Post
The S&P 500 fell 18% in 2022. For a Korean investor it fell 13%.
Buy a US stock from outside the US and you hold two positions: the stock, and the dollar. In 2022 the currency leg cushioned a Korean investor by five percentage points. In 2025 it took two and a half back. Here is why tools that report only in USD hide the second position entirely.
2022 is remembered as a bad year for US equities. The S&P 500 returned -18.1% on a total return basis, -19.4% on price alone. Nearly every year-end recap ran those numbers.
For an investor who earns and spends in Korean won, the S&P 500 returned -12.9% that year. For a Japanese investor, -6.2%. Same index, same period, same dividends, and losses that differ by a factor of three.
The difference is currency. Over 2022 the dollar rose 6.0% against the won, from 1,188.59 to 1,260.18, and 14.5% against the yen, from 115.17 to 131.81. The dollar earned back a large part of what the equities lost.
That is not luck. It is structure. And there is a good chance your tools are not showing it to you.
Buying a US stock creates two positions
When you buy an S&P 500 ETF from a won-denominated account, two things happen at once. You sell won and buy dollars. Then you use those dollars to buy equity.
So what you hold is not "US equities." It is US equities multiplied by the dollar. Your realized return is the product of two legs:
local return = (1 + equity return) x (1 + FX move) - 1
For a Korean investor in 2022: (1 - 0.181) x (1 + 0.060) - 1 = -13.2%. The published won-denominated S&P 500 total return index came in at -12.89%, so the arithmetic holds. The 0.3 point gap is just which FX series you use.
You never placed an order for that second position. There was no research, no target weight, no rebalancing rule. And it is exactly the same size as the first one. Put $100,000 into US equities and you have put $100,000 into the dollar.
It helped in 2022 and it charged you in 2025
Concluding that currency is a permanent cushion gets you exactly halfway. The sign flips.
2025 was a good year: the S&P 500 returned +17.9% total return. But the won strengthened against the dollar that year, from 1,477.86 to 1,444.55, a 2.25% appreciation. Measured in won, the same index returned +15.35%. Two and a half points went to the currency leg.
| Year | S&P 500 (USD) | S&P 500 (KRW) | What FX did |
|---|---|---|---|
| 2022 | -18.1% | -12.9% | +5.2pp cushion |
| 2023 | +26.3% | +28.6% | +2.3pp tailwind |
| 2025 | +17.9% | +15.4% | -2.5pp drag |
Two years of help, one year of drag, and a swing of two to five percentage points every single year. If you are choosing between a 0.03% expense ratio and a 0.09% one, there is a three-point variable running in the background that nobody is measuring.
Japan makes the point louder: -6.2% in 2022, +34.9% in 2023. A weakening yen turned US equities into something close to a different asset class for a Japanese holder.
The reason it is invisible is that your tools speak dollars
Most US investing apps and calculators were built for US users. In that world the base currency is the dollar, and the dollar does not move by definition. So the field labeled "return" gets a dollar return and the job is done. For that user, it is correct.
If you retire in won, pay rent in won, and buy groceries in won, that field is not the answer to your question. It is somebody else's answer.
Three places where the gap gets expensive:
Cost basis. A share bought in 2020 at 1,080 won per dollar and the same share bought in 2024 at 1,470 have completely different won-denominated cost bases. Both may show "+30%" in USD while one is +77% and the other is +28% in won. Convert at today's rate instead of the trade-date rate and that distinction disappears silently.
Aggregating everything. To put a Seoul apartment, a won deposit, a US ETF, and a Japanese stock on one screen, you have to unify the currency. A sum without that unification is not a number. And once you do unify, which date's rate you used changes the answer.
Long-horizon projections. If your 30-year net worth projection has no FX assumption anywhere in it, you have assumed the rate is frozen for 30 years. USD/KRW has traveled between roughly 1,080 and 1,478 in the last five.
Is this an argument for hedging?
No. This is not a pitch for the ticker with an H on the end. Hedging has a cost, that cost is roughly the interest rate differential between the two countries, and you do not control it. For some people hedging is right and for others it is not.
The argument is about the step before that. You cannot decide whether to hedge something you have never measured.
Three questions come first:
- What percentage of my net worth is denominated in dollars?
- What is my cost basis in my own currency, not in dollars?
- Of last year's return, how many points came from the equities and how many from the currency?
Answer those three with actual numbers and you have earned the right to have the hedging conversation. If you cannot, then hedged or unhedged, you are simply accepting a default.
How Opula handles currency
Opula settles this with a three-line rule.
On input, it always asks which currency the number is in. Then it converts to USD at the historical rate for that transaction's date before storing. It does not convert a 2020 purchase at today's rate. If the historical rate is not cached, it raises an error and tells you to sync FX first rather than quietly substituting the current rate.
Internally, everything is USD. There is no point in the system where mixed currencies get added together.
On output, it converts to whichever currency you ask for. Ask for won and your net worth, cost basis, P&L, and dividends all come back in won. And when a value genuinely cannot be converted, it says so instead of rendering a zero. That last part matters most: not fabricating a number is table stakes, telling you that you were not given one is the actual feature.
Assets that US tools do not recognize at all, like a Korean apartment, a jeonse deposit, KRX physical gold, or a single Japanese stock, pass through the same rule. The unit of account is the currency, not the asset's nationality.
The short version
If you hold US equities and live in won, your real return is the product of two numbers. One of them is in the news every day. The other moves your account by several percentage points a year and appears nowhere.
In 2022 it cut a Korean investor's loss by five points. In 2025 it took two and a half points off the gain. Either way, you did not choose it.
You should at least know how big it is.