An unrealised gain is somebody's opinion of what an asset would fetch. A realised one is what it did fetch. The distance between them is where a great many returns go to die.
10 min read · Level 4
It is usually said that marks are a necessary approximation and that the distinction between paper and banked gains is a matter of discipline rather than of measurement. There is a stronger claim worth making, which is that an unrealised gain and a realised one are not the same quantity expressed with different confidence - they are different kinds of thing, and adding them into one total is a category error that the entire reporting convention is built on. A realised gain is a completed exchange between two parties. An unrealised gain is a forecast of an exchange that has not occurred, made under conditions that will not be the conditions of the actual sale, usually by a party with an interest in the answer. That they are added together at all is an accident of the need for a single figure, and the habit of reading that figure as wealth is what allows someone to be genuinely ruined by an investment that, on paper, never had a bad year. The practical test is simple and unfashionable: of the return you are proud of, how much of it has been paid to you.
A realised gain is the difference between what something sold for and what it cost. It is settled, it is in a bank account, and nothing can change it afterwards.
An unrealised gain is the difference between what something is currently marked at and what it cost. The mark is an estimate of what a buyer would pay. In public markets it is the last traded price, which is a real transaction but is a transaction in somebody else's shares, in whatever quantity they dealt. In private markets it is a valuation produced by a process, frequently by the same manager whose performance it describes.
Both belong in a return calculation and they are not the same kind of claim. One has been tested by a buyer; the other has been asserted. The distinction disappears entirely once both are added into a single figure, which is what almost every performance statement does.
Because decisions get taken on marks as though they were money. Position sizes are judged against a portfolio total that includes unrealised gains. Borrowing is arranged against marked values. And people conclude they can afford things on the strength of a number that has never been tested by a sale.
And because the gap is at its widest exactly when it matters most. A mark is most likely to be optimistic at the top of a cycle, when comparable transactions are plentiful and flattering, and hardest to realise at the bottom, when there are no buyers at any price near it. The number is least reliable precisely when a person most wants to rely on it.
Educational, not personalised financial advice. Nothing in the Investor Academy is a recommendation to buy, sell or hold anything — what is right for you depends on your own circumstances and the rules where you live.
Buy 10,000 shares at 12.00. The cost is 120,000.
The price reaches 31.00. The position is marked at 310,000 and the unrealised gain is 190,000. That figure is in every statement, it is arithmetically correct, and no transaction has occurred.
Sell 2,000 shares at 31.00. The proceeds are 62,000, the cost basis of those shares is 24,000, and the realised gain is 38,000. That 38,000 is permanent.
The price then falls to 14.50. The remaining 8,000 shares are worth 116,000 against a cost basis of 96,000, so the unrealised gain is now 20,000.
Where does that leave the whole position?
Banked: 62,000 Still held: 116,000 Total: 178,000 against a cost of 120,000 Total gain: 58,000, which is 48.33% on the original 120,000
A gain of 58,000 is a genuinely good outcome. But set it against the peak: the unrealised gain at 31.00 was 190,000, and what survived is 58,000. That is 30.53% of the paper gain; 69.47% of it was never anything at all.
And notice what the 310,000 mark actually claimed. It was 10,000 shares multiplied by a price at which somebody had traded some shares. It did not claim that 10,000 shares could be sold at 31.00, which is a different and much stronger statement - and in a less liquid holding, attempting it is what moves the price.
Private markets make the same point with the numbers further apart. An investor puts 25,000 into a company; a later funding round values it such that the position is marked at 3.0x, which is 75,000. The company is eventually sold at a price giving 1.4x, which is 35,000. The mark was not fraudulent - it was the price a real investor paid in a real round, for a small number of shares, under the conditions of that quarter, usually with liquidation preferences attached that the marked-up common shares did not have. The 40,000 difference is 53.33% of the mark, and it is the ordinary distance between a valuation and a sale rather than evidence that anybody did anything wrong.
The first mistake is spending against an unrealised gain - committing to something on the strength of a number that no buyer has agreed to.
The second is borrowing against a mark without asking what the mark survives. A decline that would be merely unpleasant for an unborrowed holder becomes a forced sale at the worst moment when the loan is secured against the marked value.
The third is reading a fund's total value multiple as a result. Cash returned and value asserted are two different claims, and only the first has been tested.
The fourth is treating the mark of an illiquid holding as a price you could obtain for your whole position. The last trade was at somebody else's size, and size is exactly what moves a thin market.
The fifth is the reverse error, which is genuinely common: refusing to count unrealised gains at all, and so holding a position long past the point where it should have been reduced, on the grounds that it 'is not real money yet'. Both halves belong in the picture. What must not happen is that they stop being distinguishable.
Realised and unrealised, as two lines that add to a total rather than as a total. In fund terms that is cash distributed against value asserted. The habit takes one line and it is what keeps the distinction alive at the moment a decision is taken.
A last traded price, a recent funding round, a manager's model, an appraisal. Write the date of the evidence too. A mark supported by a transaction eighteen months old is a statement about eighteen months ago.
The last trade happened in somebody's quantity, not yours. For any position large relative to the daily volume of its market, write down what you think could actually be sold in a week without moving the price, and use that as the working figure.
The point of distinguishing the two is not to ignore unrealised gains but to notice when a position has grown into something you would not choose to hold at today's price. Selling part of it is how a mark becomes money, and it is the only way it ever does.
Every time you read a performance figure, size a position, consider borrowing against holdings, or assess a private fund's progress.
Do not turn the distinction into a reason to sell everything that has risen. A mark that is holding up is information, not a warning, and a holder who realises every gain the moment it appears has swapped one error for a more expensive one - and has also introduced the tax consequences of a sale, which differ so completely between countries that no general statement about them belongs here.
In private portfolios a recognisable asymmetry appears in how valuations respond to events. A funding round at a higher price produces a prompt write-up, because there is a transaction to point to. A company that is struggling but has not yet raised produces no equivalent event, so its mark tends to persist at the last round's price until something forces a revision - frequently a down round or a sale. The consequence is not deception: it is that a portfolio's marked value incorporates good news quickly and bad news only when it becomes unavoidable, and this is at its most pronounced when funding conditions are turning.
Ask when each holding was last marked and what event caused it. A portfolio whose marks all date from a more favourable period is carrying a backlog of revisions that have not happened yet.
Create a free account to read the rest and apply it to your own investing.