Much has been said concerning HMRC’s ability to treat a pension scheme as within a client’s estate on death based on the fact they had transferred within 2 years of death. Here we briefly discuss when this may apply.
There has been much discussion about HMRC’s ability to treat a pension scheme as within a client’s estate on death if they have transferred in the two years prior. Here we briefly look at when this may apply.
Historically, HMRC has tended to look at any transactions made within two years of the date of the death of a client with suspicion. For example the refusal to take benefits at selected retirement age may have been seen as an ‘omission to act’. The omission rules were designed to cover scenarios where a client was deemed to have deliberately deferred or amended benefits from a pension scheme due to ill health primarily to increase the death benefits that would be paid from the pension. The omission rulings were dropped in 2011 and so generally no longer apply – although, like all-things-Revenue, there may be exceptions!
Transfers within two year of death will be the main point that HMRC will look at. This is still very relevant – and possibly more so today with the increased flexibilities offered encouraging people to transfer. HMRC has determined that in many circumstances people have transferred benefits as they approach death for no other reason than to enhance the death benefits available (for example, a final salary scheme may only offer a spouse’s 50% annuity whereas a personal pension may offer a full return of fund). In HMRC’s view this may deprive them of funds which may end up within the client’s (or their spouse’s) estate.
More information can be found in this HMRC manual.
The details of the transfers need to be captured in form IHT409, which is a supplemental form to the IHT400 main form. This document asks about different scenarios, such as continued pension payments after death, lump sum payments and transfers. The transfer section requires notes to be filled in on the main IHT400 form and it appears to be from these notes that HMRC will decide whether or not there will be a deemed loss to the estate.
Unfortunately there is little detailed information about the full process and how they ultimately make a decision as to what is deemed to be inside or outside of the estate for IHT calculations.
A broad formula that has been muted in the press for a money purchase transfer could be;
- Look at the value of the original scheme, increase this value by investment growth for the longevity of the client's life expectancy to death and then discount this value back to today’s terms.
- You then look at the new pension scheme receiving the transfer and determine the value of the benefits the client could have immediately (so this works of retirement age cases only). This value would be based on tax free cash being taken and the residual fund. Traditionally this residual fund would have purchased an annuity with a 10 year guarantee. However, if the new scheme offers flexi access drawdown (FAD) they would use the actual fund value.
- In both cases of FAD and annuity it would be the net income figure that is used.
Taking the answer in B from the figure in A would give the deemed loss to the estate for IHT.
As part of these calculations there would be considerations for costs, discount rates, interest rates etc. which makes this a potentially very complex calculation.
This is not an HMRC published formula and should only be seen as a guide to the type of values HMRC could raise as losses to the estate for IHT.
This rule is primarily designed to catch those individuals who are looking to transfer benefits whilst in ill health for enhanced or more flexible death benefits. However, this may well also catch those individuals who are looking not so much to enhance benefits but rather get ‘all their ducks’ in a row before death so that payment of benefits can be seamless for their surviving spouse. For these reasons clients need to be very aware of the potential effects of death within two years of a pension transfer.