There are two obvious places to find it. Each works, and each costs the client something they wanted to keep. Work through the choice, then the route the menu left out.
Illustrative historic case, anonymised. Not advice, and not a recommendation of any product. Whether anything like this is suitable depends entirely on individual circumstances.
The general investment account holds a structured product the client intends to keep to term. It is showing an unrealised loss of about £100,000. The pension holds about £100,000 in cash alongside other investments. The capital blocks on the map are drawn to scale.
The pension has the cash. Releasing it is the problem. Every pound out is taxable income.
Assumes 40% throughout. If the withdrawal takes adjusted net income above £100,000 the personal allowance tapers away, and above £125,140 the rate is 45%, so the gross needed rises. A first flexible withdrawal may be taxed on an emergency basis and reclaimed later.
The GIA produces the cash. The client stops owning the product.
The loss is not new. The value had already fallen. Selling crystallises the loss; it does not create it.
Choose one to see what it costs.
Each obvious answer solves one problem and creates another.
The mistake was treating the wrappers as separate decisions. The GIA has the asset and no cash. The SIPP has cash it cannot release without 40% tax. The answer needs both at once.
The broker sells at the secondary-market price, raising broadly the £60,000 the client needs. Because the product cost £160,000, the sale also turns the £100,000 fall in value into a realised capital loss, which can potentially be carried forward against future gains.
The loss already existed; the sale realises it. It is not cash and nothing of value moves. Whether it is an allowable capital loss depends on the product’s tax treatment, and it must be claimed within four years.
The trustees buy it back in the secondary market for about £60,000. No money leaves the pension, so no pension income is taxed. The client still owns the exposure. It now sits in the pension, at a cost of £60,000.
Two separate trades at market prices. Not a transfer between wrappers and not a new contribution. Dealing costs and spread apply on both sides.
Neither wrapper, on its own, held enough of the client’s position to produce the answer.
The trades were made because the client needed cash. The tax effects followed from meeting that need across the whole position rather than inside one account.
The product later returned about 230%. That outcome did not make the decision correct. The case is about seeing an option that account-by-account analysis missed, not about the result.
The implementation solved today’s problem. Did it necessarily produce the better lifetime tax outcome?
No. It depends on two things nobody could know on the day. Compare the third route with the obvious alternative that also keeps the product: withdraw from the SIPP and leave the product in the GIA. Follow both to the product’s maturity and to the client drawing the pension.
Value at the decision point to value at maturity. Even after a 230% gain, the third route is not automatically the better one.
One decision, four outcomes. The trades were the same in each.
Why the rate decides it. In the GIA, the first £100,000 of recovery only brought the product back to its £160,000 cost, so it was never taxed. In the SIPP the same recovery escapes CGT but is taxed as income when drawn. The realised loss can offset some of that disadvantage elsewhere, if it is allowable and ultimately used. Drawn at basic rate, the third route wins. Drawn at higher rate, the £40,000 saved on the day is given back and more.