The PFIC Trap I Almost Walked Into: A US Architect’s Tokyo Tax Education

I have been building cloud infrastructure for fifteen years. The architectures I design are usually too complex to fit on a single whiteboard, but they are tractable. They have rules. The rules are documented. If something breaks, there is a runbook, or there will be one by Monday. This is, in my professional life, an environment I find comforting.

My tax life as an American expat in Tokyo, by contrast, is an environment in which the rules are documented but contradictory, the runbooks are written for a typical taxpayer who does not exist in my situation, and the cost of breaking something is not measured in downtime but in punitive multi-year retroactive penalties.

I do not find this comforting. I have, over the last several years, built my own runbook for it, and I am writing this piece because I wish someone had handed me a runbook of this kind eight years ago when I made the mistakes I am still cleaning up.

The largest mistake I made was buying a Japanese mutual fund through my Tokyo bank, in my first year here, on the recommendation of a banker who was being friendly and helpful and who did not know that the fund was, under US tax law, a Passive Foreign Investment Company.

PFICs are taxed under a regime that, if you do not elect into one of two specific alternative treatments, applies a punitive default that can result in effective marginal rates above 60% on appreciated gains, plus interest charges that compound over the holding period. The default treatment is, in my professional vocabulary, the IRS equivalent of shipping to production with no error handling and no logs.

I held the fund for three years before learning what it was. Unwinding it correctly, with the help of a US tax adviser who specialises in expat cases, cost me roughly a tenth of the position’s value in advisor fees and elective penalties. The adviser told me that the alternative, leaving it in place and continuing to hold under default treatment, would have cost considerably more over the next decade.

What finally helped me understand the structural problem, rather than just the line item I was paying for, was a deep dive into Form 8621 PFIC reporting for US expats. It walks through the mechanics of the qualified electing fund and mark-to-market elections in language that would survive a tax adviser’s reading. Most explanations of PFIC online are either too short to be operationally useful or too long to be readable. This one threads the needle.

Whenever I am thinking through the broader US-Japan compliance picture, I refer to a deep dive into the US-Japan tax treaty for remote workers. Treaty mechanics are dry. They are also, as anyone who has been double-taxed knows, the difference between a sustainable expat career and a slow financial bleed.

What I do now, operationally.

I do not buy any pooled investment vehicle that is not a US-domiciled fund without first running it through the PFIC test. The test, simplified, is that any non-US fund I might encounter is presumptively a PFIC unless proven otherwise. The exceptions are narrow. Acting as if everything non-US is a PFIC, then making exceptions when an adviser confirms otherwise, is a much safer default than the reverse.

I keep my investments primarily in US-domiciled funds held in US-domiciled accounts that I established before moving abroad. Some US brokers will close accounts of non-US-resident clients. Others will not. The handful that maintain accounts for US citizens abroad are worth their weight in administrative goodwill.

I file Form 8621 for every PFIC I hold, every year, even when the holdings are small enough that filing might not be strictly required. The cost of filing is annoyance. The cost of not filing is statute-of-limitations risk that does not fully expire until you have filed a return that includes the holding. The math, again, is in favour of filing.

I work with a tax adviser who specialises in US-Japan cases. I tried, for two years, to handle this myself with off-the-shelf software. The software produced returns that were technically filed but substantively incomplete in ways that I now know about because the adviser caught them on review. The cost of the adviser is real. The cost of the errors the adviser caught is much larger. I will not do my own returns again while I live abroad.

I read the source documentation. The IRS publishes the relevant forms and instructions. They are dense. They are also, with patience, comprehensible. Reading them once a year, even when the adviser is doing the actual filing, gives me the literacy to ask the right questions and to catch the rare cases where the adviser is operating from outdated guidance.

The thing I would tell any high-earning US citizen contemplating a move to Japan, or to any country with a complex bilateral tax relationship to the US, is that the cost of getting this right is real, and the cost of getting it wrong is much larger.

The work is not exciting. It is, in some sense, infrastructure for the rest of your life abroad. Build it carefully, document it as you go, and accept that the architecture of your finances has to be as deliberate as the architecture of the systems you ship for clients. Treat it that way and the next decade is a manageable engineering problem. Treat it casually and the IRS will, eventually, send you an outage report.