Revisiting Capacity for Loss In Retirement
There’s a lot of confusion around assessing capacity for loss, particularly for clients drawing down their portfolio. Much of it stems from the fact that the Regulator doesn’t really seem to know how it should be assessed.
There appears to be two camps in this debacle. On one side, you have an ex-regulator, paraplanners and compliance consultants who seem to think that capacity for loss is the be-all and end-all. We’ll call them the ‘compliance camp’. On the other hand, you have financial planners at the coal face. They know first hand all about the problem of trying to assess capacity for loss and reconcile this with real client goals. Both camps have very different — and hitherto irreconcilable — views on what capacity for loss means and how it should be assessed. (If you fancy a more satirical take on the subject, then read Nick Lincoln’s piece in the NMA.)
The term probably first came about in FG11/05, where it’s defined as follows:
‘By ‘capacity for loss’ we refer to the customer’s ability to absorb falls in the value of their investment. If any loss of capital would have a materially detrimental effect on their standard of living, this should be taken into account in assessing the risk that they are able to take.’
The phrase ‘capacity for loss’ doesn’t appear anywhere in COBS. Instead, COBS 9.2 uses the short-hand ‘ability to bear losses’.
Nowhere has the FCA described in detail, or prescribed how capacity for loss should be measured. And thank the good heavens for that! If it did, we’d have ended up in an even worse situation, where the Regulator effectively dictates client outcomes. On the other hand, this lack of clarity means the term is open to interpretation and has provided a stick for the compliance camp to beat planners with.
Trouble with terminology
The very term ‘capacity for loss’ itself is problematic. The compliance camp has taken it to mean that you should avoid investment risk at all costs. Even if there’s only a slight chance that investment loss will have a material impact on the client’s lifestyle. They’re content to leave inflation risk on the table and/or consign the client to a less affluent lifestyle in retirement. Just as long as they don’t take investment risk.
The very idea of quantifying, ‘how much a client can afford to lose without material impact on their lifestyle’ is nonsensical. Suppose we want to assess if a, ‘client can afford to lose X% of their portfolio without a material impact on their lifestyle’. We encounter a number of problems.
- What is X% and why is this particular figure loss relevant? Why is it 50%, not 40% or 75% or 100%? Clearly, we can’t just pluck this figure out of thin air. It has to be based on the asset class you’re invested in. And without the proverbial crystal ball, the only logical way to think about this is to look at the empirical behaviour of the underlying asset class, by asking for instance, ‘what is the historical worst case scenario for this portfolio?’ Of course, you can use all sorts of forward-looking assumptions to estimate this figure. But in reality, these are just guesses and can’t be any more valid than empirical evidence. In any case, the X% loss itself doesn’t tell us even half the story.
- The second problem is, the term ‘capacity for loss’ conjures an image of permanent irrecoverable loss. That doesn’t reflect the observed behaviour of mainstream asset classes. Overwhelming evidence shows that large losses on mainstream asset classes tend to be followed by even larger gains. In other words, capital market declines are often temporary.
The idea that we should contemplate and plan for permanent irrecoverable loss is lunacy. This is because the only logical conclusion from this ludicrous line of thinking is that a 40% or 50% or 100% irrecoverable loss in the capital markets would almost certainly mean that no form of retirement income is safe.
Annuity and DB pensions would almost certainly be bankrupt in these scenarios. So would the government. Of course, there are products that could result in 100% permanent loss to the clients (think Arch Cru, structured products etc). The Regulator has to set rules that account for the lowest common denominator but this is not what most financial planners invest their client portfolios in.
The point then is, capacity for loss can only be determined in the context of the empirical behaviour of the underlying asset class for the portfolio. In this regard, the historical worst-case scenario or even a 10th percentile scenario has to be sufficient. Again, the idea that we should consider scenarios that are worse than the historical worst-case for an asset class is ludicrous. If we do that for asset classes, we must do the same for annuities, DB and other income sources. And the only logical conclusion here is that nothing is safe and the client has no capacity for loss in the event of a doomsday scenario. This does nothing to aide financial planning in any way. We should all pack up and go home.
- There’s a third problem with the question, ‘can you afford X% drop in your portfolio without a material impact on your lifestyle?’ This is that the answer depends, not just on X (the depth of the loss), but on when (i.e. the timing of the loss) in our client’s retirement plan this event occurs and how long (the duration) it lasts.
A 50% decline in the second year of retirement that takes three years to recover has a radically different impact on the portfolio than a 50% decline in the 15th year that takes eight months to recover. This is another reason why using traditional cashflow that assumes straight-line linear return is flawed. Even if you build in losses into the cashflow projections, you can’t adequately account for the depth, timing and duration of the market declines — not to mention the recovery — in a way that reflects the empirical behaviour of the underlying asset classes.
The beauty of using extensive historical data in retirement planning is that you can run multiple scenarios (e.g 1,000 rolling periods). These help you assess the impact of the depth, timing and duration of losses on a client’s lifestyle, while also accounting for the recovery in a way that reflects the underlying behaviour of the asset classes.
- The term ‘capacity for loss’ doesn’t account for the alternative to taking risk. In other words, does the client have the capacity not to take risk? For someone in retirement, avoiding investment risk often means accepting the risk of running out of money and/or severe inflationary risk, which can do untold damage to their income. The FCA glossed over this in the (no 4) footnote in the FG11/05 paper above. But inflation is almost always guaranteed to damage the purchasing power of money. In addition, the risk of running out of money in cash is often greater than the risk of running out of money with an equities/bond portfolio. The upshot is, if you decide that a client has no capacity to bear equity risk and you recommend cash instead, you confine the client to penury in retirement. But never mind, you’ll have met the FCA’s capacity for loss requirement.
- Clients are often prepared to make material adjustments to their lifestyle if market conditions so dictate. This is Financial Planning 101 - we make ongoing changes to our savings or spending along the way. In fairness to COBS, it does say that advisers must take account of any ‘detrimental effect’ on the client’s lifestyle. Sadly, the compliance camp has taken this to mean that detrimental effect to the client’s lifestyle should be avoided at all costs. In reality, a client may be prepared to reduce their spending by say 10% or 20% during periods of extreme market decline to allow their portfolio to recover. To the extent that this is discussed, modelled and agreed with the client in advance, there’s nothing in COBS that prevents advisers from doing this.
All of this leads to the question, how do you adequately assess a client’s risk capacity in retirement? (I prefer the term risk capacity and I think we should bin the phrase ‘capacity for loss’. But I’m not holding my breath.) I approach the answer to this question through the prism of the two opposing retirement income frameworks.
- Safety-first: this is how many in the compliance camp tend to think of risk capacity, although they rarely articulate it so elegantly. This approach says, ‘all essential spending in retirement must be secured using only guaranteed income sources such as annuity, DB and State Pensions.’ Clients should not rely on investments for any of their essential income. Under this framework, risk capacity is black and white. If a client’s essential income requirement is fully covered by guaranteed sources of income, they have capacity to take investment risk. If it isn’t, they don’t have capacity to take risk and shouldn’t invest. The problem with this school of thought is that most clients can’t meet their desired goal without taking some risk. This approach would lead everyone, except the very wealthy, to buying an annuity in retirement or keeping most of their money in cash. It would invariably consign them to a lower standard of living in retirement. Under this approach, the only people who should transfer out of DB schemes are people who can afford their lifestyle without relying on the DB income.
- Probability-based: this approach assess risk capacity in the context of the client’s goals and the empirical behaviour of the underlying asset classes they invest in. Here, the question is, ‘given what we know about the empirical behaviour of this asset classes, will a severe scenario lead to a material reduction in lifestyle? If so, can the client cope with this?’ Or, ‘what kind of market scenario will result in a material impact to the client’s lifestyle?’ This is essentially what the sustainable withdrawal framework does. It assesses whether (near) worst-case scenarios for the underlying asset class result in material damage to the client’s lifestyle. If it does, you quantify this, assess if the client can live with this and pre-agree with clients before investing.
Katie is 66 and has a £500k retirement portfolio and a £300k house, which is paid for. She has just started to get her £8,500 a year State Pension. She reckons she needs a total of £25,000 (gross) a year to maintain her lifestyle, which means she needs £16,500 pa from her portfolio. Katie’s attitude to risk is ‘balanced’ but how do we calculate the beast that is capacity for loss?
- Mr Safety-First says: Katie relies on her retirement portfolio for 2⁄3 of her income. Therefore, any investment loss will have a material impact on her lifestyle. Accordingly, Katie should buy a secure income of £16.5K and only invest whatever is left. (Never mind that an index-linked annuity of £16.5k pa will likely require virtually all her savings. A level annuity of the same amount will require around 2/3rd of her fund but this will leave her exposed to inflationary risk in retirement.)
- Mrs Probability-Based says: Katie needs £16,500 from a portfolio of £500K, which is a withdrawal rate of 3.3%. This is net of total fees of 1.5%pa. Based on the Global Balanced Portfolio that matches Katie’s Risk tolerance, this withdrawal rate has a success rate of 97%. This is based on data that covers all the 888 30-year rolling periods since 1915. The historical worst-case scenario (red line in the chart below) is that this portfolio is exhausted at age 92. But that assumes there are no changes made to the plan along the way. The chart also highlights a particular scenario (in purple) in Jan 1973, when the client’s portfolio fell to £312k ( or £238K after accounting for inflation) by age 67, a decline of c40% in nominal terms (or c60% real terms)! As you can see, the model reflects the empirical behaviour of these asset classes, and the subsequent recovery. The scenario didn’t end up to be particularly disastrous.
So, without mentioning the horrible phrase ‘capacity for loss’, probability-based has considered the client’s ability to bear losses. The losses in question have been defined in terms of the empirical behaviour of the underlying asset classes that Katie invests in.
The point of all this is that risk capacity can be easily accounted for in the context of the client’s goal and behaviour of the underlying assets they invest in. The sustainable withdrawal framework accounts for this, without even having to mention the phrase. Thus financial planners adopting the framework do so as part and parcel of the planning process.