't simply swap into the "cheaper" UK-listed twin. That segmentation meant the two prices could be driven by separate regional flows — index rebalancing, currency moves, local investor sentiment — even though the underlying cash flows were identical by contract.
Worked example
Suppose Royal Dutch trades at $60 and Shell at $36. Fair-value parity says Royal Dutch should be 1.5 times Shell, i.e. $54. Royal Dutch is rich by $6, about 11%. A fund shorts $60 of Royal Dutch and buys $54 of Shell (1.5 shares of Shell scaled to match), fully hedged against the combined company's actual business risk. If the ratio corrects fully back to 1.50, the position captures the 11% gap regardless of which direction the overall market moves. The catch: the position can (and did, for LTCM in 1998) go further from parity before it converges, and the fund needs the capital and patience to hold through that widening — precisely the moment LTCM ran out of both.
What this means in practice
Royal Dutch and Shell finally unified into a single share class in 2005, permanently closing the trade. The episode is now the standard teaching example for why "the price is mathematically wrong" and "the price will correct soon" are different claims. An arbitrage can be certain in its endpoint and still unprofitable, or ruinous, if the capital backing it isn't large and patient enough to survive the path.
Don't confuse a hedged position with a safe one. Being short the rich leg and long the cheap leg removes exposure to the underlying business, but not exposure to the spread widening further before it narrows — and that widening is exactly what breaks undercapitalized arbitrageurs.
Related concepts
Practice in interviews
Further reading
- Froot, Dabora, 'How Are Stock Prices Affected by the Location of Trade?' (Journal of Financial Economics, 1999)
- Lowenstein, When Genius Failed