The Stock Bonus Isn't the Twist. Read the Clause Underneath It.

Buried in the same tentative agreement is a sentence almost nobody has quoted: if SK hynix posts a loss, part of employees' wages get deferred.

Not the bonus. Wages. Management and labour would negotiate how much base pay to defer in order to protect jobs and accelerate recovery, with the deferred amount paid as a lump sum once the company returns to profit. Put that next to the stock bonus and the deal stops being a compensation story. SK hynix has converted its workforce into a shock absorber for the memory cycle — geared to the upside through equity, geared to the downside through deferred pay.

That's a far more interesting thing than a big number.

First, where the number actually comes from

The headline figure needs a caveat, because it isn't a payout — it's a projection.

Projected 2026 operating profit 250 trillion won (~$179B) — some brokerage estimates

Profit-sharing formula 10% of annual operating profit

Resulting bonus pool ~25 trillion won

Employees ~35,000

Average pretax bonus ~700 million won (figures up to 779M won / $547K are circulating)

The year isn't finished. If operating profit lands below the brokerage estimate, the bonus falls with it — the formula is mechanical. Actual payouts also vary by job grade and individual performance, so "every employee gets $547,000" isn't right even if the average holds.

The structure — with one important correction

→ 40% cash, paid in the award year → 40% in SK hynix shares, sellable immediately → 20% in shares, distributed over the following two years → Plus a 6.3% wage increase, expanded pension benefits, shift allowances, housing loan support

The correction: 60% stock is the default, not a mandate. Employees can elect to take more in stock — up to 100% of the bonus. And for the first year of the new system, employees with genuine financial needs can opt for a cash payout.

That materially changes the "workers had this imposed on them" framing. There's an opt-out, and there's an opt-up.

Why the union actually objected

Here's the part that explains the pushback better than "stock is volatile."

Twelve months ago, SK hynix and the union overhauled this exact system. They scrapped the ceiling on payouts, adopted the 10%-of-operating-profit formula, and agreed to maintain it for ten years. Under that deal, 80% of the bonus was paid in cash in the award year, with 20% deferred — also in cash.

Now management wants 40% cash.

The union's stated objection wasn't that shares are risky. It was that changing the payment method undermines the intent of an agreement signed a year ago and meant to last a decade. That's a governance complaint, not a market-risk complaint, and it's a legitimate one.

The buyback and the bonus are the same instrument

This is the connection I'd most want you to see.

In the same window, SK hynix announced an enormous share buyback — reported in the tens of trillions of won, with an additional multi-billion-dollar tranche disclosed as the pay deal landed.

The bonus is paid in treasury shares. So the company repurchases its own stock — supporting the price — and then distributes that stock to employees instead of cash. The buyback and the compensation package are two halves of one transaction.

The effect: cash that would have left the business as bonuses stays on the balance sheet for fab construction and HBM capacity, while employees receive an asset whose price the company is actively supporting.

That is elegant. It is also worth being clear-eyed about: employees are being paid in an instrument at the top of the most violent cycle in semiconductors, in a year when the company expects record profit — which is precisely when the shares are most expensive.

The cycle is the whole point

Memory has always been the most brutally cyclical business in technology, and SK hynix knows it better than anyone: the company deferred wages during the 2023 downturn and has used voluntary unpaid leave in past troughs. Management called this its "crisis-overcoming DNA."

That history is why the new structure makes sense from the company's side. In an up-cycle, a cash-only bonus drains capital at exactly the moment you need it for capacity — and in memory, the winner of the next cycle is decided by capex committed during this one. In a down-cycle, deferred wages preserve jobs and cash.

The workforce now flexes with the cycle in both directions. For a company facing Samsung, Micron and a rapidly scaling CXMT, that flexibility is a genuine competitive asset.

Where I land

Both of the reasons you'd expect — preserving cash, aligning incentives — are real. But I'd frame it more bluntly: SK hynix has transferred a meaningful slice of cycle risk from shareholders to employees, and paid them extremely well for accepting it.

Whether that's a good trade depends entirely on where we are in the cycle. If HBM demand holds through 2028, the workers who took stock will look brilliant. If memory does what memory has always eventually done, they will have been paid at the peak in an asset that fell — and had their wages deferred on top.

The opt-out matters. So does the fact that the union still has to ratify this.

Question for you: would you take 60% of a career-defining bonus in your employer's stock, at the top of the best year your industry has ever had? The company is essentially offering employees a leveraged position on their own job security — and I'm not sure most people would price that correctly.

Address

Boston, Massachusets

Contacts

news@techproduit.com

Get your weekly newspaper