Why tax is a question for your own jurisdiction

What actually determines the answer 

The starting point is not the account. It is the person holding it, and specifically where that person is resident for tax purposes. Residence in this sense is a legal test rather than a description of where somebody feels settled, and each jurisdiction writes its own: a count of days present over a period, the location of a permanent home, the place where personal and economic ties are strongest, a separate domicile concept inherited from older law, or some combination of all of them. Those tests were not drafted to agree with one another, so a person can satisfy more than one at the same time, or move between them part way through a period. Where two states both claim somebody, the conflict is resolved between those states under arrangements neither the person nor the broker is a party to.

A second variable sits underneath the first. Even within one jurisdiction, identical closed positions can fall under different rules depending on how the activity is characterised: whether it is treated as occasional dealing by a private person or as a business carried on, which category of asset or contract the instrument is assigned to, whether anything is held through a company rather than personally, and how the relevant period is defined and when it ends. Characterisation is a legal conclusion drawn from facts about the person, and most of those facts are ones a broker never sees.

Why the firm on the other side cannot answer it 

A broker knows what happened on an account. It knows what was opened and closed, at what price and at what time, what was charged for it, and what was paid in and taken out. That is a complete record of one account and a small fraction of the file a tax question turns on. It does not show where the holder is resident this period or was resident in the last one, what other income or holdings exist, whether losses were carried in from an earlier period, whether this account is one of several, or whether any of it sits inside a structure rather than being held personally. An answer produced from the visible fraction would be an answer to a narrower question than the one being asked.

There is a second reason, and it is the same distinction this module drew earlier between executing an instruction and advising on one. Advising on a tax position is a professional service, delivered by people who are qualified in the jurisdiction whose law is being applied and answerable for the advice they give in it. A firm licensed to execute orders holds no such qualification and carries no such responsibility, so a confident answer from it would carry the appearance of authority without any of the substance behind it.

What an adviser asks for 

What the firm can produce is the input rather than the answer, and the earlier lesson on statements and records described what those documents are. From an adviser's side the framing is different. The adviser is reconstructing a period from primary documents and will not work from a figure that cannot be traced back to one, because a figure that cannot be traced cannot be defended if it is questioned. The reconstruction generally calls for the following.

• Account statements covering the whole period without gaps, as the firm issued them rather than as a screenshot or a copied figure.
• The closed position ledger: instrument, direction, size, opening and closing time, opening and closing price, and the result of each position.
• Every cost line charged during the period, separated by type, including commission and any financing adjustment on a position held past a daily cut off.
• Deposits and withdrawals with their dates and the accounts they moved between, so transfers can be separated from results.
• The currency the account is denominated in, and the rate and date of any conversion that was applied.
• The positions still open at the end of the period, with their opening details and their unrealised result as at that date.
• The terms in force during the period, since the basis on which costs were charged is part of the record rather than context for it.

Why the closing balance is not the figure 

The number most readily to hand is the one on the account screen, and it is the one least likely to answer anything. A balance at a point in time reflects everything that has moved through the account, money paid in and taken out included, and it makes no distinction between a result and a transfer. The arithmetic below separates them on assumed figures.

The eighth row is the one worth pausing on. In the adverse case the balance finished the period higher than it started while the trading produced a debit, because the deposits were larger than the loss. Read on its own, the balance said the opposite of what happened. Three separate quantities in the block have some claim to the word result and none of them is equal to another: the realised component before costs, the realised component after costs, and the equity figure, which folds in the unrealised result on positions still open at the period end.

Which of those a return calls for, on what basis, and whether costs are subtracted before or after, differs between regimes, and this page states the answer for none of them. The questions themselves are worth knowing even unanswered, because they determine which records matter: whether a result counts when a position closes or at a period end regardless of whether it closed, whether costs are recognised and which of them, whether a loss in one period bears on another at all, whether the location of the counterparty matters alongside the residence of the client, and how the account currency interacts with the currency a return is filed in. Each of those has a different answer in different places, and in several places the answer has changed within living memory.

Why generic tax content is worth little 

Tax pages published by brokers are common, and they share a structural problem rather than a drafting one. They are written for an audience whose residence is unknown, so every sentence in them is either general enough to decide nothing or specific enough to be wrong for most of the people reading it. A reader cannot tell from the page which of the two a given sentence is, because the page cannot tell either.

The problem compounds across time and across languages. Tax law changes on its own schedule, so a page that was accurate when it was written becomes a false statement in a later period without anything on it visibly changing. A site published in many languages multiplies that, because a correction made once is applied unevenly and a translated page can go on carrying a statement the source page no longer makes. The same failure runs through a confident answer in a forum from somebody whose residence, structure and prior periods are invisible. Confidence is not evidence, and the circumstances that would make an answer transferable from one person to another are precisely the ones nobody states.

Where practitioners disagree 

Whether a broker should publish tax content at all is genuinely contested. One view holds that readers will look somewhere regardless, that what they find elsewhere is usually worse, and that a carefully bounded general explainer carrying heavy qualification does more good than silence does. The other holds that any statement about tax coming from a counterparty is read as authoritative however it is qualified, that it decays silently across years and locales, and that the firm has an interest in the subject sounding simpler than it is. The position taken on this page is the second one, which is why it describes the shape of the question and the records that answer it and states no treatment.

The second disagreement is about when professional advice is proportionate. One tradition holds that advice on a small amount of occasional activity costs more than it saves, and that published official guidance is enough for a straightforward position in most places. Another points out that the characterisation of an activity is drawn from facts accumulating from the first period onward, so the decision about what to record is taken early whether or not anybody takes it deliberately, and reconstructing a period years afterwards from an incomplete record costs more than the advice would have. The threshold is disputed and this page does not set one. What neither side disputes is that the records are the input to the answer either way, and that a record not kept at the time is rarely recoverable later.